2009

• A. Anderson, “Virtually There: Digital Manufacturing Simulates Survival in a Bear Market,” Automation World, pp. 40, 2009.
  [Bibtex]

  @article{
  Anderson09,
     Author = {Anderson, Alex},
     Title = {Virtually There: Digital Manufacturing Simulates Survival in a Bear Market},
     Journal = {Automation World},
     Pages = {40},
     Month = {April 2009},
     Abstract = {Proponents of the technology insist that despite the high cost of admission, digital manufacturing is just the ticket to help lead manufacturers out of the economic downturn.
  Dick Slansky is the first person to admit that he was a little “optimistic” when making his predictions about the adoption rate that digital manufacturing, and virtual commissioning in particular, would see over the past year. “It hasn’t grown as quickly as I thought it would,” concedes Slansky, senior PLM and Discrete Manufacturing analyst at ARC Advisory Group Inc., in Dedham, Mass.
  
  A little over a year ago, he predicted the overall product lifecycle management (PLM) market would grow at a compounded annual growth rate of 14.5 percent and swell from $7.5 billion in 2006 to more than $14.6 billion by 2011. While he is still putting together his updated estimates to include 2007 and 2008 numbers, it is already clear that those numbers are too ambitious. And Slansky wasn’t alone in his optimism. Industry analyst firm CIMdata, which specializes in PLM, was even more bullish, predicting that the digital manufacturing tools market would more than double between 2006 and 2011, growing from an estimated $436 million to more than $1 billion.
  
  “Several factors have constrained the growth, not the least of which are the economy and the relatively high cost of the solution,” says Slansky. Nevertheless, interest is there, and growth has been steady, despite the fact that a full deployment can often cost more than $100,000 per seat. More to the point, Slansky insists that rather than being an albatross dragging a company down, digital manufacturing and virtual commissioning can make a real impact on a company’s survival prospects in the current economic climate.
  
  “The inclination is to hunker down, cut R&D, cut spending. Innovation is going to be the difference between surviving or not as we come out of the downturn. And engineering reform is still going to be a key difference maker.”
  
  Traditionally, mechanical engineers and control engineers didn’t work together. The mechanical engineer would get the model of the finished product, design the manufacturing line to produce it and then call in the control engineer to create the code that would run the line. As the old military cliché goes, no plan survives contact with the enemy—or in this case, the production floor—and finding the design did not work as advertised was commonplace. Dozens of different factors such as collisions between robots, bad ergonomics or having the wrong machines altogether, could rise up to snafu the plan, which then had to be reworked or otherwise fixed. Since this was done in the real world, it inevitably meant metal had been bent, which had to be unbent, which in turn is time consuming and expensive. Digital manufacturing and virtual commissioning promised to do away with all that, and, according to Slansky and a host of others, it’s starting to keep that promise.
  
  Slansky defines digital manufacturing as the integration of digital models of product designs with manufacturing processes. The manufacturing engineer uses powerful 3D simulation tools to create a virtual model of all the production processes. Once complete, the virtual environment can be used to validate, synchronize and optimize the production process—to make sure it works as advertised.
  
  Virtual commissioning is that element of digital manufacturing whereby all elements of a production line are computer-simulated prior to physical set-up. Using detailed mechanical CAD models that include the conveyor systems, robotic work cells, packaging equipment and the control systems, programmable logic controllers (PLCs), motors, sensors and drives, the entire production line, or even the entire factory, can be modeled and run to see if it will work in real life the way it’s supposed to.
  
  “They can then use advanced simulation tools to validate that all of these components are synchronized, controlled, and will function in harmony to operate as intended and physically produce the product,” wrote Slansky in his January 2009 report: “The Real Story on Virtual Commissioning: Is the Technology Ready to Speed New Product Introduction?”
  
  Additionally, “Engineers can use similar tools to design and program the control logic and pre-validate the connections, logic, and dynamics of the planned physical installation.  Virtual commissioning eliminates many of the time- and resource-consuming tasks that engineers must otherwise perform on real physical equipment prior to its use.”
  
  The promised benefits can be eye opening. According to CIMdata, organizations using digital manufacturing can expect to see lead time to market reduced by an average of 30 percent, the number of design changes reduced by 65 percent and time spent in the manufacturing planning process reduced by 40 percent. On the other side of the equation, production throughput can be increased by an average of 15 percent and overall production costs can be cut by as much as 13 percent.
  
  Writes Slansky, “Not only does this significantly reduce the product launch lifecycle, it also allows engineers to optimize these processes before they are physically installed. By integrating the digital product design with a virtually designed manufacturing environment, it is possible to modify either the product or the production process/equipment without incurring prohibitive costs or delays...virtual commissioning provides real and tangible benefits and addresses one of manufacturing’s fundamental issues: getting product to market in a timely and competitive manner.”
  
  Virtual commissioning
  
  By its nature, a lot of the conversations surrounding virtual commissioning deal with the setting up of new production lines/facilities. Considering the economy, there may not be a lot of “new” set-ups taking place, but Slansky says that virtual commissioning pays off any time a company needs to reengineer a line, and is arguably even more important in such circumstances.
  
  Given the dour economy, he says, you may have situations in which you are doing limited production runs and need to reconfigure the product lines more frequently, or you may have to consolidate production from three plants into two, without abandoning products. Virtual commissioning shines in such cases, helping you be more flexible and reducing downtime.
  
  Stefan Linner, head of marketing for Siemens PLM Software’s Tecnomatix product line, agrees. “Even when companies close down factories and lines, they still need to manufacture product, just in different quantities. If they condense into one location, they may have to switch over existing capacity. If you reduce the number of plants, you may create bottlenecks in existing plants. Digital manufacturing helps avoid that.
  
  “From a high level, some people may think if the manufacturing/production goes down, the planning goes down too, but I think the exact opposite is the case,” Linner adds. “You have to be even more rigorous in planning to eliminate any inefficiencies in your process.”
  
  Siemens PLM Software (formerly UGS Corp.), a Plano, Texas-based division of Siemens’ Automation and Drives Group, is one of only two vendors offering a comprehensive factory simulation and modeling solution to the market today: the other is the French-based PLM supplier Dassault Systèmes, with its Delmia product line. Dassault has announced a series of relationships with plant floor automation vendors, including Rockwell Automation Inc., Milwaukee, to integrate Delmia directly with their PLC’s to smooth rollouts and enhance collaboration at an earlier stage in the production process. While at this time last year, both vendors were busily integrating their partnerships and acquisitions and promising future solutions, now they are rolling those solutions out to the market.
  
  “The savings are tremendous. We are talking millions of dollars,” says Jordan Merhib, sales manager at Applied Manufacturing Technologies Inc. (AMT), an Orion, Mich.-based engineering services company. “For example, one of the big three auto manufacturers had gone out to their usual line builders to automate the end of a press line. They wanted to automate the unloading of that line, and the speed and volume made it very complicated. The bids that came in were all over the map in terms of price and time. They came to us and asked us to look at these five proposals and figure out why they were so different. We simulated each one and saw that none of them would work. We modeled one that would work and they took that and included it in an updated RFP (request for proposals). They figured that we saved them more than $3 million on that one line.”
  
  AMT provides project resources, project engineering, project management, consulting and engineering outsourcing to clients. Like virtual commissioning, which took off in the automotive sector first and is expanding to other vertical industries, AMT started in automotive and has seen its client list expand to include aerospace, consumer packaged goods, industrial machinery and alternative energy.
  
  “When someone wants to put in a factory or configure a line, the best approach is to do it in the virtual world first and figure out the answers to any questions before you start cutting metal,” says Merhib, adding that the simulation group has become a “huge” part of AMT’s business. “Back in the day, people had to do things the hard way: on a drafting table, and with real-world trial and error. Now, for the most part, for anyone who comes to us to help them figure out how to do something, we will use some kind of simulation software before embarking on the work.
  
  “Despite the state of the economy, this continues to be an important and thriving part of our business,” Merhib continues. “People may not be buying a lot of new equipment, but they are trying to get more efficiency out of the equipment they have and consolidating production to conserve costs.”
  
  AMT, which uses Siemens PLM’s Technomatix for its simulation work, isn’t the only engineering services company to see the area as important. Fulda, Germany-based EDAG GmbH & Co. is implementing a Delmia automation solution from Dassault Systèmes.
  
  “The ability to virtually optimize and validate control behavior in a production cell is particularly advantageous for us as an engineering partner with our own associated line builder,” says Karina Schäfer, head of the digital factory at EDAG. “It lets us move from the construction site into the office, which in turn, creates better working conditions.”
  
  More to the point, the way engineering services firms are getting involved makes it more likely that some of the smaller manufacturers and integrators in the market may get involved in what is still a pricy area. Says AMT’s Merhib, “For us, the costs get amortized, while our customers get the full benefits today.”
  
  Subdued enthusiasm
  
  While somewhat subdued in his predictions, Slansky hasn’t lost any of his enthusiasm for the potential impact that digital manufacturing and virtual commissioning can have on businesses, despite the up-front investment. “Manufacturers are going to need technology like this. They may have to make some tech investments up front, but it will pay off for them in the long term.”
  
  “Given the intense competition for market share and the absolute necessity to get products to market in time to hit the market window for consumer acceptance, companies are looking for any competitive edge,” Slansky concluded in his January report. “Streamlining the design, validation, and commissioning of production systems will be a key factor for manufacturers in meeting requirements for cost-effective and efficient product launches and hitting market windows. Virtual commissioning and other digital manufacturing tools have emerged at precisely the right time in a global manufacturing climate where accelerating time to market and implementing manufacturing process optimization will be critical to a company’s success.”  },
        Year = {2009} }

• Y. Lei, D. Djurdjanovic, L. G. Barajas, G. Workman, S. R. Biller, and J. Ni, “DeviceNet Network Health Monitoring Using Physical Layer Parameters,” Journal of Intelligent Manufacturing, pp. 8, 2009.
  [Bibtex]

  @article{
  Yong08,
     Author = {Lei, Yong and Djurdjanovic, Dragan and Barajas, Leandro G. and Workman, Gary and Biller, Stephan R. and Ni, Jun},
     Title = {DeviceNet Network Health Monitoring Using Physical Layer Parameters},
     Journal = {Journal of Intelligent Manufacturing},
     Pages = {8},
     Abstract = {Since the 1980s, the manufacturing environment has seen the introduction of numerous sensor/actuator bus and network protocols for automation systems, which led to increased manufacturing productivity, improved inter- changeability of devices from different vendors, facilitated flexibility and reconfigurability for various applications and improved reliability, while reducing installation and maintenance costs. However, such heightened manufacturing integration facilitated by industrial networks also leads to dramatic consequences of improper or degraded network operation. This paper presents a novel Network Health Management system that provides diagnostic and prognostic information for DeviceNet at the device and network level. It extracts features from analog waveform communication signals, as well as logic and timing features from digital packet logs. These features are used to evaluate the network system performance degradation by applying multidimensional clustering techniques. In addition, this work proposes a hybrid prognostics structure using combined physical and logic layer features to provide fault location information that cannot easily be realized with analog or digital data independently. Furthermore, an intermittent connection diagnostic algorithm which analyzes patterns of interrupted and error packets on the network was developed. This tool can be used as a packet source identification method which uses joint analog features and digital information inferred from analog waveforms. A test-bed was constructed and the experiments of network impedance mismatch, cable degradation, and intermittent connections were conducted in laboratory environment. Experimental results show that the proposed system can detect degradations of the network and identify the location of the intermittent connection successfully. Field tests performed in an industrial environment were also conducted and their results are discussed.},
     Keywords = {Network fault diagnosis, Field buses, DeviceNet, Industrial networks, Fault diagnosis, Network automation systems, Network instrumentation },
     Year = {2009} }

2008

• L. G. Barajas, M. B. Egerstedt, E. W. Kamen, and A. Goldstein, “Stencil printing process modeling and control using statistical neural networks,” IEEE Transactions on Electronics Packaging Manufacturing, vol. 31, iss. 1, pp. 9-18, 2008.
  [Bibtex]

  @article{
  Barajas+NN08,
     Author = {Barajas, Leandro G. and Egerstedt, Magnus B. and Kamen, Edward W. and Goldstein, Alex},
     Title = {Stencil printing process modeling and control using statistical neural networks},
     Journal = {IEEE Transactions on Electronics Packaging Manufacturing},
     Volume = {31},
     Number = {1},
     Pages = {9-18},
     Abstract = {This paper presents a neural network model for the Stencil Printing Process (SPP) in Surface Mount Technology (SMT) manufacturing of Printed Circuit Boards (PCB). A practical model description that decomposes the overall steady-state process in independently modelled subspaces is provided. The neural network model can be updated in real-time procuring a method to control the process by dynamically searching the optimal set point of the control variables. The optimization is performed by minimizing the mean squared error with respect to the desired solder brick height; furthermore, in the case when multiple solutions exist, the set point that yields the lowest variance is used. The process simulator is mainly suitable for offline testing and debugging of more complex closed-loop control algorithms for the SPP optimization providing a common and realistic framework for algorithm performance evaluation. The soundness of this paper is based on the fact that the estimation of the sampled moments of the probability distributions is made using a statistically significant number of data samples from each board, for each component type, for each printing direction, and for each pad orientation.},
     Keywords = {Stencil printing process model, statistical neural networks, surface mount technology, closed loop control},
     Year = {2008} }

2004

• X. Xu, “A method based on linguistic aggregation operators for group decision making with linguistic preference relations,” Information Sciences, vol. 166, pp. 19-30, 2004.
  [Bibtex]

  @article{
  Xu04,
     Author = {Xu, Xeshui},
     Title = {A method based on linguistic aggregation operators for group decision making with linguistic preference relations},
     Journal = {Information Sciences},
     Volume = {166},
     Pages = {19-30},
     Abstract = {In this paper, we define some operational laws of linguistic variables and develop some new aggregation operators such as linguistic geometric averaging (LGA) operator, linguistic weighted geometric averaging (LWGA) operator, linguistic ordered weighted geometric averaging (LOWGA) operator and linguistic hybrid geometric averaging (LHGA) operator, etc., which can be utilized to aggregate preference information taking the form of linguistic variables, and then study some desirable properties of the operators. Based on the LGA and the LHGA operators, we propose a practical method for group decision making with linguistic preference relations. The method is straightforward and has no loss of information. Finally, an illustrative numerical example is also given.},
     Keywords = {group decision making, linguistic preference relations, aggregation operators, operational laws},
     Year = {2004} }

2003

• L. G. Barajas and M. B. Egerstedt, “Automatic generation of weighted quality characteristic for multi-output systems,” Document in preparation, 2003.
  [Bibtex]

  @article{
  Barajas+Egerstedt03,
     Author = {Barajas, Leandro G. and Egerstedt, Magnus B.},
     Title = {Automatic generation of weighted quality characteristic for multi-output systems},
     Journal = {Document in preparation},
        Year = {2003} }

• Z. Ling Chunxian, D. Milos, W. Martin, and H. Christopher, “Fine pitch stencil printing using enclosed printing systems,” Soldering & Surface Mount Technology, vol. 15, iss. 1, pp. 43-49, 2003.
  [Bibtex]

  @article{
  Ling03,
     Author = {Ling Chunxian, Zou and Milos, Dusek and Martin, Wickham and Christopher, Hunt},
     Title = {Fine pitch stencil printing using enclosed printing systems},
     Journal = {Soldering & Surface Mount Technology},
     Volume = {15},
     Number = {1},
     Pages = {43 - 49},
     Abstract = {Enclosed print heads have recently been developed as an improvement on the traditional squeegee methods for solder paste printing. They offer the opportunity of widening the printing process window and reducing process waste. Consequently, this work was undertaken to evaluate some aspects of enclosed print head printing, and it has been shown to be a robust process. A number of performance factors were established: with increased humidity the paste degradation was limited due to its sealed paste reservoir; the system also permitted successful intermittent printing over a 5 day period; printing is much more tolerant to distorted substrates than some squeegee blades, and hence improves printing on non-planar surfaces; significant reduction in paste wastage occurs, since paste ageing is reduced.},
     Keywords = {Stencils, Solder Pastes, Assembly},
     Year = {2003} }

2002

• D. W. Coit, B. T. Jackson, and A. E. Smith, “Neural network open loop control system for wave soldering,” Journal of Electronics Manufacturing, vol. 11, iss. 1, pp. 95-105, 2002.
  [Bibtex]

  @article{
  Coit02,
     Author = {Coit, D. W. and Jackson, B. T. and Smith, A. E.},
     Title = {Neural network open loop control system for wave soldering},
     Journal = {Journal of Electronics Manufacturing},
     Volume = {11},
     Number = {1},
     Pages = {95-105},
     Abstract = {This paper describes the development of neural networks for the prediction of (1) printed circuit card surface temperature during a wave soldering process, and (2) the quality level of the circuit card assembly soldered connections. Using a combination of production data and design of experiment data, a set of hierarchically connected neural networks were developed and validated. These networks predict thermal behavior of a printed circuit card assembly at various points in the solder process based on process settings and circuit card design data. Then these predictions are used as inputs, together with other parameters, to estimate the quality of solder connections. The system can be used to decrease the number of solder connection defects, reduce set-up and preparation time between lots, and lead to consistent, repeatable process settings without trial production runs or operator tuning efforts. For the wave soldering process studied, this is especially important since the batch size is quite small, quality demands are stringent, and process settings are changed frequently C5290. C7410D. C7480.},
     Keywords = {electronic engineering computing. neurocontrollers., printed circuit manufacture. . process control., production engineering computing. quality control., statistical analysis. wave soldering},
     Year = {2002} }

• R. Durairaj, G. J. Jackson, N. N. Ekere, G. Glinski, and C. Bailey, “Correlation of solder paste rheology with computational simulations of the stencil printing process,” Soldering & Surface Mount Technology, vol. 14, iss. 1, pp. 11-17, 2002.
  [Bibtex]

  @article{
  Durairaj02,
     Author = {Durairaj, R. and Jackson, G. J. and Ekere, N. N. and Glinski, G. and Bailey, C.},
     Title = {Correlation of solder paste rheology with computational simulations of the stencil printing process},
     Journal = {Soldering & Surface Mount Technology},
     Volume = {14},
     Number = {1},
     Pages = {11-17},
     Abstract = {Soldering technologies continue to evolve to meet the demands of the continuous miniaturisation of electronic products, particularly in the area of solder paste formulations used in the reflow soldering of surface mount devices. Stencil printing continues to be a leading process used for the deposition of solder paste onto printed circuit boards (PCBs) in the volume production of electronic assemblies, despite problems in achieving a consistent print quality at an ultra-fine pitch. In order to eliminate these defects a good understanding of the processes involved in printing is important. Computational simulations may complement experimental print trials and paste characterisation studies, and provide an extra dimension to the understanding of the process. The characteristics and flow properties of solder pastes depend primarily on their chemical and physical composition and good material property data is essential for meaningful results to be obtained by computational simulation.This paper describes paste characterisation and computational simulation studies that have been undertaken through the collaboration of the School of Aeronautical, Mechanical and Manufacturing Engineering at Salford University and the Centre for Numerical Modelling and Process Analysis at the University of Greenwich. The rheological profile of two different paste formulations (lead and lead-free) for sub 100 micron flip-chip devices are tested and applied to computational simulations of their flow behaviour during the printing process.____________Solder paste characterization and rheological studies coupled with computational simulations offer an insight into the mechanisms of the stencil printing process that would otherwise not be available. The comparisons between a lead and lead-free paste sample presented here are only one example of the vast characterization studies possible. For example, the same procedures can be applied to the comparison of samples with varying rheological properties of the flux system whilst maintaining the sample particle material and size distribution or vice –versa. Further rheological testing and development of the computational models is required however to account for microscopic phenomena that may have a significant influence on the paste behavior at a macroscopic scale. This includes the phenomenon of wall slip where flux rich regions adjacent to a solid surface form a lubricating layer, reducing or eliminating the shear rate of the solid particles. The application of these results at a macroscopic scale to the boundary conditions of aperture filling and emptying simulations at a microscopic scale will further enhance the understanding of the critical parameters and conditions that will be required to maintain stencil printing as a leading solder deposition process for the continued miniaturization of electronic components,},
     Keywords = {Solder Pastes, Rheology, Stencils},
     Year = {2002} }

• A. Johnson and B. Boyes, “Stencil Printing Basics: Equipment Materials and Process Requirements,” Chip Scale Review, iss. January/February, 2002.
  [Bibtex]

  @article{
  Johnson02,
     Author = {Johnson, Alden and Boyes, Bob},
     Title = {Stencil Printing Basics: Equipment Materials and Process Requirements},
     Journal = {Chip Scale Review},
     Number = {January/February},
     Abstract = {Optimizing fine-pitch electronics assembly requires a careful review of the complete process. Each variation of device, lead spacing, substrate, material and build-rate schedule requires specific parameters to ensure high yields. Additionally, selecting the best combination of materials, equipment and data analysis systems can greatly reduce development time and costs. In stencil printing, defects are typically caused by one or more of the following: poor alignment between the substrate and stencil, incorrect paste chemistry or variations in the amount of paste deposited. To eliminate defects, the capabilities of the printing equipment and the materials selected (substrate, paste type and stencil design) should be examined closely.},
        Year = {2002} }

• S. H. Mannan, “Solder paste reflow modeling,” Soldering & Surface Mount Technology, vol. 14, iss. 1, pp. 18-23, 2002.
  [Bibtex]

  @article{
  Mannan02,
     Author = {Mannan, S. H.},
     Title = {Solder paste reflow modeling},
     Journal = {Soldering & Surface Mount Technology},
     Volume = {14},
     Number = {1},
     Pages = {18-23},
     Abstract = {Solder paste printing and reflow are well established processes for producing solder joints in electronic assemblies. Solder paste consists of a dense suspension of solder particles in a liquid medium (vehicle) that acts as an oxide reducing agent (flux) during reflow, cleaning the metal surfaces of oxides. This paper reports on attempts to model the physical and chemical processes occurring during solder paste reflow using computational fluid dynamics (CFD). Axisymmetric, two-dimensional and three-dimensional models are described, and a method of reproducing oxide-like behaviour in these models is introduced},
     Keywords = {assembling. computational fluid dynamics. oxidation., printed circuit manufacture. . reflow soldering. surface, chemistry. surface cleaning. surface contamination., surface mount technology},
     Year = {2002} }

• V. Torra, “Learning weights for the quasi-weighted means,” Fuzzy Systems, IEEE Transactions on, vol. 10, pp. 653-666, 2002.
  [Bibtex]

  @article{
  Torra02,
     Author = {Torra, V.},
     Title = {Learning weights for the quasi-weighted means},
     Journal = {Fuzzy Systems, IEEE Transactions on},
     Volume = {10},
     Pages = {653-666},
     Abstract = {In this paper, we study the determination of weights for quasi-weighted means (also called quasi-linear means) when a set of examples is given. We consider first a simple case, the learning of weights for weighted means, and then we extend the approach to the more general case of a quasi-weighted mean. We consider the case of a known arbitrary generator . This paper finishes considering the use of parametric functions that are suitable when the values to aggregate are measure values or ratio ones.},
     Keywords = {Aggregation operators, data fusion, parameter determination, quasi-linear means, quasi-weighted means, weighted mean},
     Year = {2002} }

2001

• L. G. Barajas, E. W. Kamen, and A. Goldstein, “On-line enhancement of the stencil printing process,” Circuits Assembly, vol. 12, iss. 3, pp. 32-36, 2001.
  [Bibtex]

  @article{
  Barajas01,
     Author = {Barajas, Leandro G. and Kamen, Edward W. and Goldstein, Alex},
     Title = {On-line enhancement of the stencil printing process},
     Journal = {Circuits Assembly},
     Volume = {12},
     Number = {3},
     Pages = {32-36},
     Month = {March},
     Abstract = {A assembly lines are related to the stencil printing process, and approximately 30 percent of the total manufacturing cost is due test and rework expenses. Many factors the quality of the stencil printing process, including squeegee pressure, printing speed, solder paste viscosity and rheology, and air and humidity within the production environment. Efforts have been made to enhance the of the printing process by using feedback control, where automatic optical inspection (AOI) measurements of deposited solder paste bricks are used to adjust machine settings such as squeegee pressure and printing speed. However, the success that has achieved in using on-line feedback control existing surface-mount lines appears to be rather limited. Some new stencil printers do include closed-loop controller to maintain desired squeegee pressures and printing speeds in printing directions.However, these controllers do not ensure that the deposited bricks will have desired characteristics such as the correct.},
     Keywords = {defects related SPP, fuzzy logic, SPC, neural network, assembling, automatic optical inspection, fuzzy control, fuzzy logic, printed circuit manufacture, printed circuit testing, process control, production testing, soldering},
     Year = {2001} }

• R. Durairaj, T. A. Nguty, and N. N. Ekere, “Critical factors affecting paste flow during the stencil printing of solder paste,” Soldering & Surface Mount Technology, vol. 13, iss. 2, pp. 30-34, 2001.
  [Bibtex]

  @article{
  Durairaj01,
     Author = {Durairaj, R. and Nguty, T. A. and Ekere, N. N.},
     Title = {Critical factors affecting paste flow during the stencil printing of solder paste},
     Journal = {Soldering & Surface Mount Technology},
     Volume = {13},
     Number = {2},
     Pages = {30 - 34},
     Abstract = {The paste printing process accounts for the majority of assembly defects, and most defects originate from poor understanding of the effect of printing process parameters on the printing performance. As the current product miniaturisation trend continues, area array type package solutions are now being designed into products. The assembly of these devices requires the printing of very small solder paste deposits. The printing of solder pastes through small stencil apertures typically results in stencil clogging and incomplete transfer of paste to the PCB pads. At the very narrow aperture sizes required for flip-chip applications, the paste rheology becomes crucial for consistent paste withdrawal. This is because, for smaller paste volumes, surface tension effects become dominant over viscous flow. Proper understanding of the effect of the key material, equipment and process parameters, and their interactions, is crucial for achieving high print yields. During the aperture filling and emptying sub-process, the solder paste experiences forces/stresses as it interacts with the stencil aperture walls and the pad surfaces, which directly impact the paste flow within the apertures. As the substrate and stencil separate, the frictional/adhesive force on the stencil walls competes directly with the adhesives/pull force on the PCB pads, often resulting in incomplete paste transfer or skipping/clogged apertures. In this paper, we investigate the effect of stencil design on the printing process and in particular the effect on paste transfer efficiency.____________This paper has presented results of a study of the effect of stencil design on the printing process and, in particular, the effects on paste transfer efficiency. Proper understanding of the effects of aperture design is essential because, during the aperture filling and emptying sub-processes , the solder paste experiences forces/stresses as it interacts with the stencil aperture walls and the pad surfaces which directly impact the paste flow within the apertures. The analysis shows that, as the substrate and stencil separate, the frictional/adhesive forces on the stencil walls compete directly with the adhesives/pull force on the PCB pads, often resulting in incomplete paste transfer or skipping/ clogged apertures. This means that the aspect/area ratio is an important factor in addressing the challenges of printing solder pastes at ultra-fine geometries such as are required for flip-chip technology.The results show that for good paste transfer the aspect and area ratios should be higher than 1.25 and above 0.6 respectively, as lower paste height deposits were obtained for apertures with aspect and area ratios below 1.25 and 0.6 respectively. Experimental observation also showed that skipping occurs mainly for the apertures with ratios well below these figures, but the more dominant parameter appears to be the area ratio. At very small geometries, such as those now being used for flip chip applications, the stencil design should be based on both the area/aspect ratios to ensure efficient paste transfers and consistency in paste deposit thickness and volume from pad-to-pad, and from board-to-board. It should also be noted that the paste particle size and other paste properties directly impact on paste transfer, and further improvements in paste formulation will help improve printing performance if these ratios can be lowered.},
     Keywords = {Solder Paste, Stencil Printer, Viscosity, Pastes},
     Year = {2001} }

• N. N. Ekere, D. He, and L. Cai, “The influence of wall slip in the measurement of solder paste viscosity,” Components and Packaging Technologies, IEEE Transactions on [see also Components, Packaging and Manufacturing Technology, Part A: Packaging Technologies, IEEE Transactions on], vol. 24, iss. 3, pp. 468-473, 2001.
  [Bibtex]

  @article{
  Ekere01,
     Author = {Ekere, N. N. and He, D. and Cai, L.},
     Title = {The influence of wall slip in the measurement of solder paste viscosity},
     Journal = {Components and Packaging Technologies, IEEE Transactions on [see also Components, Packaging and Manufacturing Technology, Part A: Packaging Technologies, IEEE Transactions on]},
     Volume = {24},
     Number = {3},
     Pages = {468-473},
     Abstract = {The wall slip phenomena is known to have a significant effect on the measurement of the viscosity of dense suspensions. In the measurement of the viscosity of solder pastes the effect of wall slip is such that the measured viscosity (also called the apparent viscosity) is much lower than the true viscosity of the paste. Therefore, correction needs to be applied to the measured viscosity in order to obtain the true viscosity of the solder paste. In this paper, we present work on the modeling of the influence of wall slip on viscosity measurement, and a model for predicting the true viscosity based on measurements using parallel plate viscometer. The apparent viscosity values measured at two different plate gaps, but at the same applied shear rate (also called the apparent shear rate), is used for predicting the true viscosity, the wall slip velocity and the thickness of the boundary slip layer. The model was validated using results from solder paste samples measured at three different plate gaps (H=0.5 mm, 1.0 mm and 1.5 mm). Our results show that the predicted values of the true viscosity using the data measured at any two gaps are in reasonably good agreement. The results also show that the influence of the wall slip is significant and that the ratio of the predicted viscosity to the apparent viscosity decreases with increasing apparent shear rate.A model for predicting the true viscosity of solder pastes has been presented in this paper. The apparent viscosities measured at different gaps, but at the same applied shear rate was used for predicting the true viscosity, the wall slip velocity and the thickness of the boundary layer. The model was validated by viscosity measurements on a solder paste. Results showed that the predicted values for the true viscosity using the data measured at any two gaps are in reasonably good agreement. Our measurements and predicted true viscosity values also showed that, first, as the apparent shear rate increases the slip velocity increases; whilst the ratio of the true viscosity to the apparent viscosity decreases. Secondly, at a small gap the influence of the wall slip on the viscosity measurement is more significant than that at a larger gap. For solder paste with solid volume fraction phi=0.5, the thickness of the slip layer was estimated in the range of 0.15a to 0.19a. These results are in good agreement with those reported by other workers. The model presented in this paper has very wide applicability, and can also be used for correcting for the effect of wall slip in viscosity measurements for other suspensions such as conductive adhesives and under-fills used in the assembly of electronic devices.},
     Keywords = {reflow soldering, surface mount technology, printed circuit manufacture, slip flow, viscometers, wall slip, solder paste viscosity, dense suspensions, apparent viscosity, parallel plate viscometer, plate gaps, shear rate, boundary slip layer, 0.5 to 1.5},
     Year = {2001} }

• K. P. Ellis, F. J. Vittes, and J. E. Kobza, “Optimizing the performance of a surface mount placement machine,” Electronics Packaging Manufacturing, IEEE Transactions on [see also Components, Packaging and Manufacturing Technology, Part C: Manufacturing, IEEE Transactons on], vol. 24, iss. 3, pp. 160-170, 2001.
  [Bibtex]

  @article{
  Ellis01,
     Author = {Ellis, K.P. and Vittes, F.J. and Kobza, J.E.},
     Title = {Optimizing the performance of a surface mount placement machine},
     Journal = {Electronics Packaging Manufacturing, IEEE Transactions on [see also Components, Packaging and Manufacturing Technology, Part C: Manufacturing, IEEE Transactons on]},
     Volume = {24},
     Number = {3},
     Pages = {160-170},
     Abstract = {Process planning is an important and integral part of effectively operating a printed circuit board (PCB) assembly system. A PCB assembly system generally consists of different types of placement machines, testing equipment, and material handling equipment. This research develops a new solution approach to determine the component placement sequence and feeder arrangement for a turret style surface mount-placement machine often used in PCB assembly systems. This solution approach can be integrated into a process planning system to reduce assembly time and improve productivity. The algorithm consists of a construction procedure that uses a set of rules to generate an initial component placement sequence and feeder arrangement along with an improvement procedure to improve the initial solution. An industrial case study conducted at Ericsson, Inc., using a Fuji CP4-3 machine and actual PCB data, is presented to demonstrate the performance of the proposed solution approach. The solutions obtained using the proposed solution approach are compared to those obtained using state of the art PCB assembly process optimization software. For all PCBs in the case study, the proposed solution approach yielded lower placement times than the commercial software, thus generating additional valuable production capacity. This research is applicable for both researchers and practitioners in printed circuit board assembly systems.},
     Keywords = {assembly planning, printed circuit manufacture, surface mount technology, optimisation, assembly time, process planning, PCB assembly system, component placement sequence, feeder arrangement, turret style surface mount-placement machine, productivity, c},
     Year = {2001} }

• S. L. Ho, M. Xie, L. C. Tang, K. Xu, and T. N. Goh, “Neural network modeling with confidence bounds: a case study on the solder paste deposition process,” Electronics Packaging Manufacturing, IEEE Transactions on [see also Components, Packaging and Manufacturing Technology, Part C: Manufacturing, IEEE Transactons on], vol. 24, iss. 4, pp. 323-332, 2001.
  [Bibtex]

  @article{
  Ho+Xie01,
     Author = {Ho, S.L. and Xie, M. and Tang, L.C. and Xu, K. and Goh, T.N.},
     Title = {Neural network modeling with confidence bounds: a case study on the solder paste deposition process},
     Journal = {Electronics Packaging Manufacturing, IEEE Transactions on [see also Components, Packaging and Manufacturing Technology, Part C: Manufacturing, IEEE Transactons on]},
     Volume = {24},
     Number = {4},
     Pages = {323-332},
     Abstract = {The formation of reliable solder joints in electronic assemblies is a critical issue in surface mount manufacturing. Stringent control is placed on the solder paste deposition process to minimize soldering defects and achieve high assembly yield. Time series process modeling of the solder paste quality characteristics using neural networks (NN) is a promising approach that complements traditional control charting schemes deployed on-line. We present the study of building a multilayer feedforward neural network for monitoring the solder paste deposition process performance. Modeling via neural networks provides not only useful insights in the process dynamics, it also allows forecasts of future process behavior to be made. Data measurements collected on ball grid array (BGA) and quad flat pack (QFP) packages are used to illustrate the NN technique and the forecast accuracies of the models are summarized. Furthermore, in order to quantify the errors associated with the forecasted point estimates, asymptotically valid prediction intervals are computed using nonlinear regression. Simulation results showed that the prediction intervals constructed give reasonably satisfactory coverage percentages as compared to the nominal confidence levels. Process control using NN with confidence bounds provides more quality information on the performance of the deposition process for better decision making and continuous improvement.The drive toward product miniaturization and the concomitant decrease in component lead pitch has increased the importance of monitoring the amount of printed solder paste in the deposition process. In this paper, we presented the study of NN in modeling and predicting the behavior of solder paste deposition process. This approach can lead to better decision making and continuous improvement. Results from our studies showed that the feed-forward NN architecture exhibits low prediction errors, and hence is a viable alternative compared to physical models. Furthermore, the use of prediction intervals is proposed, in addition to the forecasted point estimates. We demonstrated that constructing the intervals via nonlinear regression techniques showed reasonably good coverage probabilities. Due to this forecasting capability, integrating the neural models with confidence bounds for continuous process monitoring in a close loop framework indeed provides quality information on the performance of the deposition process. The time series solder paste height data can be learned by the NN to detect occurrence of decreasing trend. Hence, early detection of process deterioration is made possible and timely remedial actions can be initiated.The drive toward product miniaturization and the concomitant decrease in component lead pitch has increased the importance of monitoring the amount of printed solder paste in the deposition process. In this paper, we presented the study of NN in modeling and predicting the behavior of solder paste deposition process. This approach can lead to better decision making and continuous improvement. Results from our studies showed that the feed-forward NN architecture exhibits low prediction errors, and hence is a viable alternative compared to physical models. Furthermore, the use of prediction intervals is proposed, in addition to the forecasted point estimates. We demonstrated that constructing the intervals via nonlinear regression techniques showed reasonably good coverage probabilities. Due to this forecasting capability, integrating the neural models with confidence bounds for continuous process monitoring in a close loop framework indeed provides quality information on the performance of the deposition process. The time series solder paste height data can be learned by the NN to detect occurrence of decreasing trend. Hence, early detection of process deterioration is made possible and timely remedial actions can be initiated.},
     Keywords = {reflow soldering, multilayer perceptrons, ball grid arrays, packaging, statistical analysis, surface mount technology, process monitoring, printed circuit manufacture, feedforward neural nets, neural network modeling, confidence bounds, solder paste dep},
     Year = {2001} }

• R. Prasad, “Step 4 — Printing,” Surface Mount Technology, iss. April, 2001.
  [Bibtex]

  @article{
  Prasad01,
     Author = {Prasad, Ray},
     Title = {Step 4 — Printing},
     Journal = {Surface Mount Technology},
     Number = {April},
     Abstract = {In the reflow soldering of surface mount assemblies, solder paste connects the leads or terminations of surface mount components to the lands. There are many variables in this process paste, screen printer, paste application method and printing process. In printing solder paste, the substrate is placed on the work holder, held firmly mechanically or by vacuum, and aligned with the aid of tooling pins or vision. A screen or stencil is used to apply solder paste. This article, after briefly describing the paste printing process, will focus on some key issues in squeegee and stencil materials and then discuss printing processes for fine-pitch and through-hole components in a mixed surface mount assembly.To achieve good printing results, a combination of the right paste material (right rheology, i.e., viscosity, metal content, largest powder size and lowest flux activity possible for the application), the right tools (printer, stencil and squeegee blade) and the right process (good registration and clean sweep) are necessary.  Metal squeegees and stencils are the most commonly used printing tools even though they are slightly more expensive than other alternatives. As the industry moves to finer pitches, laser cut and electroformed stencil use, again more expensive options, are becoming more common.  Even the best pastes, equipment and application methods alone cannont ensure acceptable results. The user must control process and equipment variables to achieve good print quality. This is even more critical in fine-pitch printing because it requires very accurate stencil aperture alignment to the land patterns. Finally, there are times when it may be more cost effective to reflow solder through-hole components if they can withstand the reflow temperature.},
        Year = {2001} }

• L. Tenorio, “Statistical Regularization of Inverse Problems,” SIAM Review, vol. 43, iss. 2, pp. 347-366, 2001.
  [Bibtex]

  @article{
  Tenorio01,
     Author = {Tenorio, Luis},
     Title = {Statistical Regularization of Inverse Problems},
     Journal = {SIAM Review},
     Volume = {43},
     Number = {2},
     Pages = {347-366},
     Abstract = {In experimental sciences we often need to solve inverse problems. That is, we want to obtain information about the internal structure of a physical system from indirect noisy observations. Often the problem is not whether a solution exists; on the contrary, there are too many solutions that .t the data to a chosen tolerance level. The goal is to use prior information to determine a physically meaningful solution. Here, we present some of the basic questions that arise. We describe methods that can be used to find inversion estimates as well as ways to assess their performance.},
     Keywords = {Fredholm integral equations, Backus–Gilbert method, reproducing kernels, Tikhonov regularization, splines, cross-validation, wavelets, minimax estimation},
     Year = {2001} }

• Y. Wei, S. B. Gelfand, and J. V. Krogmeier, “Noise-constrained least mean squares algorithm,” Signal Processing, IEEE Transactions on, vol. 49, iss. 9, pp. 1961-1970, 2001.
  [Bibtex]

  @article{
  Wei01,
     Author = {Wei, Yongbin and Gelfand, S.B. and Krogmeier, J.V.},
     Title = {Noise-constrained least mean squares algorithm},
     Journal = {Signal Processing, IEEE Transactions on},
     Volume = {49},
     Number = {9},
     Pages = {1961-1970},
     Abstract = {We consider the design of an adaptive algorithm for finite impulse response channel estimation, which incorporates partial knowledge of the channel, specifically, the additive noise variance. Although the noise variance is not required for the offline Wiener solution, there are potential benefits (and limitations) for the learning behavior of an adaptive solution. In our approach, a Robbins-Monro algorithm is used to minimize the conventional mean square error criterion subject to a noise variance constraint and a penalty term necessary to guarantee uniqueness of the combined weight/multiplier solution. The resulting noise-constrained LMS (NCLMS) algorithm is a type of variable step-size LMS algorithm where the step-size rule arises naturally from the constraints. A convergence and performance analysis is carried out, and extensive simulations are conducted that compare NCLMS with several adaptive algorithms. This work also provides an appropriate framework for the derivation and analysis of other adaptive algorithms that incorporate partial knowledge of the channel.},
     Keywords = {least mean squares methods, adaptive signal processing, transient response, convergence of numerical methods, random processes, time-varying channels, AWGN, noise-constrained least mean squares algorithm, adaptive algorithm design, finite impulse respon},
     Year = {2001} }

• H. K. Yii, N. Morad, and M. S. Hitam, “Optimisation of a Solder Paste Printing Process Parameters Using a Hybrid Intelligent Approach,” Neural Network World, vol. 11, iss. 2, pp. 109-127, 2001.
  [Bibtex]

  @article{
  Yii+Morad+Hitam01,
     Author = {Yii, Hee Kim and Morad, Norhashimah and Hitam, Muhammad Suzuri},
     Title = {Optimisation of a Solder Paste Printing Process Parameters Using a Hybrid Intelligent Approach},
     Journal = {Neural Network World},
     Volume = {11},
     Number = {2},
     Pages = {109-127},
     Abstract = {This paper describes a method of modelling and optimising the solder paste printing process using an artificial intelligence approach. A hybrid approach combining the backpropagation neural network and genetic algorithm to model and subsequently optimise the process is developed using actual data collected from a manufacturing plant. Results obtained showed that the neural network developed was able to model the process successfully and the genetic algorithm developed was able to optimise the process parameters using various optimisation criteria.},
     Keywords = {Solder paste printing process, neural networks, genetic algorithms, bootstrap method, multiple regression method, multi-objective optimisation},
     Year = {2001} }

2000

• L. Godo and V. Torra, “On aggregation operators for ordinal qualitative information,” Fuzzy Systems, IEEE Transactions on, vol. 8, pp. 143-154, 2000.
  [Bibtex]

  @article{
  Godo00,
     Author = {Godo, L. and Torra, V.},
     Title = {On aggregation operators for ordinal qualitative information},
     Journal = {Fuzzy Systems, IEEE Transactions on},
     Volume = {8},
     Pages = {143-154},
     Abstract = {In many fuzzy systems applications, values to be aggregated are of a qualitative nature. In that case, if one wants to compute some type of average, the most common procedure is to perform a numerical interpretation of the values, and then apply one of the well-known (the most suitable) numerical aggregation operators. However, if one wants to stick to a purely qualitative setting, choices are reduced to either weighted versions of max-min combinations or to a few existing proposals of qualitative versions of ordered weighted average (OWA) operators. In this paper, we explore the feasibility of defining a qualitative counterpart of the weighted mean operator without having to use necessarily any numerical interpretation of the values. We propose a method to average qualitative values, belonging to a (finite) ordinal scale, weighted with natural numbers, and based on the use of finite t-norms and t-conorms defined on the scale of values. Extensions of the method for other OWA-like and Choquet integral-type aggregations are also considered},
     Keywords = {fuzzy systems},
     Year = {2000} }

• L. Li and P. Thompson, “Stencil printing process development for flip chip interconnect,” Electronics Packaging Manufacturing, IEEE Transactions on [see also Components, Packaging and Manufacturing Technology, Part C: Manufacturing, IEEE Transactons on], vol. 23, iss. 3, pp. 165-170, 2000.
  [Bibtex]

  @article{
  Li00,
     Author = {Li, L. and Thompson, P.},
     Title = {Stencil printing process development for flip chip interconnect},
     Journal = {Electronics Packaging Manufacturing, IEEE Transactions on [see also Components, Packaging and Manufacturing Technology, Part C: Manufacturing, IEEE Transactons on]},
     Volume = {23},
     Number = {3},
     Pages = {165-170},
     Abstract = {Traditional methods for forming flip chip interconnection include evaporation and electroplating. While both options have proven performance records, their costs are too high for many of today's cost-sensitive applications. Motorola's Interconnect Systems Laboratory (ISL) developed a low cost flip chip interconnect alternative, which electroless plates Ni/Au as the under bump metallurgy (UBM), deposits solder paste using a stencil or photoresist mask, and reflows the paste to form solder bumps. This paper is focused on the stencil print process development for wafer solder bumping. Solder paste selection, characterization, stencil design, and process parameter optimizations are critical factors for successful fine pitch stencil printing. Fine mesh eutectic solder pastes (Type 5: -500+635, Type 6: -635) with different flux vehicles were evaluated for printability and reflow studies. Paste and flux modification and co-development with vendors for this specific application were conducted. An optimal paste, with specified viscosity and desired print, reflow, and cleaning behavior, was developed. Wafer stencil design rules were established to deposit the right amount of solder paste to form the required reflowed solder bump height without bridges between pads. Printing and reflow design of experiments were performed to establish the baseline and optimum process parameters. Solder bump characterizations, including bump height and uniformity, composition, shear force, and scanning electron microscope (SEM) bump profile and cross section microstructure analysis, were conducted. Development and characterization results are reported. This stencil print solder bump process is demonstrated on 4, 5, 6, and 8-in device wafers with pitches down to 200 /spl mu/m. Solder bumps are formed on wafers without bridges or missing bumps. Wafer level yield is >95% on die basis. Bump height standard deviation within die is less than 3%, and range is less than 15% of the average bump height. The 63 Sn/37 Pb eutectic-bumped functional device dice were flip chip assembled to test boards with Cu/Ni/Au pad metallization, and underfilled for interconnect reliability studies. All reliability requirements were met.},
     Keywords = {printing, flip-chip devices, masks, reflow soldering, fine-pitch technology, viscosity, surface cleaning, integrated circuit reliability, integrated circuit yield, integrated circuit interconnections, chip scale packaging, stencil printing process devel},
     Year = {2000} }

• K. Lin and S. Yao, “Solder thickness variation with respect to soldering parameters,” Components and Packaging Technologies, IEEE Transactions on [see also Components, Packaging and Manufacturing Technology, Part A: Packaging Technologies, IEEE Transactions on], vol. 23, iss. 4, pp. 661-664, 2000.
  [Bibtex]

  @article{
  Lin00,
     Author = {Lin, Kwang-Lung and Yao, S.},
     Title = {Solder thickness variation with respect to soldering parameters},
     Journal = {Components and Packaging Technologies, IEEE Transactions on [see also Components, Packaging and Manufacturing Technology, Part A: Packaging Technologies, IEEE Transactions on]},
     Volume = {23},
     Number = {4},
     Pages = {661-664},
     Abstract = {This study investigated the growth behavior of solder interconnect on solder bonding pad. The solder thickness was investigated with respect to parameters including bonding pad area, reflow temperature, solder paste volume, loading, and die size (metallization area). Solder thickness was found to have a maximum value with respect to the ratio of solder paste volume to pad size. Loading and the reflow temperature showed no effect on solder thickness. An enlargement in metallization area reduces the solder thickness, while greater solder paste volume increases the solder thickness.The solder thickness obtained from the reflow of solder paste is related to the solder pad area on the substrate, as well as to the metallization area of the die. A maximum solder thickness is achieved with respect to the ratio of solder paste volume/pad area, and the solder thickness increases linearly with respect to increasing ratio below 0.004. The three factors that are important in controlling the solder thickness are pad area, solder volume, and die size (metallization area), rather than reflow temperature or die weight.},
     Keywords = {reflow soldering, integrated circuit packaging, lead bonding, solder thickness variation, soldering parameters, growth behavior, bonding pad area, reflow temperature, solder paste volume, loading, die size, metallization area},
     Year = {2000} }

• O. Stan and E. W. Kamen, “A local linearized least squares algorithm for training feedforward neural networks,” Neural Networks, IEEE Transactions on, vol. 11, iss. 2, pp. 487-495, 2000.
  [Bibtex]

  @article{
  Sta+Kam00,
     Author = {Stan, O. and Kamen, Edward W.},
     Title = {A local linearized least squares algorithm for training feedforward neural networks},
     Journal = {Neural Networks, IEEE Transactions on},
     Volume = {11},
     Number = {2},
     Pages = {487-495},
     Abstract = {In training the weights of a feedforward neural network, it is well known that the global extended Kalman filter (GEKF) algorithm has much better performance than the popular gradient descent with error backpropagation in terms of convergence and quality of solution. However, the GEKF is very computationally intensive, which has led to the development of efficient algorithms such as the multiple extended Kalman algorithm (MEKA) and the decoupled extended Kalman filter algorithm (DEKF), that are based on dimensional reduction and/or partitioning of the global problem. In this paper we present a new training algorithm, called local linearized least squares (LLLS), that is based on viewing the local system identification subproblems at the neuron level as recursive linearized least squares problems. The objective function of the least squares problems for each neuron is the sum of the squares of the linearized backpropagated error signals. The new algorithm is shown to give better convergence results for three benchmark problems in comparison to MEKA, and in comparison to DEKF for highly coupled applications. The performance of the LLLS algorithm approaches that of the GEKF algorithm in the experiments.},
     Keywords = {feedforward neural nets, linearisation techniques, least squares approximations, Kalman filters, filtering theory, convergence, computational complexity, backpropagation, local linearized least squares algorithm, feedforward neural network training, wei},
     Year = {2000} }

• V. Torra, “The WOWA operator and the interpolation function W*: Chen and Otto’s interpolation method revisited,” Fuzzy Sets and Systems, vol. 113, iss. 3, pp. 389-396, 2000.
  [Bibtex]

  @article{
  Torra00,
     Author = {Torra, Vicenç},
     Title = {The WOWA operator and the interpolation function W*: Chen and Otto's interpolation method revisited},
     Journal = {Fuzzy Sets and Systems},
     Volume = {113},
     Number = {3},
     Pages = {389-396},
     Abstract = {Chen and Otto introduced in 1995 an interpolation method for defining membership functions from a set of points. In this paper we report some errors found in that algorithm and introduce some new conditions on the boundaries that are more general than the ones in the original work. We construct, in this way, a new interpolation method and prove that it has a smooth shape on the boundaries.  This study is related to the WOWA operator, an aggregation function that generalizes both the weighted mean and the OWA operator. WOWA uses an interpolation method to calculate a vector of weights as a combination of other two vectors. The new method is introduced to implement this combination.},
     Keywords = {WOWA operator, membership functions},
     Year = {2000} }

• Z. Uykan, C. Guzelis, M. E. Celebi, and H. N. Koivo, “Analysis of input-output clustering for determining centers of RBFN,” Neural Networks, IEEE Transactions on, vol. 11, pp. 851-858, 2000.
  [Bibtex]

  @article{
  Uykan00,
     Author = {Uykan, Z. and Guzelis, C. and Celebi, M.E. and Koivo, H.N.},
     Title = {Analysis of input-output clustering for determining centers of RBFN},
     Journal = {Neural Networks, IEEE Transactions on},
     Volume = {11},
     Pages = {851-858},
     Abstract = {The key point in design of radial basis function networks is to specify the number and the locations of the centers. Several heuristic hybrid learning methods, which apply a clustering algorithm for locating the centers and subsequently a linear least-squares method for the linear weights, have been previously suggested. These hybrid methods can be put into two groups, which will be called as input clustering (IC) and input-output clustering (IOC), depending on whether the output vector is also involved in the clustering process. The idea of concatenating the output vector to the input vector in the clustering process has independently been proposed by several papers in the literature although none of them presented a theoretical analysis on such procedures, but rather demonstrated their effectiveness in several applications. The main contribution of this paper is to present an approach for investigating the relationship between clustering process on input-output training samples and the mean squared output error in the context of a radial basis function network (RBFN). We may summarize our investigations in that matter as follows: (1) A weighted mean squared input-output quantization error, which is to be minimized by IOC, yields an upper bound to the mean squared output error. (2) This upper bound and consequently the output error can be made arbitrarily small (zero in the limit case) by decreasing the quantization error which can be accomplished through increasing the number of hidden units},
     Keywords = {heuristic programming, learning (artificial intelligence), least squares approximations, minimisation, pattern clustering, radial basis function networks},
     Year = {2000} }

1999

• R. L. Mahajan, “Neural nets for modeling, optimization and control in semiconductor manufacturing,” Proceedings of SPIE – The International Society for Optical Engineering, vol. 3812, pp. 176-187, 1999.
  [Bibtex]

  @article{
  Mahajan99,
     Author = {Mahajan, Roop L.},
     Title = {Neural nets for modeling, optimization and control in semiconductor manufacturing},
     Journal = {Proceedings of SPIE - The International Society for Optical Engineering},
     Volume = {3812},
     Pages = {176-187},
     Abstract = {This paper provides an overview of our recent work in the development of neural network models for optimization and control of electronic manufacturing processes. The concepts of physical-neural network models and model transfer are described and demonstrated to be effective in building accurate neural network models economically. Process diagnostic techniques using multiple neural networks are reviewed and shown to be accurate for fault diagnosis. Finally, recent strategies in integration of statistical and neural network tools for process control are discussed. Several examples from electronics manufacturing such as chemical vapor deposition and fine pitch stencil printing are described to illustrate application of the basic concepts discussed.},
     Keywords = {Mathematical models, Semiconductor device manufacture,, Optimization, Process control, Failure analysis, Chemical vapor, deposition, Electronic manufacturing processes, Process diagnostic technique,, Fault diagnosis, C 723.4 Artificial Intelligence, C 9},
     Year = {1999} }

• J. V. Stevenson and D. Drabenstadt, “Stencil printer optimization study,” Surface Mount Technology, vol. 13, iss. 11, pp. 58, 60, 62-3, 1999.
  [Bibtex]

  @article{
  Stevenson99,
     Author = {Stevenson, J. V. and Drabenstadt, D.},
     Title = {Stencil printer optimization study},
     Journal = {Surface Mount Technology},
     Volume = {13},
     Number = {11},
     Pages = {58, 60, 62-3},
     Abstract = {The solder paste printing process is perhaps the most critical operation for assembling a PCB populated with surface mounted components. Industry experts agree that solder defects are caused, in part, by poorly printed solder paste. As many as 60% of soldering defects can be traced to improper deposition. Therefore, accuracy and repeatability of the stencil printer is highly desirable throughout the entire printing process. There are many variables that can effect the formulation of a solder paste pad. Some industry experts suggest that there are a minimum of 39 process variables that must be controlled. By reducing the variation in the printed paste pad area, the opportunity exists for achieving a six-sigma process. To optimize the stencil printing process, the six-sigma philosophies of reducing process variation and targeting an average response were used to complete this task. The six-sigma tool, a mathematically sound statistical technique, was used to accomplish this undertaking. After mapping out the entire printing operation, a list of factors were generated for the semiautomatic printer and solder paste used for this experiment. The semiautomatic printer used on the line had a manual setting with a gauge for the print pressure setting. The solder was a high-metal-content, fine-pitch mesh paste},
     Keywords = {assembling, circuit reliability, design of experiments, fine-pitch technology, optimisation, printed circuit manufacture, printed circuit testing, soldering, surface, mount technology, defects related SPP, Critical Variables},
     Year = {1999} }

• R. R. Yager and D. P. Filev, “Induced ordered weighted averaging operators,” Systems, Man and Cybernetics, Part B, IEEE Transactions on, vol. 29, pp. 141-150, 1999.
  [Bibtex]

  @article{
  Yager+Filev99,
     Author = {Yager, R.R. and Filev, D.P.},
     Title = {Induced ordered weighted averaging operators},
     Journal = {Systems, Man and Cybernetics, Part B, IEEE Transactions on},
     Volume = {29},
     Pages = {141-150},
     Abstract = {We briefly describe the Ordered Weighted Averaging (OWA) operator and discuss a methodology for learning the associated weighting vector from observational data. We then introduce a more general type of OWA operator called the Induced Ordered Weighted Averaging (IOWA) Operator. These operators take as their argument pairs, called OWA pairs, in which one component is used to induce an ordering over the second components which are then aggregated. A number of different aggregation situations have been shown to be representable in this framework. We then show how this tool can be used to represent different types of aggregation models},
     Keywords = {artificial intelligence, cybernetics, fuzzy logic, inference mechanisms},
     Year = {1999} }

1998

• M. S. Branicky, V. S. Borkar, and S. K. Mitter, “A unified framework for hybrid control: model and optimal control theory,” Automatic Control, IEEE Transactions on, vol. 43, iss. 1, pp. 31-45, 1998.
  [Bibtex]

  @article{
  Bra+Bor98,
     Author = {Branicky, M.S. and Borkar, V.S. and Mitter, S.K.},
     Title = {A unified framework for hybrid control: model and optimal control theory},
     Journal = {Automatic Control, IEEE Transactions on},
     Volume = {43},
     Number = {1},
     Pages = {31-45},
     Abstract = {We propose a very general framework that systematizes the notion of a hybrid system, combining differential equations and automata, governed by a hybrid controller that issues continuous-variable commands and makes logical decisions. We first identify the phenomena that arise in real-world hybrid systems. Then, we introduce a mathematical model of hybrid systems as interacting collections of dynamical systems, evolving on continuous-variable state spaces and subject to continuous controls and discrete transitions. The model captures the identified phenomena, subsumes previous models, yet retains enough structure to pose and solve meaningful control problems. We develop a theory for synthesizing hybrid controllers for hybrid plants in all optimal control framework. In particular, we demonstrate the existence of optimal (relaxed) and near-optimal (precise) controls and derive "generalized quasi-variational inequalities" that the associated value function satisfies. We summarize algorithms for solving these inequalities based on a generalized Bellman equation, impulse control, and linear programming.},
     Keywords = {hierarchical systems, optimal control, automata theory, dynamic programming, state-space methods, differential equations, continuous time systems, optimal control, differential equations, automata theory, mathematical model, dynamical systems, state spa},
     Year = {1998} }

• D. Filev and R. R. Yager, “On the issue of obtaining OWA operator weights,” Fuzzy Sets and Systems, vol. 94, iss. 2, pp. 157-169, 1998.
  [Bibtex]

  @article{
  Filev+Yager98,
     Author = {Filev, Dimitar and Yager, Ronald R.},
     Title = {On the issue of obtaining OWA operator weights},
     Journal = {Fuzzy Sets and Systems},
     Volume = {94},
     Number = {2},
     Pages = {157-169},
     Abstract = {We first investigate the issue of obtaining the weights associated with the OWA aggregation in the situation when we have observed data on the arguments and the aggregated value. We next introduce a family of OWA operators called exponential OWA operators. Finally, we look at a simple procedure for generating the weights given a required degree of orness.},
        Year = {1998} }

• L. Gopalakrishnan, “Continuous improvement of the solder paste deposition process through designed experiments,” IPC/SMTA Electronics Assembly Expo. Proceedings of the Technical Program, 1998.
  [Bibtex]

  @article{
  Gopala98,
     Author = {Gopalakrishnan, L.},
     Title = {Continuous improvement of the solder paste deposition process through designed experiments},
     Journal = {IPC/SMTA Electronics Assembly Expo. Proceedings of the Technical Program},
     Abstract = {Several researchers have concluded that the solder paste deposition process has the most significant impact on SMT process yields and product reliability. The objective of this research was to evaluate the parameters affecting solder paste stencil printing and to develop a fundamental understanding of the complex interactions among them, especially in a contract assembly environment. Multiple experiments were carried out using a design of experiments approach. The components considered varied from 12 mil pitch peripheral leaded devices to large input/output BGAs. A Taguchi-based experimental methodology was used to screen the parameters that significantly influenced the stencil printing process. Factors considered in the screening experiments included solder paste type, squeegee material, stencil thickness, aperture size, print pressure, print speed, and separation speed. Full factorial experiments were then conducted to study the combined effect of the significant factors on the response variables. The volume of the solder paste deposits and the overall print quality were the response variables used. Comprehensive statistical analyses were performed, and optimal material and process parameters were identified. The response surface methodology was used to identify process windows for solder paste stencil printing. This process was further improved through the use of higher print speeds, ranging up to 7 in/s. Systematic experimentation was then carried out to test the limits of this technology for different solder paste formulations using both polyurethane and metal squeegees},
     Keywords = {assembling. ball grid arrays. circuit reliability. design, of experiments. . integrated circuit packaging. printed, circuit design. printed circuit manufacture. . printed, circuit testing. soldering. surface mount technology., Taguchi methods, Volume: S},
     Year = {1998} }

• L. Gopalakrishnan and K. Srihari, “Solder paste deposition through high speed stencil printing for a contract assembly environment,” Journal of Electronics Manufacturing, vol. 8, iss. 2, pp. 89-101, 1998.
  [Bibtex]

  @article{
  Gopala-Srihari98,
     Author = {Gopalakrishnan, L. and Srihari, K.},
     Title = {Solder paste deposition through high speed stencil printing for a contract assembly environment},
     Journal = {Journal of Electronics Manufacturing},
     Volume = {8},
     Number = {2},
     Pages = {89-101},
     Abstract = {Maximizing the throughput of an assembly line while concurrently ensuring good quality and high reliability is the primary focus of a contract printed circuit board assembly environment. Reduction of the cycle time associated with the assembly process, therefore, is of prime importance. The surface mount assembly steps include stencil printing, component placement and reflow soldering. Stencil printing is typically not the bottleneck in the assembly sequence. However, there are specific assemblies in which the stencil printing operation is the bottleneck. This research focused on reducing the cycle time associated with the stencil printing step by increasing the print speed. Designed experiments were carried out on commercially available solder pastes to investigate their ability to be stencil printed at high speeds (76 to 178 mm per second). A wide range of component types, including ball grid arrays and 0.3-mm (12 mil) pitch ultrafine pitch components, were studied. The heights of the solder paste deposits and the print definition obtained were the response variables considered. This study also determined the impact of various squeegee materials and powder sizes on solder paste deposition. It was observed that some newer commercially available solder paste formulations could be consistently and repeatably deposited at print speeds up to 178 mm (7 inches) per second},
     Keywords = {assembling. printed circuit manufacture. reflow soldering., surface mount, precise repeatable pads},
     Year = {1998} }

• D. He, N. N. Ekere, and M. A. Currie, “The behavior of solder pastes in stencil printing with vibrating squeegee,” Components, Packaging, and Manufacturing Technology, Part C, IEEE Transactions on [see also Components, Hybrids, and Manufacturing Technology, IEEE Transactions on], vol. 21, iss. 4, pp. 317-324, 1998.
  [Bibtex]

  @article{
  He98,
     Author = {He, Da and Ekere, N. N. and Currie, M. A.},
     Title = {The behavior of solder pastes in stencil printing with vibrating squeegee},
     Journal = {Components, Packaging, and Manufacturing Technology, Part C, IEEE Transactions on [see also Components, Hybrids, and Manufacturing Technology, IEEE Transactions on]},
     Volume = {21},
     Number = {4},
     Pages = {317-324},
     Abstract = {Stencil printing of solder paste with sinusoidally vibrated squeegee is a new technique developed in recent years used for the assembly of printed circuit boards (PCB's) in surface mount technology (SMT). Understanding of the behavior of solder paste under the action of vibrating squeegee is needed to optimize the process parameters. Two vibration experiments on solder paste were conducted. In the first experiment, a prototype of vibrating squeegee system was used to simulate the printing process and in the second experiment paste samples were packed in a cylindrical container which was horizontally vibrated. Experimental results validate the prior theoretical predictions. Suitable ranges of vibration parameters were found.The main points of the theoretical analysis of the process of stencil printing of solder paste using a vibrating squeegee were briefly introduced in this paper. To validate the theoretical prediction two experiments were conducted. The first experiment simulated the stencil printing of solder paste using a vibrating squeegee. Experimental results demonstrated that the application of a vibrating squeegee can generate a liquid rich layer at the squeegee blade and paste roll interface. This liquid rich layer acts as a lubricational agent between the blade and the paste roll that can reduce the squeegee blade resistance on the paste roll. Empirical evidence has shown that a good paste roll is essential for aperture filling and emptying, and thus more consistent deposits on the pads of a substrate. The application of the vibrating squeegee can also help to reduce the potential for the paste roll to stick on the squeegee blade at the end of a printing stroke. For paste sample I used in this experiment, the suitable range of the vibration frequency is from 80–200 Hz, the suitable range of the amplitude is from 0.1–0.37 mm. In the suitable ranges, the effect of application of high frequency and small amplitude is more significant than otherwise.In the second experiment solder paste sample was vibrated inside a cylindrical container that, in some extent, could reflect the response of paste packed inside apertures to the vibration of the squeegee. Experimental results showed that under vibration a liquid rich layer is generated around the container wall, which implies that vibration may help the transfer of paste from apertures to substrate, thus reducing clogging defects. Microscope observation showed that under vibration the arrangement of solder particles is more uniform than that without vibration. Uniform arrangement of solder particles may reduce the bulk viscosity because of the reduction of lubricational forces among neighboring particles. The use of the expired paste, sample II, in this study may suggest that solder pastes with high metal loads can be used in stencil printing with a vibrating squeegee. Pastes with high metal load can maintain deposit resolutions and reduce bridging defects.},
     Keywords = {reflow soldering, printing, printed circuit manufacture, assembling, surface mount technology, solder pastes, stencil printing, vibrating squeegee, printed circuit boards, assembly, surface mount technology, process parameters, printing process, cylindr},
     Year = {1998} }

• S. Kalluri and G. R. Arce, “Adaptive weighted myriad filter algorithms for robust signal processing in α-stable noise environments,” Signal Processing, IEEE Transactions on, vol. 46, pp. 322-334, 1998.
  [Bibtex]

  @article{
  Kalluri+Arce98,
     Author = {Kalluri, S. and Arce, G.R.},
     Title = {Adaptive weighted myriad filter algorithms for robust signal processing in α-stable noise environments},
     Journal = {Signal Processing, IEEE Transactions on},
     Volume = {46},
     Pages = {322-334},
     Abstract = {Stochastic gradient-based adaptive algorithms are developed for the optimization of weighted myriad filters (WMyFs). WMyFs form a class of nonlinear filters, motivated by the properties of α-stable distributions, that have been proposed for robust non-Gaussian signal processing in impulsive noise environments. The weighted myriad for an N-long data window is described by a set of nonnegative weights {wi }i=lN and the so-called linearity parameter K>0. In the limit, as K→∞, the filter reduces to the familiar weighted mean filter (which is a constrained linear FIR filter). Necessary conditions are obtained for optimality of the filter weights under the mean absolute error criterion. An implicit formulation of the filter output is used to find an expression for the gradient of the cost function. Using instantaneous gradient estimates, an adaptive steepest-descent algorithm is then derived to optimize the weights. This algorithm involves a very simple update term that is computationally comparable to the update in the classical LMS algorithm. The robust performance of this adaptive algorithm is demonstrated through a computer simulation example involving lowpass filtering of a one-dimensional chirp-type signal in impulsive noise},
     Keywords = {FIR filters, adaptive filters, adaptive signal processing, circuit optimisation, filtering theory, noise, nonlinear filters, optimisation, statistical analysis},
     Year = {1998} }

• C. S. Kenney, A. J. Laub, and M. S. Reese, “Statistical Condition Estimation for Linear Least Squares,” SIAM Journal on Matrix Analysis and Applications, vol. 19, iss. 4, pp. 906-923, 1998.
  [Bibtex]

  @article{
  Kenney98,
     Author = {Kenney, C. S. and Laub, A. J. and Reese, M. S.},
     Title = {Statistical Condition Estimation for Linear Least Squares},
     Journal = {SIAM Journal on Matrix Analysis and Applications},
     Volume = {19},
     Number = {4},
     Pages = {906-923},
     Abstract = { Statistical condition estimation is applied to the linear least squares problem. The method obtains componentwise condition estimates via the Fréchet derivative. A rigorous statistical theory exists that determines the probability of accuracy in the estimates. The method is as computationally efficient as normwise condition estimation methods, and it is easily adapted to respect structural constraints on perturbations of the input data. Several examples illustrate the method.},
     Keywords = {conditioning, sensitivity, linear least squares},
     Year = {1998} }

• J. C. Lagarias, J. A. Reeds, M. H. Wright, and P. E. Wright, “Convergence Properties of the Nelder-Mead Simplex Method in Low Dimensions,” SIAM Journal on Optimization, vol. 9, iss. 1, pp. 112-147, 1998.
  [Bibtex]

  @article{
  Lagarias98,
     Author = {Lagarias, Jeffrey C. and Reeds, James A. and Wright, Margaret H. and Wright, Paul E.},
     Title = {Convergence Properties of the Nelder-Mead Simplex Method in Low Dimensions},
     Journal = {SIAM Journal on Optimization},
     Volume = {9},
     Number = {1},
     Pages = {112-147},
     Abstract = {The Nelder--Mead simplex algorithm, first published in 1965, is an enormously popular direct search method for multidimensional unconstrained minimization. Despite its widespread use, essentially no theoretical results have been proved explicitly for the Nelder--Mead algorithm. This paper presents convergence properties of the Nelder--Mead algorithm applied to strictly convex functions in dimensions 1 and 2. We prove convergence to a minimizer for dimension 1, and various limited convergence results for dimension 2. A counterexample of McKinnon gives a family of strictly convex functions in two dimensions and a set of initial conditions for which the Nelder--Mead algorithm converges to a nonminimizer. It is not yet known whether the Nelder--Mead method can be proved to converge to a minimizer for a more specialized class of convex functions in two dimensions.},
     Keywords = {direct search methods, Nelder--Mead simplex methods, nonderivative optimization},
     Year = {1998} }

• A. Lotfi and M. Howarth, “Industrial application of fuzzy systems: Adaptive fuzzy control of solder paste stencil printing,” Information Sciences, vol. 107, iss. 1-4, pp. 273-285, 1998.
  [Bibtex]

  @article{
  Lotfi+Howarth98a,
     Author = {Lotfi, A. and Howarth, M.},
     Title = {Industrial application of fuzzy systems: Adaptive fuzzy control of solder paste stencil printing},
     Journal = {Information Sciences},
     Volume = {107},
     Number = {1-4},
     Pages = {273-285},
     Abstract = {This paper presents an adaptive fuzzy control algorithm for the control of the solder paste stencil printing stage of surface mount printed circuit board (PCB) assembly. The proposed method of automatic solder paste stencil printing consists of four blocks: fuzzy feature extraction, defect classification of paste deposits, adaptive fuzzy rule-based model identification and subsequently an optimal control action for the stencil printing process. Experimental results are presented to illustrate the capability of the algorithm.An algorithm for automatic control of solder paste stencil printing process of SMT is developed to reduce the number of board failures using an adaptive fuzzy controller. The prototype of the proposed automatic control is implemented and it demonstrates the capability of a fuzzy system to model and produce a good control accion for solder paste stencil printing.},
     Keywords = {Surface mount technology, Printed circuit boards, Fuzzy sets,, Fuzzy control, Adaptive algorithms, Feature extraction, Knowledge, based systems, Adaptive control systems, Optimal control systems,, Identification (control systems), Solder paste stencil p},
     Year = {1998} }

• B. Ries, “3-D Post Printing Inspection,” Circuits Assembly, iss. June, pp. 40-47, 1998.
  [Bibtex]

  @article{
  Ries98,
     Author = {Ries, Bob},
     Title = {3-D Post Printing Inspection},
     Journal = {Circuits Assembly},
     Number = {June},
     Pages = {40-47},
     Month = {June},
     Keywords = {3D 2D, 10x, Volume: Single best predictor, sampled and 100%, in-line, off-line inspection , Symptoms and possible causes and actions},
     Year = {1998} }

1997

• L. El Ghaoui and H. Lebret, “Robust Solutions to Least-Squares Problems with Uncertain Data,” SIAM Journal on Matrix Analysis and Applications, vol. 18, iss. 4, pp. 1035-1064, 1997.
  [Bibtex]

  @article{
  Ghaoui+Lebret97,
     Author = {El Ghaoui, Laurent  and Lebret, Hervé},
     Title = {Robust Solutions to Least-Squares Problems with Uncertain Data},
     Journal = {SIAM Journal on Matrix Analysis and Applications},
     Volume = {18},
     Number = {4},
     Pages = {1035-1064},
     Abstract = {We consider least-squares problems where the coefficient matrices A,b are unknown but bounded. We minimize the worst-case residual error using (convex) second-order cone programming, yielding an algorithm with complexity similar to one singular value decomposition of A. The method can be interpreted as a Tikhonov regularization procedure, with the advantage that it provides an exact bound on the robustness of solution and a rigorous way to compute the regularization parameter. When the perturbation has a known (e.g., Toeplitz) structure, the same problem can be solved in polynomial-time using semidefinite programming (SDP). We also consider the case when A,b are rational functions of an unknown-but-bounded perturbation vector. We show how to minimize (via SDP) upper bounds on the optimal worst-case residual. We provide numerical examples, including one from robust identification and one from robust interpolation.},
     Keywords = {least-squares problems, uncertainty, robustness, second-order cone programming, semidefinite programming, ill-conditioned problem, regularization, robust identification, robust interpolation},
     Year = {1997} }

• R. Fierro and F. L. Lewis, “A framework for hybrid control design,” Systems, Man and Cybernetics, Part A, IEEE Transactions on, vol. 27, iss. 6, pp. 765-773, 1997.
  [Bibtex]

  @article{
  Fierro97,
     Author = {Fierro, R. and Lewis, F.L.},
     Title = {A framework for hybrid control design},
     Journal = {Systems, Man and Cybernetics, Part A, IEEE Transactions on},
     Volume = {27},
     Number = {6},
     Pages = {765-773},
     Abstract = {This paper presents a hybrid system framework which considers simultaneously the control and decision-making issues. This reconfigurable framework can accommodate a wide range of situations, from aircraft control systems to mobile manipulators. A continuous-state plant is supervised by a discrete-event system which is based on a theory of linked finite state machines. The composite system is viewed as an iterative process where a task is carried out by changing the structure of the continuous-state plant. An algorithm for a hybrid control design is provided and illustrated through a mobile manipulator example.},
     Keywords = {finite state machines, continuous time systems, discrete event systems, intelligent control, control system synthesis, mobile robots, manipulators, position control, fuzzy logic, closed loop systems, hybrid control design, decision-making, reconfigurabl},
     Year = {1997} }

• A. Johnson, “Optimizing stencil printing for fine-pitch circuit boards,” Electronics Engineer, iss. December, 1997.
  [Bibtex]

  @article{
  Johnson97,
     Author = {Johnson, Alden},
     Title = {Optimizing stencil printing for fine-pitch circuit boards},
     Journal = {Electronics Engineer},
     Number = {December},
     Month = {December},
     Abstract = {View printing of fine-pitch PCBs as a complete process - a combination of substrate, stencil, paste, and paste-deposition equipment. Following general guidelines will help you devise a system suited to your production environment.},
     Keywords = {General Recomendations},
     Year = {1997} }

• F. K. H. Lau and V. W. S. Yeung, “A hierarchical evaluation of the solder paste printing process,” Journal of Materials Processing Technology, vol. 69, iss. 1-3, pp. 79-89, 1997.
  [Bibtex]

  @article{
  Lau97,
     Author = {Lau, Francis K. H. and Yeung, Vincent W. S.},
     Title = {A hierarchical evaluation of the solder paste printing process},
     Journal = {Journal of Materials Processing Technology},
     Volume = {69},
     Number = {1-3},
     Pages = {79-89},
     Abstract = {Solder paste printing is one of the most critical processes in applying Surface Mount Technology. Investigation has shown that PCBA defectives derived out from reflow soldering inspection, in-circuit testing, and function testing up to final quality inspection, are heavily induced by poor solder paste printing. A hierarchy has been developed according to the stages and importance of the set-up of the printing process. The hierarchy consists of six levels: (i) geometry of the stencil openings resulting from fabrication method employed; (ii) solder paste matching; (iii) waiting time effects; (iv) squeegee materials selection; (v) printing machine operation parameter design; and (vi) reflow soldering process parameter design. The mechanisms, responses and factors within each level have been evaluated and optimized. Relationships amongst the levels were formulated to generate the complete set-up. Taguchi's Method was applied to the evaluation of level 5 and level 6, using L27 and L18 orthogonal arrays, respectively. ANOVAs were also run to test the significance of the results. The significant control factors were identified and optimized accordingly. The settings and rules of the hierarchy were established. Confirmation trials and follow-up showed that there was up to 25% reduction in the defect rate and 20% saving of the labour cost. The findings of the project illustrate a complete mechanism for material processing evaluation and improvement.},
     Keywords = {Solder paste printing, Reflow soldering, SMT processing, Hierarchical evaluation, Taguchi's method, Defect rate, Quality control, defects related SPP, General Recomendations, big components, metal squeegee, angle, speed, pressure, cleaning interval, sap},
     Year = {1997} }

1996

• S. J. M. A. E. Twomey, “Artificial neural network approach to the control of a wave soldering process,” Intelligent Engineering Systems Through Artificial Neural Networks, vol. 6, pp. 889-894, 1996.
  [Bibtex]

  @article{
  Twomey96,
     Author = {Twomey, J. M. Smith A. E.},
     Title = {Artificial neural network approach to the control of a wave soldering process},
     Journal = {Intelligent Engineering Systems Through Artificial Neural Networks},
     Volume = {6},
     Pages = {889-894},
     url = {ASME Fairfield NJ USA},
     Abstract = {This research proposes a novel artificial neural network based approach for the control of wave soldering processes of electronic circuit boards. The network approach is committee networks by cross-validation. This method maximizes the use of the entire data set for model building and evaluation. The approach is compared to the two most common methods of building and evaluating neural networks. (Author abstract) 9 Refs.},
     Keywords = {Soldering, Neural networks, Quality control, Optimization, Parameter estimation, Mathematical models, Wave soldering},
     Year = {1996} }

• R. R. Yager, “On mean type aggregation,” Systems, Man and Cybernetics, Part B, IEEE Transactions on, vol. 26, pp. 209-221, 1996.
  [Bibtex]

  @article{
  Yager96,
     Author = {Yager, R.R.},
     Title = {On mean type aggregation},
     Journal = {Systems, Man and Cybernetics, Part B, IEEE Transactions on},
     Volume = {26},
     Pages = {209-221},
     Abstract = {We introduce and define the concept of mean aggregation of a collection of n numbers. We point out that the lack of associativity of this operation compounds the problem of the extending mean of n numbers to n+1 numbers. The closely related concepts of self identity and the centering property are introduced as one imperative for extending mean aggregation operators. The problem of weighted mean aggregation is studied. A new concept of prioritized mean aggregation is then introduced. We next show that the technique of selecting an element based upon the performance of a random experiment can be considered as a mean aggregation operation},
     Keywords = {fuzzy logic},
     Year = {1996} }

1995

• C. W. Chung and K. Srihari, “A “Closed Loop” System For The Realtime Control Of Solder Paste Stencil Printing,” Journal Of Electronics Manufacturing, vol. 5, iss. 2, pp. 99-109, 1995.
  [Bibtex]

  @article{
  Chung95,
     Author = {Chung, C.W. and Srihari, K.},
     Title = {A "Closed Loop" System For The Realtime Control Of Solder Paste Stencil Printing},
     Journal = {Journal Of Electronics Manufacturing},
     Volume = {5},
     Number = {2},
     Pages = {99-109},
        Year = {1995} }

• L. Nyberg, “`First article’ inspection and what to look for,” Surface Mount Technology, vol. 9, iss. 11, pp. 48-49, 1995.
  [Bibtex]

  @article{
  Nyberg95,
     Author = {Nyberg, Lindley},
     Title = {`First article' inspection and what to look for},
     Journal = {Surface Mount Technology},
     Volume = {9},
     Number = {11},
     Pages = {48-49},
     Abstract = {The inspection of the first board in surface mount technology assemblies should be detailed in order to minimize the chance of costly error before a full-blown production run. Three major steps for first article inspection are described. These steps are: first article inspection for stencil printer; first article inspection for pick and place; and first article inspection for reflow soldering. This description is for a simple assembly of 0.025 inches-pitch parts of larger on FR-4 substrates using Sn63/Pb37 paste.},
     Keywords = {Printed circuit boards, Surface mount technology, Inspection,, Screen printing, Soldering, Quality control, Soldering alloys, First article inspection, Stencil printer, Reflow soldering, E 714.2 Semiconductor Devices and Integrated Circuits, E 913.3.1, },
     Year = {1995} }

• E. Sachs, A. Hu, and A. Ingolfsson, “Run by run process control: combining SPC and feedback control,” Semiconductor Manufacturing, IEEE Transactions on, vol. 8, iss. 1, pp. 26-43, 1995.
  [Bibtex]

  @article{
  Sachs95,
     Author = {Sachs, E. and Hu, A. and Ingolfsson, A.},
     Title = {Run by run process control: combining SPC and feedback control},
     Journal = {Semiconductor Manufacturing, IEEE Transactions on},
     Volume = {8},
     Number = {1},
     Pages = {26-43},
     Abstract = {The run by run controller provides a framework for controlling a process which is subject to disturbances such as shifts and drifts as a normal part of its operation. The run by run controller combines the advantages of both statistical process control (SPC) and feedback control. It has three components: rapid mode, gradual mode, and generalized SPC. Rapid mode adapts to sudden shifts in the process such as those caused by maintenance operations.  Gradual mode adapts to gradual drifts in the process such as those caused by build-up of deposition inside a reactor. The choice between the two modes is determined by the outcome from generalized SPC which allows SPC to be applied to a process while it is being tuned. The run by run controller has been applied to the control of a silicon epitaxy process in a barrel reactor. Rapid mode recovered the process within 3 runs after a disturbance.  Gradual mode reduced the variation of the process by a factor of 2.7 as compared to historical data.},
     Keywords = {closed loop systems, statistical process control, integrated circuit manufacture, vapour phase epitaxial growth, semiconductor growth, silicon, run by run process control, SPC, feedback control, statistical process control, rapid mode, gradual mode, mai},
     Year = {1995} }

• S. Venkateswaran, K. Srihari, J. H. Adriance, and G. R. Westby, “A software system for the process control and trouble shooting of solder paste stencil printing,” Surface Mount International, pp. 713-179, 1995.
  [Bibtex]

  @article{
  Venka95,
     Author = {Venkateswaran, S. and Srihari, K. and Adriance, J.H. and Westby, G.R.},
     Title = {A software system for the process control and trouble shooting of solder paste stencil printing},
     Journal = {Surface Mount International},
     Pages = {713-179},
        Year = {1995} }

1994

• C. W. Chung, K. Srihari, J. H. Adriance, and G. R. Westby, “Closed loop process control for solder paste stencil printing,” Surface Mount International Conference and Exposition. Proceedings of the Technical Program, 1994.
  [Bibtex]

  @article{
  Chung94,
     Author = {Chung, C. W. and Srihari, K. and Adriance, J. H. and Westby, G. R.},
     Title = {Closed loop process control for solder paste stencil printing},
     Journal = {Surface Mount International Conference and Exposition. Proceedings of the Technical Program},
     Abstract = {The deposition of solder paste in the proper amount and in the right place strongly influences process yields in the surface mount assembly of printed circuit boards (PCBs). The move towards fine and ultra-fine pitch surface mount components (SMCs) has increased the importance of the solder paste printing process. A prototype real time closed-loop process control system was designed and developed in this research to monitor the solder paste stencil printing process for the assembly of surface mount PCBs. This decision support process control system achieves `optimal' print conditions through an iterative approach that was both proactive and reactive in nature},
     Keywords = {assembling. closed loop systems. fine-pitch technology., iterative methods. printed circuit manufacture. printed, circuit testing. process control. soldering. surface mount, technology},
     Year = {1994} }

• K. Feldmann and J. Sturm, “Closed loop quality control in printed circuit assembly,” Components, Packaging, and Manufacturing Technology, Part A, IEEE Transactions on [see also Components, Hybrids, and Manufacturing Technology, IEEE Transactions on], vol. 17, iss. 2, pp. 270-276, 1994.
  [Bibtex]

  @article{
  Feldmann94,
     Author = {Feldmann, Klaus and Sturm, Jürgen},
     Title = {Closed loop quality control in printed circuit assembly},
     Journal = {Components, Packaging, and Manufacturing Technology, Part A, IEEE Transactions on [see also Components, Hybrids, and Manufacturing Technology, IEEE Transactions on]},
     Volume = {17},
     Number = {2},
     Pages = {270-276},
     Abstract = {Almost 30-50% of fabrication costs in electronics production are caused by testing and repair operations. The reasons for that problem are many. First of all, there is still a lack of knowledge concerning the conditions for robust process control. On the other hand, uncertain specification of product quality causes unnecessary inspection, testing, and repair actions. Thus, a strategy for yield improvement requires strong dedication to causal relationships and specification of process and product quality. In the first step, critical process parameters and their causal effects on product quality must be analyzed. In the second step, once critical process parameter are addressed, control strategies can be developed to assure high first pass yields. This paper gives an overview of ongoing work that has been performed to establish a SMT production line with closed loop quality control via direct process monitoring capabilities and integrated inspection operations. For different inspection tasks, 3-D laser and X-ray inspection is applied. Process parameters as well as test and inspection results are gathered on line. Thus, interrelationships between different process steps, beginning from material inspection, solder paste application, components placement, reflow soldering, and final inspection, can be analyzed. Moreover, variation of process parameters conforming to different approaches to design of experiments (DOE) can be performed while running production. On-line diagnosis of process parameters and inspection results enables a closed loop quality control.},
     Keywords = {quality control, computerised monitoring, surface mount technology, assembling, production testing, automatic optical inspection, production engineering computing, printed circuit testing, printed circuit manufacture, closed loop quality control, printe},
     Year = {1994} }

• M. T. Lamata, “A model of decision with linguistic knowledge,” Mathware and Soft Computing, vol. 3, pp. 253-263, 1994.
  [Bibtex]

  @article{
  Lamata94,
     Author = {Lamata, M.T.},
     Title = {A model of decision with linguistic knowledge},
     Journal = {Mathware and Soft Computing},
     Volume = {3},
     Pages = {253-263},
     Abstract = {The aim of this paper is to develop a new aggregating method for the decision problem in which the possible values of rewards are known in linguistic terms. We show new operators for solving this problem, as well as the way in which OWA operators provide us with an adequate framework for representing the optimism degree of the decision maker in case we have no information about the real state.},
     Keywords = {decision-making, linguistic label, uncertainty, operators},
     Year = {1994} }

• Y. Li, R. L. Mahajan, and J. Tong, “Design factors and their effect on PCB assembly yield-statistical and neural network predictive models,” Components, Packaging, and Manufacturing Technology, Part A, IEEE Transactions on [see also Components, Hybrids, and Manufacturing Technology, IEEE Transactions on], vol. 17, iss. 2, pp. 183-191, 1994.
  [Bibtex]

  @article{
  Li+Mahajan94,
     Author = {Li, Y. and Mahajan, R.L. and Tong, J.},
     Title = {Design factors and their effect on PCB assembly yield-statistical and neural network predictive models},
     Journal = {Components, Packaging, and Manufacturing Technology, Part A, IEEE Transactions on [see also Components, Hybrids, and Manufacturing Technology, IEEE Transactions on]},
     Volume = {17},
     Number = {2},
     Pages = {183-191},
     Abstract = {This study relates circuit board design features to assembly yields. Data used were collected over a period of one year from two circuit board assembly shops at AT&T. Design parameters that may affect the assembly yield were identified using knowledge of the assembly process. These parameters were then quantified for a set of board designs and related to the actual assembly yield by the statistical regression models and the artificial neural network (ANN) models. These models are able to predict the assembly yield with a root mean square (RMS) error of less than 5%. They can be used to predict the assembly yield for new board designs on the same line. Alternatively, they can be used to compare the performance of different lines by comparing the expected yield for a given design with the actual yield.},
     Keywords = {printed circuit manufacture, printed circuit design, neural nets, modelling, statistical analysis, design factors, PCB assembly yield, statistical models, neural network predictive models, circuit board design features, assembly yields, circuit board as},
     Year = {1994} }

• S. H. Mannan, N. N. Ekere, I. Ismail, and E. K. Lo, “Squeegee deformation study in the stencil printing of solder pastes,” Components, Packaging, and Manufacturing Technology, Part A, IEEE Transactions on [see also Components, Hybrids, and Manufacturing Technology, IEEE Transactions on], vol. 17, iss. 3, pp. 470-476, 1994.
  [Bibtex]

  @article{
  Mannan94,
     Author = {Mannan, S. H. and Ekere, N. N. and Ismail, I. and Lo, E. K.},
     Title = {Squeegee deformation study in the stencil printing of solder pastes},
     Journal = {Components, Packaging, and Manufacturing Technology, Part A, IEEE Transactions on [see also Components, Hybrids, and Manufacturing Technology, IEEE Transactions on]},
     Volume = {17},
     Number = {3},
     Pages = {470-476},
     Abstract = {We report on the results of an experimental comparison of different types of squeegee blade used in the stencil printing of solder pastes for reflow soldering in SMT, concentrating on paste heights (scooping) and printing defects. We show how our experimental results for squeegee deformation into stencil apertures lead to the construction of a model for squeegee deformation. The model takes into account the force on the squeegee due to solder paste flow and some of the non-Newtonian properties of the solder paste. An explanation is proposed for the differences in paste heights between apertures of different orientations.},
     Keywords = {printing, soldering, surface mount technology, printed circuit manufacture, squeegee deformation study, stencil printing, solder pastes, reflow soldering, SMT, paste heights, printing defects, nonNewtonian properties, aperture orientation, equations, _|},
     Year = {1994} }

1993

• M. S. Husman, J. P. Rukavina, and G. Yuan, “A study of solder paste volumes for screen printing,” Proceedings of the Technical Program. NEPCON West’93, 1993.
  [Bibtex]

  @article{
  Husman93,
     Author = {Husman, M. S. and Rukavina, J. P. and Yuan, Guo},
     Title = {A study of solder paste volumes for screen printing},
     Journal = {Proceedings of the Technical Program. NEPCON West'93},
     Abstract = {Solder paste deposition is a very important factor in the surface mount assembly process. It is well known that an accurate and consistent volume of solder paste is the key to low-defect reflow soldering. This paper describes an investigation of the effect of aperture direction on solder paste volumes and their variations. The study considered two squeegee materials, a 90 durometer rubber squeegee and a metal squeegee, as well as two different component lead pitches, 25 mil and 50 mil. The purpose of this study is threefold. (i) To compare solder paste volumes deposited in aperture directions parallel to and perpendicular to the print direction. If a significant volume difference is noted, stencil design guidelines should reflect this. From the authors' experience, the size and shape of apertures for a four-sized symmetrical component are the same in both the horizontal and vertical directions. (ii) To examine the consistency of solder paste depositions in both aperture directions. If strong statistical evidence suggests that solder paste volumes are more consistent on one aperture direction than the other, design engineers may find this information useful for printed circuit board layout. (iii) To compare the results of (i) and (ii) for two different squeegee materials. This will be useful for squeegee material selection},
     Keywords = {assembling. circuit layout. printed circuit design., printed circuit manufacture. soldering. surface mount, technology, Volume: Single best predictor},
     Year = {1993} }

• C. Wang and G. Zhongguo Jixie, “Real time quality monitor and control of the SMT laser microsolder joints,” Zhongguo Jixie Gongcheng/China Mechanical Engineering, vol. 4, iss. 6, pp. 9-11, 1993.
  [Bibtex]

  @article{
  Wang93,
     Author = {Wang, Chunqing and Zhongguo Jixie, Gongcheng},
     Title = {Real time quality monitor and control of the SMT laser microsolder joints},
     Journal = {Zhongguo Jixie Gongcheng/China Mechanical Engineering},
     Volume = {4},
     Number = {6},
     Pages = {9-11},
     Abstract = {In this paper, a new type of system has been developed for monitoring and controlling the quality of SMT laser microsolder joints in the soldering process, and a separated laser heating and infrared detecting unit has been designed in which the InSb thermoelectric pickup was adopted as the infrared detector cell. The experiment results show that the infrared radiation signal from the soldering joints includes the useful information of the heating and melting process of the solder paste. A microcomputer quality monitor and closed-loop control method has been achieved tentatively. (Translated author abstract) 3 Refs.},
     Keywords = {Soldered joints,  Surface mount technology,  Soldering,  Monitoring,  Quality control,  Process control,  Infrared detectors,  Control systems,  Real time systems,  Soldering alloys,  Heating,  Melting,  Laser applications,  Laser microsoldering,  Close},
     Year = {1993} }

• G. R. Yost, “Acquiring knowledge in Soar,” Expert, IEEE [see also IEEE Intelligent Systems], vol. 8, iss. 3, pp. 26-34, 1993.
  [Bibtex]

  @article{
  Yost93,
     Author = {Yost, G.R.},
     Title = {Acquiring knowledge in Soar},
     Journal = {Expert, IEEE [see also IEEE Intelligent Systems]},
     Volume = {8},
     Number = {3},
     Pages = {26-34},
     Abstract = {The strategies used in designing limited method-based tools can be applied to the much broader method that underlies Soar, called the problem space computational model (PSCM). Taql (for task acquisition language), which is the computer representation of PSCM, is discussed. Taql, together with its programming environment, comprises the Soar-based expert-system development tool. A Taql specification consists of a set of Taql constructs, each of which describes some aspect of a PSCM knowledge role. Each of Taql's 17 constructs is a list consisting of a type and a name for the Taql construct instance, followed by a list of arguments. Each argument specifies some aspect of the construct's PSCM knowledge role. Taql's effectiveness is attributed to a combination of PSCM's simplicity and flexibility and the principled way in which Taql was designed.},
     Keywords = {specification language, knowledge acquisition, Soar, method-based tools, problem space computational model, Taql, task acquisition language, PSCM, programming environment, expert-system development tool, specification, argument, expert systems, formal s},
     Year = {1993} }

1991

• A. B. Suppelsa and H. F. Liebman, “Successful implementation of closed loop semi-aqueous cleaning,” Proceedings of the Technical Program National Electronic Packaging and Production Conference, vol. 1, pp. 557-565, 1991.
  [Bibtex]

  @article{
  Suppelsa91,
     Author = {Suppelsa, A. B.  and Liebman, H. F.},
     Title = {Successful implementation of closed loop semi-aqueous cleaning},
     Journal = {Proceedings of the Technical Program National Electronic Packaging and Production Conference},
     Volume = {1},
     Pages = {557-565},
     Abstract = {Following extensive materials and process testing, and equipment development, a Closed Loop spray cleaning cell utilizing the Semi-Aqueous, Terpene hydro-carbon BioAct EC7R, has been installed at the Motorola Inc., Communications Manufacturing Facility, at Plantation, Florida. This cleaning cell, which represents the present best viable alternative to CFC cleaning, is now being used to clean and to remove post reflow solder paste residues from soldered surface mount PCBs used in portable communications equipment. This paper will report the system development, production implementation and use history to date, review the terpene technology equipment developments to date, explain the environmental non-impact of system, outline the cleaning cell process philosophy and construction, and overall cleaning experience to date. (Author abstract) 58 Refs},
     Keywords = {Printed Circuits,  Integrated Circuit Manufacture - Cleaning,  Solvents,  Hydrocarbons,  Soldering - Decontamination,  Semi-Aqueous Cleaning,  CFC,  Reflow Soldering,  PCB},
     Year = {1991} }

1990

• J. A. Owczarek and F. L. Howland, “A study of the off-contact screen printing process. I. Model of the printing process and some results derived from experiments,” Components, Hybrids, and Manufacturing Technology, IEEE Transactions on [see also IEEE Trans. on Components, Packaging, and Manufacturing Technology, Part A, B, C], vol. 13, iss. 2, pp. 358-367, 1990.
  [Bibtex]

  @article{
  Owczarek90a,
     Author = {Owczarek, J.A. and Howland, F.L.},
     Title = {A study of the off-contact screen printing process. I. Model of the printing process and some results derived from experiments},
     Journal = {Components, Hybrids, and Manufacturing Technology, IEEE Transactions on [see also IEEE Trans. on Components, Packaging, and Manufacturing Technology, Part A, B, C]},
     Volume = {13},
     Number = {2},
     Pages = {358-367},
     Abstract = {A physical model of the screen printing process is described. Interrupted printing tests provided guiding information on the flow pattern in the paste during the deposition process. The paste flow region ahead of the squeegee was divided into three regions: the pressurization region (region I), the downward screen cross-flow region (region II), and the paste accumulation region (region III). During printing, the squeegee tip becomes deformed and its angle of attack decreases. Region I extends from the leading edge of the deformed squeegee tip to the beginning of the deformation region. As a result of pressure buildup in region I ahead of the squeegee, there is a flow of paste under the squeegee. The paste deposition process depends mainly on the flow process in region I. Analysis of experiments indicates that the squeegees used deform so that their fronts can be approximated by a wedge. The angles of hard squeegees decrease by approximately 20 degrees from the undeformed angle of 45 degrees . The corresponding angle decrease for soft squeegees is on the order of 30-40 degrees . The vertical forces acting on the squeegees were estimated to be between 60 and 47 lbf. The average speed of the paste under the squeegee, caused by the developed hydrodynamic pressure, was found to be between 0.1 and 0.2 of the squeegee speed.},
     Keywords = {off-contact screen printing process, printing process, physical model, guiding information, flow pattern, paste, deposition process, pressurization region, downward screen cross-flow region, paste accumulation region, squeegee tip, hard squeegees, verti},
     Year = {1990} }

• J. A. Owczarek and F. L. Howland, “A study of the off-contact screen printing process. II. Analysis of the model of the printing process,” Components, Hybrids, and Manufacturing Technology, IEEE Transactions on [see also IEEE Trans. on Components, Packaging, and Manufacturing Technology, Part A, B, C], vol. 13, iss. 2, pp. 368-375, 1990.
  [Bibtex]

  @article{
  Owczarek90b,
     Author = {Owczarek, J.A. and Howland, F.L.},
     Title = {A study of the off-contact screen printing process. II. Analysis of the model of the printing process},
     Journal = {Components, Hybrids, and Manufacturing Technology, IEEE Transactions on [see also IEEE Trans. on Components, Packaging, and Manufacturing Technology, Part A, B, C]},
     Volume = {13},
     Number = {2},
     Pages = {368-375},
     Abstract = {For pt.I see ibid., vol.13, no.2, p.358-67, 1990. The analysis presented is concerned with the derivation of equations which allows calculation of the thickness of dry paste deposited on a substrate in the off-contact screen printing process. The rheological properties of the paste are approximated by those of a power-law fluid. Deformed squeegee shape in the paste deposition region is considered to be that of a wedge. The equations show that the height of the deposited paste depends on the magnitude of the equivalent paste height under the squeegee and on the paste flow speed under the squeegee. The equivalent paste height under the squeegee depends, for a given screen, on the deformability of the squeegee and on the emulsion height. The paste flow speed under the squeegee depends, for a given paste and screen, on the magnitude of the pressure buildup ahead of the squeegee. The pressure buildup in turn is a function of the squeegee deformation characteristics and of squeegee translational speed.},
     Keywords = {off-contact screen printing process, model, thickness, dry paste, rheological properties, power-law fluid, squeegee shape, paste deposition region, equivalent paste height, paste flow speed, deformability, emulsion height, pressure buildup, squeegee tra},
     Year = {1990} }

1988

• R. R. Yager, “On ordered weighted averaging aggregation operators in multicriteria decisionmaking,” Systems, Man and Cybernetics, IEEE Transactions on, vol. 18, pp. 183-190, 1988.
  [Bibtex]

  @article{
  Yager88,
     Author = {Yager, R.R.},
     Title = {On ordered weighted averaging aggregation operators in multicriteria decisionmaking},
     Journal = {Systems, Man and Cybernetics, IEEE Transactions on},
     Volume = {18},
     Pages = {183-190},
     Abstract = {The author is primarily concerned with the problem of aggregating multicriteria to form an overall decision function. He introduces a type of operator for aggregation called an ordered weighted aggregation (OWA) operator and investigates the properties of this operator. The OWA's performance is found to be between those obtained using the AND operator, which requires all criteria to be satisfied, and the OR operator, which requires at least one criteria to be satisfied},
     Keywords = {decision theory, optimisation, set theory},
     Year = {1988} }

1983

• S. Kirkpatrick, C. D. Gelatt Jr., and M. P. Vecchi, “Optimization by Simulated Annealing,” Science, vol. 220, iss. 4598, pp. 671-680, 1983.
  [Bibtex]

  @article{
  Kirkpatrick83,
     Author = {Kirkpatrick, S.  and Gelatt, Jr., C. D.  and Vecchi, M.P.},
     Title = {Optimization by Simulated Annealing},
     Journal = {Science},
     Volume = {220},
     Number = {4598},
     Pages = {671-680},
     Month = {13 May},
     Abstract = {In this article we briefly review the central constructs in combinatorial optimization and in statistical mechanics and then develop the similarities between the two fields. We show how the Metropolis algorithm for approximate numerical simulation of the behavior of a manybody system at a finite temperature provides a natural tool for bringing the techniques of statistical mechanics to bear on optimization. We have applied this point of view to a number of problems arising in optimal design of computers.},
        Year = {1983} }

1975

• J. J. Filliben, “The Probability Plot Correlation Coefficient Test for Normality ,” Technometrics, vol. 17, iss. 1, pp. 111-117, 1975.
  [Bibtex]

  @article{
  Filliben75,
     Author = {Filliben, James J.},
     Title = {The Probability Plot Correlation Coefficient Test for Normality },
     Journal = {Technometrics},
     Volume = {17},
     Number = {1},
     Pages = {111-117},
     Abstract = {This paper introduces the normal probability plot correlation coefficient as a test statistic in complete samples for the composite hypothesis of normality. The proposed test statistic is conceptually simple, is computationally convenient, and is readily extendible to testing non-normal distributional hypotheses. An empirical power study shows that the normal probability plot correlation coefficient compares favorably with 7 other normal test statistics. Percent points are tabulated for n = 3(1)50(5)100.},
        Year = {1975} }

1967

• H. W. Lilliefors, “On the Kolmogorov-Smirnov test for normality with mean and variance unknown,” Journal of the American Statistical Association, vol. 62, pp. 399-402, 1967.
  [Bibtex]

  @article{
  Lilliefors67,
     Author = {Lilliefors, H. W.},
     Title = {On the Kolmogorov-Smirnov test for normality with mean and variance unknown},
     Journal = {Journal of the American Statistical Association},
     Volume = {62},
     Pages = {399-402},
        Year = {1967} }

1952

• T. W. Anderson and D. A. Darling, “Asymptotic theory of certain “goodness of fit” criteria based on stochastic processes,” Annals of Mathematical Statistics, iss. 23, pp. 193-212, 1952.
  [Bibtex]

  @article{
  Anderson-Darling52,
     Author = {Anderson, T. W. and Darling, D. A.},
     Title = {Asymptotic theory of certain "goodness of fit" criteria based on stochastic processes},
     Journal = {Annals of Mathematical Statistics},
     Number = {23},
     Pages = {193-212},
     Keywords = {Anderson-Darling normality test},
     Year = {1952} }

1951

• H. Robbins and S. Monro, “A stochastic approximation method,” Annals of Mathematical Statistics, vol. 22, pp. 400-407, 1951.
  [Bibtex]

  @article{
  Robbins+Monro51,
     Author = {Robbins, H. and Monro, S.},
     Title = {A stochastic approximation method},
     Journal = {Annals of Mathematical Statistics},
     Volume = {22},
     Pages = {400-407},
        Year = {1951} }

2009

• L. G. Barajas, A. L. Thomaz, and H. I. Christensen, “Whitepaper: Ushering in the Next Generation of Factory Robotics & Automation,” in National Workshop on Challenges to Innovation in Advanced Manufacturing: Industry Drivers and R&D Needs, Gaithersburg, MD, 2009.
  [Bibtex]

  @inproceedings{
  Barajas09NIST,
     Author = {Barajas, Leandro G. and Thomaz, Andrea L. and Christensen, Henrik I.},
     Title = {Whitepaper: Ushering in the Next Generation of Factory Robotics & Automation},
     BookTitle = {National Workshop on Challenges to Innovation in Advanced Manufacturing: Industry Drivers and R&D Needs},
     Editor = {National Institute of Standards and Technology},
     Address= {Gaithersburg, MD},
     Abstract = {The manufacturing capability and sustainability of the U.S. industry has been losing ground to its Asian and European competitors for the last few decades. For example, Japanese and German companies currently dominate the market of industrial Robotics and Automation (R&A) solutions with the support of low price Chinese manufacturers. Given the high labor cost in North American markets, the only viable option for U.S. industries to compete with a global market is via state-of-the-art R&A. Furthermore, most capital-intensive and wealth producing industries in the U.S. neither have the technical expertise nor the manufacturing capability to survive without cost effective R&A, which places these industries in a precarious state of vulnerability to disruptive technologies that may redefine the value stream map of their respective businesses.
  The unfortunate reality is that the domestic production of consumer products using conventional processes could soon cease to exist based on a 30-year track record of global outsourcing pressure toward regions with low labor and investment costs.  The transformational development and establishment of next generation manufacturing assembly processes using the latest in dexterous and intelligent robotics and lean production technologies will provide the necessary competitive edge for a variety of affordable products for the future.  As a result, jobs will be retained as some will shift from line work to technical support and operation of the robotic systems.
  The structured environment existent in current production facilities that enables robots to perform their tasks actually limits flexibility and drives a significant cost penalty for using robots. There has been some progress in enabling robots to operate in manufacturing operations with less structure, but robot capability in this area is very limited. This “robot capability gap” persists and limits the range of applications and business conditions under which robotics provide a feasible commercial alternative to other means of implementing manufacturing processes. This gap is especially evident in the automobile industry when examining the final assembly process.
  From an end user perspective, we believe that a new generation of assembly automation can be anticipated to significantly reduce the reliance on fixturing, mechanized structuring, and conventional sense-plan-act programming. This capability would enable assembly automation with a set of little or no more infrastructure requirements than a completely manual process would. These new assembly processes will exploit the existence of a flexible robot perception system as an integral component of a three-part strategy that includes: 1) highly flexible robots/end effectors, 2) flexible perception, and 3) safe integration/harmony with people, which are also performing tasks in the assembly process. The cognitive component of the perception system would facilitate the “assignment” of the automation to a set of assembly tasks and/or assistance to others performing a task not yet appropriate for automation. This capability will also enable the rapid “reassignment” of the automation to other tasks as required by production mix and business needs. Many U.S. manufacturing domains stand to benefit from the flexibility and productivity that this form of dynamic automation brings to the assembly process.   Multi-purpose robots that can safely collaborate with human workers will elevate the capabilities of existing assembly workers in the pursuit of providing quality products to end-users.
  A key factor in creation and adoption of the next generation manufacturing technologies is the development of flexible perception and human-like control technologies. In addition, by taking a leadership role in the development and adoption of such emerging technologies, we could ensure that the jobs created in this new area stay in the U.S.  These jobs can only be created and retained if a technological edge can be found that overcomes the attraction to low-cost labor regions. Through the pervasive use of intelligent R&A that can be as flexible and as easily trained as people, related industry jobs could also be moved from offshore to the U.S. as a direct result of this new technical capability.
  Our goal is to see revolutionary advancements in dexterous robotics leveraged in a new energy efficient automation environment that combines the best possible mix of human and machine capabilities.  These next generation robots include “safe robot” technologies that allow the seamless integration of people and dexterous robots in one lean process.  The key factor for the success of this approach is that the new systems leverage the infrastructure and flexible material processes that traditional manual systems use rather than expensive and traditionally inefficient automation methods.  This substitution enables a substantial reduction in R&A support investment that can normally be up to 10 times the cost of the robot themselves.
  From a scientific point of view, this endeavor encompasses a wide range of disciplines. Even when current commoditized hardware capabilities are almost at the level required to enable us to cross the capability gap, the actual integrated control and communications software systems are still lagging due to the heavy burden of current legacy systems. The historical paradigm for controlling R&A systems relies upon the system designer being able to specify a priori every requirement and possible condition of the system. This approach leaves no room for changing conditions, adaptability, plasticity, and in general, learning.
  One of the main hindrances that is currently preventing the evolution of the next generation R&A is the lack of standards of performance and test methods. Every R&A manufacturer attempts to keep their customer base captive by having closed and mostly incompatible systems. Most of the major specifications of these systems are given in terms of mechanical or electrical characteristics rather than in terms of overall system performance. NIST could play a vital role in advent of the new wave of R&A technologies by facilitating the dialog among interested parties and establishing both system standards and evaluation metrics in order to be able to track the level of capability improvement of such systems. Such specifications should not only encompass hardware and software metrology targets, but also high-level system qualitative and quantitative capability measurements for standardized processes. In a way, this will enable an R&A revolution equivalent to the one observed on the computer industry in the mid 1980’s. Effects of this achievement will be reflected deep into the fabric of industry and ultimately into the entire society; but in this case instead of putting a computer in every home or pocket, it will enable the pervasive use of functional R&A in all areas of our daily lives, from the factory plant floor to even your kitchen floor.
  We predicate that there is a unique opportunity to make progress in this arena by harnessing collaborations between industry and academia.  Our existing collaboration between Georgia Tech and General Motors is one such good example.  In our collaborative efforts to bring cutting-edge R&A technology from the labs to the factory floor, we are forced to reconcile some of the real issues involved with integrating flexible R&A with existing manufacturing processes and to focus on technologies that deliver real value added to the end customer. },
        Year = {2009} }

2003

• L. G. Barajas, M. B. Egerstedt, E. W. Kamen, and A. Goldstein, “Process control in a high-noise environment using a limited number of measurements,” in American Control Conference, 2003. Proceedings of the 2003, Denver, CO, 2003, pp. 597-602.
  [Bibtex]

  @inproceedings{
  Barajas03,
     Author = {Barajas, Leandro G. and Egerstedt, Magnus B. and Kamen, Edward W. and Goldstein, Alex},
     Title = {Process control in a high-noise environment using a limited number of measurements},
     BookTitle = {American Control Conference, 2003. Proceedings of the 2003},
     Address= {Denver, CO},
     Volume = {1},
     Pages = {597-602},
     Abstract = {In this paper, we develop a hybrid control algorithm that generates adequate control values for processes where only a limited number of function evaluations are available for the control law generation. This situation arises, for example, in stencil printing processes in printed circuit board manufacturing, where the cost associated with multiple function evaluations is prohibitive. The proposed control scheme is based on a weak-search algorithm that can be used in the presence of large amounts of noise as well as when only a limited amount of information is known about the process. The control law is given by a modified version of a constrained conjugated-gradient method, transitioned into a windowed-smoothed block-form of the least-squares affine estimator.},
     Keywords = {LS, hybrid control, noise, least squares, affine estimator, conjugated gradient
  algorithm theory
  conjugate gradient methods
  function evaluation
  least squares approximations
  printed circuit manufacture
  printing industry
  process control
  constrained conjugated gradient method
  control law generation
  control values
  high noise environment
  hybrid control algorithm
  least square affine estimator
  multiple function evaluations
  printed circuit board manufacturing
  process control
  stencil printing processes
  windowed smoothed block form},
     Year = {2003} }

2002

• S. Aravamudhan, D. Santos, G. Pham-Van-Diep, and F. Andres, “A study of solder paste release from small stencil apertures of different geometries with constant volumes,” in Electronics Manufacturing Technology Symposium, 2002. IEMT 2002. 27th Annual IEEE/SEMI International, 2002, pp. 159-165.
  [Bibtex]

  @inproceedings{
  Aravamudhan02,
     Author = {Aravamudhan, S. and Santos, D. and Pham-Van-Diep, G. and Andres, F.},
     Title = {A study of solder paste release from small stencil apertures of different geometries with constant volumes},
     BookTitle = {Electronics Manufacturing Technology Symposium, 2002. IEMT 2002. 27th Annual IEEE/SEMI International},
     Pages = {159-165},
     Abstract = {Stencil printing is a critical first step in surface mount assembly. It is often cited that about 50% or more of the defects found in the assembly of PCBs are attributed to stencil printing. Manufacturing techniques for the assembly of certain flip chips, chip scale packages and fine pitch ball grid arrays are testing the limits of current stencil printing capabilities. A thorough understanding of basic stencil printing principles would facilitate the design of printers, stencils and pastes, and would ultimately permit the extension of reliable print techniques to the very fine print arena. For small apertures, solder paste volume and consistency are critical to solder joint reliability. The work described in this paper examines the release performance of various solder pastes from a variety of aperture sizes and geometries. The focus of this study is a comparison of square versus circular apertures when the nominal volume of paste to be deposited is kept constant. This method of study is contrasted with published work wherein squares versus circles have been studied, but, in those, the dimensions (not volumes) were the same (e.g., 12 mil diameter circle as compared to a 12 mil (on a side) square aperture).},
     Keywords = {Stencil Printing, Fine Feature, CSP, Flip Chips, Square vs. Circle},
     Year = {2002} }

• L. G. Barajas, E. W. Kamen, A. Goldstein, M. B. Egerstedt, and B. Small, “A closed-loop hybrid control algorithm for stencil printing,” in Surface Mount Technology Association International Conference (SMTA02), Boston, MA, 2002, pp. 51-58.
  [Bibtex]

  @inproceedings{
  Barajas02,
     Author = {Barajas, Leandro G. and Kamen, Edward W. and Goldstein, Alex and Egerstedt, Magnus B. and Small, Benjamin},
     Title = {A closed-loop hybrid control algorithm for stencil printing},
     BookTitle = {Surface Mount Technology Association International Conference (SMTA02)},
     Editor = {SMTA},
     Address= {Boston, MA},
     Pages = {51-58},
     Abstract = {In this paper a control algorithm, that generates adequate machine settings for the Stencil Printing Process (SPP) in Surface Mount Technology manufacturing, is presented.  The SPP is characterized by high process-noise levels, and by requiring constant solder-paste volume deposition at all times.  Due to the prohibitive cost associated with extensive testing, only a limited amount of information can be used in the decision-making.  The major constraint when controlling the SPP is that it is not possible to evaluate the output of the system an arbitrary number of times, since for each calculation the printing and inspection of at least one or even two new boards is required.  These difficulties are addressed by the controller proposed in this paper, which is based on an algorithm for generating a sequence of iterative values that converge to an optimum set of machine parameters for a desired solder paste volume.  The merit of the control is that it minimizes the variance and the steady state error of the weighted sample mean versus the desired height, which improves the quality of the process.  In addition, it considers print direction and different component types independently.  Finally, this controller also automatically corrects for the process errors associated with the loss of the solder-paste working-viscosity-point, which can occur due to unscheduled stops on the line, and for incorrect specification of machine parameters given by the operator.},
     Keywords = {Stencil printing process control, least squares, affine estimator, Volume: Single best predictor, LS},
     Year = {2002} }

• A. Marin, P. Svasta, and D. Simion-Zanescu, “Theoretical and experimental aspects in the depositing process of solder and adhesive pastes in surface mount technology,” in Polymers and Adhesives in Microelectronics and Photonics, 2002. POLYTRONIC 2002. 2nd International IEEE Conference on, 2002, pp. 150-154.
  [Bibtex]

  @inproceedings{
  Marin02,
     Author = {Marin, A. and Svasta, P. and Simion-Zanescu, D.},
     Title = {Theoretical and experimental aspects in the depositing process of solder and adhesive pastes in surface mount technology},
     BookTitle = {Polymers and Adhesives in Microelectronics and Photonics, 2002. POLYTRONIC 2002. 2nd International IEEE Conference on},
     Pages = {150-154},
     Abstract = {The paper contributes to a detailed comparison of some materials, like solder pastes and conductive adhesives, used in the assembling technology process of the electronic components, on interconnecting supports. The study dealt also with the analyze of different geometrical configurations, in the depositing processes of the above-mentioned materials. The goal is to determine an optimum correlation of the functional parameters, for the depositing process, in realizing prototype and low manufacturing series applications. The study is focused only on the technology process in depositing materials, for interconnecting components and electronic modules, the other steps in the technological flow being considered only like input and output data. From all known technological processes, it was selected the one adapted to our purpose and it were established the necessary physical and technical parameters. By these means, we have studied the physical properties of the materials, pointing out their characteristics and the influence of these ones on the depositing process. It were made critical observations, concerning the validity of the proposed models and it were pointed out the further researches that we must conduct, in order to obtain better practical results.},
     Keywords = {surface mount technology, adhesives, reflow soldering, rheology, surface tension, viscosity, non-Newtonian flow, polymer melts, conducting polymers, filled polymers, printed circuit manufacture, solder pastes, conductive adhesives, assembling technology},
     Year = {2002} }

• G. Pham-Van-Diep, “Real time visualization and prediction of solder paste flow in the circuit board print operation,” in Advanced Technology Symposium, Boston, MA, 2002, pp. 41-49.
  [Bibtex]

  @inproceedings{
  Pham-Van-Diep02,
     Author = {Pham-Van-Diep, Gerald},
     Title = {Real time visualization and prediction of solder paste flow in the circuit board print operation},
     BookTitle = {Advanced Technology Symposium},
     Editor = {SMTA},
     Address= {Boston, MA},
     Publisher = {SMTA},
     Pages = {41-49},
     Abstract = {Three studies are undertaken to understand the dependence of aperture fill and stencil release on solder paste print definition. The first study focuses on the role of pastes. Seven pastes are compared and ranked by release performance. Second, three stencil-forming techniques are compared. Chemical etch, laser machined and electro-formed stencils from two manufacturers are studied to determine critical design parameters. Third, a model describing the release mechanism including sheer and adhesion forces is developed and compared to experimental results. Additionally, a real-time visualization technique and a model for solder paste flow into stencil apertures during the squeegee operation are described and results are discussed.},
     Keywords = {Surface Mount, Fine Feature, Transfer Efficiencies, Solder Paste, Stencil Printing},
     Year = {2002} }

• R. R. Yager and V. Kreinovich, “Main ideas behind OWA lead to a universal and optimal approximation scheme,” in Fuzzy Information Processing Society, 2002. Proceedings. NAFIPS. 2002 Annual Meeting of the North American, 2002, pp. 428-433.
  [Bibtex]

  @inproceedings{
  Yager02,
     Author = {Yager, R.R. and Kreinovich, V.},
     Title = {Main ideas behind OWA lead to a universal and optimal approximation scheme},
     BookTitle = {Fuzzy Information Processing Society, 2002. Proceedings. NAFIPS. 2002 Annual Meeting of the North American},
     Pages = {428-433},
        Year = {2002} }

2001

• A. D. Alawani, K. Srihari, and J. DiLella, “Developing a robust stencil printing process for diverse interconnection technologies,” in Pan Pacific Symposium, Kauai, Hawaii, 2001.
  [Bibtex]

  @inproceedings{
  Alawani01,
     Author = {Alawani, Ashish D. and Srihari, K. and DiLella, Jude},
     Title = {Developing a robust stencil printing process for diverse interconnection technologies},
     BookTitle = {Pan Pacific Symposium},
     Address= {Kauai, Hawaii},
     Abstract = {The area array assembly process is extremely complicated due to the large number of material, equipment and process parameters involved. The intricacy has been enhanced by the increasing trend towards finer pitches and higher interconnects. The complexity of solder paste deposition increases with decrease in component pitch. The need for high assembly yields and product reliability, and the difficulties associated with rework have made tighter process control quite essential during area array assembly. That is why solder paste printing has become one of the most critical processes in area array assembly. There is not a single rule for stencil design and printing for area array packages. Due to the geometry and reliability differences, the level and nature of control is different for each of the different component types. Hence, developing a process for a variety of components together is a challenging task. In the present research effort, the no-clean stencil printing process was studied and characterized for diverse, advanced surface mount packages including large Input/Output (I/O) Plastic Ball Grid Arrays (PBGAs), Ceramic Ball Grid Arrays (CBGAs) Chip Scale Ball Grid Array (CSBGA) and a Micro BGA (?BGA). Additionally, printing was characterized for a Through-Hole Component (THC), which was to be reflow-soldered along with the area array components. The challenge was to identify a single stencil thickness and paste type for a diverse range of solder volumes with approximately 130,000 cubic mils of solder paste for the THC (with adequate hole-fill) at one extreme and at least 650 cubic mils for the ?BGA at the other. The approach used was to first find the optimal combinations for the sub-systems and then modify these combinations using mature consideration and calculated guesses to identify sub-optimal, but the most practical parameters for the entire system. Multiple experiments were carried out using a Design of Experiments (DOE) based approach. Three different stencil thickness and two paste types were studied to identify the optimal materials for the overall system. The initial experiment identified that solder volumes as low as for the ?BGA can be deposited successfully using a Type III paste, provided the appropriate area ratios are maintained. This can be achieved by controlling the ratio between the aperture feature size and stencil thickness. The optimal thickness for the diverse range of components on the test vehicle used for the present research was identified to be 6 mils. Three different aperture shapes, namely, circular, oval and square with rounded corners were evaluated for paste transfer and in every single case circular apertures performed the best. Adequate hole-fill could not be achieved for the THC by using a polyurethane squeegee. Hence, further experiments were conducted to establish a comparison between the print performances of a metal and a polyurethane squeegee. It was observed that the use of a metal squeegee increased the solder paste deposition at the area array sites by at least 20 - 30 % and resulted in a radical improvement in hole-fill (from 50 % previously) to 70 %. Comprehensive statistical analysis was performed to establish these findings. In summary, this paper comprehensively documents the various aspects involved in the development of a robust stencil printing process for a variety of interconnection technologies.},
     Keywords = {Design of Experiments, Surface Mount Technology, Printed Circuit Boards.},
     Year = {2001} }

• B. Beair, “Real time statistical process control of the screen print process,” in SMTA International, 2001.
  [Bibtex]

  @inproceedings{
  Beair01,
     Author = {Beair, Bill},
     Title = {Real time statistical process control of the screen print process},
     BookTitle = {SMTA International},
     Abstract = {Increased component density found on today's printed circuit boards (PCBs), coupled with faster assembly cycle time requirements and the increased cost of screen print defects, necessitate continuous process monitoring of manufacturing operations. Paste rheology, board warp, aperture size and many other assembly features drive the need for continuous monitoring of solder deposition volume. Current screen print equipment offers 2D and 3D paste inspection capability, but fails to archive the inspection data to permit continuous monitoring of solder volume. Access to historical data, trend analysis, and data set comparisons facilitates process monitoring. This paper provides an insight into continuous monitoring and the positive effects it can produce in the electronics market. Defining the Paste Deposition Process Window There are many factors that come into play when defining solder volume control limits. To determine the control limits the minimum and maximum solder volume must be determined. Factors to consider when evaluating solder volume include plating on the circuit card, plating on the component leads and the component lead and circuit card pad dimensions. The online IPC-SM-782-Land Pattern Calculator (www.ipc.org/html/fsresources.htm) is a good tool to establish initial control limits. The control limits can then be compared to the theoretical print volume calculated from the stencil apertures. Stencil apertures aspect ratios will change the mean of the volume printed.},
        Year = {2001} }

• G. Beliakov, “Methods of construction of OWA operators from data,” in Fuzzy Systems, 2001. The 10th IEEE International Conference on, 2001, pp. 184-187.
  [Bibtex]

  @inproceedings{
  Beliakov02,
     Author = {Beliakov, G.},
     Title = {Methods of construction of OWA operators from data},
     BookTitle = {Fuzzy Systems, 2001. The 10th IEEE International Conference on},
     Volume = {1},
     Pages = {184-187},
     Abstract = {This paper investigates the problem of obtaining the weights of the ordered weighted aggregation (OWA) operators from observations. The problem is formulated as a restricted least squares and uniform approximation problems. We take full advantage of the linearity of the problem. In the former case, a well known technique of non-negative least squares is used. In a case of uniform approximation, we employ a recently developed cutting angle method of global optimisation. Both presented methods give results superior to earlier approaches, and do not require complicated nonlinear constructions. Additional restrictions, such as degree of orness of the operator, can be easily introduced},
     Keywords = {convergence
  convergence of numerical methods
  cutting angle method
  deterministic global optimisation
  fuzzy set theory
  least squares approximations
  linear least squares
  optimisation
  ordered weighted aggregation operators
  weights learning},
     Year = {2001} }

• W. E. Coleman, “Stencil print performance studies,” in SMTA International, Chicago, IL, 2001.
  [Bibtex]

  @inproceedings{
  Coleman01,
     Author = {Coleman, William E.},
     Title = {Stencil print performance studies},
     BookTitle = {SMTA International},
     Address= {Chicago, IL},
     Abstract = {Print performance of Electroformed and Laser-Cut stencils are studied over a wide range of aperture sizes. Relative paste volume, dispersion of paste volume, paste height, and dispersion of paste height are the measurables used to judge the stencil print performance. A new evaluation tool "PAR slope" (the slope of the curve obtained for a scatter plot of relative paste volume vs. Print Area Ratio) is also used to judge the stencil print performance. Results of this test show that Electroformed stencils provide acceptable solder paste bricks for Print Area Ratios down to as low as .5, much lower that the normally guideline of .65. This is a significant result for applications like 16 mil pitch QFP’s, uBGA’s, and Flip Chip’s. A stencil technology selection guideline based on Print Area Ratios and SMD component types is proposed.The typical evaluation to measure stencil performance has been to look at % paste release and/or the dispersion of solder paste volume for specific components / aperture sizes. This study looks at variety of responses over a wide range of aperture sizes. It deals with issues like how consistent is the % paste release and the paste height over a broad range of Print Area Ratios; what is the average relative paste volume over a range of Print Area Ratios; what is the standard deviation of the relative solder paste volume and solder paste height over a range Print area Ratios. From the data and corresponding responses a stencil ranking system is developed for the four stencil types tested. A stencil technology selection design guideline is proposed based on specific aperture designs / stencil thickness. The study indicates that under a fixed set of process / equipment conditions E-FAB stencils provide acceptable print performance down to a Print Area Ratio of .5.},
     Keywords = {stencil, print area ratio, solder volume, PCB, laser, and electroformed.},
     Year = {2001} }

• A. Johnson and A. Flori, “High Density/Fine Feature Solder Paste Printing,” in Electronics Assembly Process Exhibition and Conference, APEX, San Diego, CA, 2001.
  [Bibtex]

  @inproceedings{
  Johnson01,
     Author = {Johnson, Alden and Flori, Albert},
     Title = {High Density/Fine Feature Solder Paste Printing},
     BookTitle = {Electronics Assembly Process Exhibition and Conference, APEX},
     Address= {San Diego, CA},
     Abstract = {To achieve low PPM defects in solder paste printing of high density and fine feature sizes certain rules must be followed in the original manufacturing design.  The stencil printer must be capable of aligning to the substrate but there are six other factors that have to be considered before a consistent print can be achieved.   These factors include stencil thickness; aperture size and shape (hole wall aspect ratio) squeegee type and paste particle size. This paper will discuss the parameters that should be optimized to achieve high yields and some possible trade off if you are not willing to incorporate these rules in your manufacturing design.__________Solder paste printing in the Surface Mount industry has come a long way in the in the past fifteen years.  At first the 50 mil pitch devices were a problem as far as defect rate, now as packages shrink in size and increase in lead count we see the defect rate very high 100 to 200 PPM defect rate on the 20 mill pitch and below while manufacturing sees six sigma quality rates on the larger pitch components.  It is well documented that stencil printing contributes to high defect rates for the surface mount assembly process the root cause of the defect may be too complex to ascertain.   The paste printing process has four major variables, printer, stencil, substrate and solder pasteBased on the analysis results, the following is recommended:The following factor combinations provide the best print quality and consistent solder paste heights: a stencil aperture at 10 mils with square shapes, stencil thickness of 4, solder paste type 5, and polyethylene blades 90 durometer.  If trade offs are to be made the defect rates will increase resulting in higher manufacturing costs repair costs or warranty repairs.},
     Keywords = {ANOVA},
     Year = {2001} }

• M. Painaik, D. L. Santos, and J. A. McLenaghan, “Guidelines for fine feature stencil printing,” in SMTA International, Chicago, IL, 2001.
  [Bibtex]

  @inproceedings{
  Painaik01,
     Author = {Painaik, Mandar and Santos, Daryl L. and McLenaghan, A. James},
     Title = {Guidelines for fine feature stencil printing},
     BookTitle = {SMTA International},
     Address= {Chicago, IL},
     Abstract = {Recent technology advances and the evolution of components such as flip chips, chip scale packages, 0201s, and fine pitch leaded devices, have presented process challenges for most electronics manufacturers. The focus of this paper addresses the issues relative to fine feature component stencil printing on Printed Wiring Boards (PWBs). The paper discusses the stencil design, optimization of stencil printing and stencil wipe frequency for 0201 and CSP/micro-Ball Grid Array components. The fine feature stencil printing process is quantified for repeatable, efficient manufacturing in high volume production.Today, CSPs and 0201 devices are assembled in portable electronic products. These products are subject to harsh conditions during their service life and reliability is a concern. CSPs and 0201 components must reside on boards with other non-fine pitch devices. Therefore, the ability to deposit solder not only for CSPs and 0201s but also for SMT components is still a necessity. Printingrelated defects are common for these components due to area ratio violations. It is common because the transfer efficiency for CSPs is less that 60%. Overprinting is one solution, but it can lead to bridging and solder balling. Proper stencil design can help improve the transfer efficiency and increase the consistency of the process without over-sizing of the apertures. The stencil design parameters, apart from size, that need careful consideration are aperture shape, taper and electropolish. It can be seen from the above results that medium taper and high polish work the best for CSP and 0201 applications. Also, for CSP applications, it is recommended to use a square aperture for higher volume transfer.},
     Keywords = {BGA, CSP, 0201, stencil design.},
     Year = {2001} }

• M. S. Ramkumar, R. Clouthier, and R. Ghaffarian, “Print process characterization for BGA, fine pitch BGA and CSP components,” in SMTA International, Chicago, IL, 2001.
  [Bibtex]

  @inproceedings{
  Ramkumar01,
     Author = {Ramkumar, S. Manian and Clouthier, Richard and Ghaffarian, Reza},
     Title = {Print process characterization for BGA, fine pitch BGA and CSP components},
     BookTitle = {SMTA International},
     Address= {Chicago, IL},
     Abstract = {The goal of this characterization study is to conduct a formal design of experiment, to understand the impact of three important print process parameters, along with BGA, CSP pitch and I/O density on solder paste print quality and volume. The primary objective is to determine an optimum stencil print process for mixed technology PWB, containing various BGA, Fine Pitch BGA (FPBGA) and CSP components, ranging in pitch from 0.5 to 1.27 mm and I/Os from 48 to 784. The experiment was carried out using OSP coated test vehicles, provided by JPL, containing a myriad of BGA and CSP components. No Clean paste was selected for this study, with Type III and Type IV particle size and viscosity ranging from 900-1000 Kcps. Other parameters for the experiment included: Stencil Type - E-Fab Stencil Thickness - 5.65 and 4.65 mils Print Mechanism - Metal Blade 45 degree angle Print Speed - 2 mm/sec Print Pressure - Optimum setting to achieve good rolling of paste in front of the squeegee As part of the experiment, RIT conducted both qualitative and quantitative analysis. The qualitative analysis included visual examination of the actual bricks of paste deposited and recording any anomalies observed. The quantitative analysis involved the measurement of true 3D volume and height of solder paste print. This paper will present the findings of this study in detail.The analysis carried out at RIT indicates that the area aspect ratio plays a very important role in the paste printing process for mixed technology boards, containing various BGA, FPBGA and CSP devices. Other parameters that are also important include solder particle size, stencil thickness, stencil type and print parameters. Figure 8 shows that consistent and acceptable % transfer can be obtained for most of the BGA, Fine Pitch BGA and CSP components on the board (>60% transfer). The one device that did not show good results was U12, which has the smallest aperture size and pitch (0.5mm) when compared to other components. This indicates that further investigation is necessary, utilizing stencils which are thinner than 4.65 mils and different Type IV pastes. The best transfer (>120%) was obtained for component U3, which is a TSOP (0.8mm pitch), with rectangular apertures. Aperture shape was not considered as part of this experiment but could be a considered as a factor in future experimentation. Component U5, which is a 0.25mm pitch flip chip, was not used in the analysis, because the print quality was observed to be consistently very poor for the given print parameters. The inspection system was also not equipped with the proper magnification to inspect such a small device. Further experimentation is currently being planned, to include additional factors such as No-Lead paste and a thinner stencil (<4.65mils).},
        Year = {2001} }

• A. Rybalov and R. R. Yager, "Controlled clustering, uni-norm operators and OWA operators," in IFSA World Congress and 20th NAFIPS International Conference, 2001. Joint 9th, 2001, pp. 1935-1939.
  [Bibtex]

  @inproceedings{
  Rybalov+Yager01,
     Author = {Rybalov, A. and Yager, R.R.},
     Title = {Controlled clustering, uni-norm operators and OWA operators},
     BookTitle = {IFSA World Congress and 20th NAFIPS International Conference, 2001. Joint 9th},
     Volume = {4},
     Pages = {1935-1939},
     Abstract = {Clustering processes now have widespread applications. Their purpose is to separate data into groups of similar characteristics. Usually these processes are data driven, and, thus, we don't have effective mechanism to control them. The standard procedure is to set the number of clusters (in supervised clustering we can also set centers of clusters). But as soon as we fix the number of clusters we don't have further control on the number of points in each cluster. So, the problem now is how to manage the distribution of points. The distribution of points in clusters reflects the level of concentration: if almost all points are in one cluster then the level of concentration is very high; if points are equally distributed across clusters then the level of concentration is low. Therefore, the first problem is to find the way to measure the level of concentration. After this we proceed to describe the clustering process that permits us to control concentration. We then apply uni-norm operators and OWA operators to managing of the clustering process},
     Keywords = {artificial intelligence, pattern clustering},
     Year = {2001} }

• W. Sauer and H. Wohlrabe, "Machine and process capability coefficient of solder paste printers," in Electronics Technology: Concurrent Engineering in Electronic Packaging, 2001. 24th International Spring Seminar on, Dept. of Electr. Eng., Tech. Univ. Dresden, Germany, 2001, pp. 26-31.
  [Bibtex]

  @inproceedings{
  Sauer01,
     Author = {Sauer, W. and Wohlrabe, H.},
     Title = {Machine and process capability coefficient of solder paste printers},
     BookTitle = {Electronics Technology: Concurrent Engineering in Electronic Packaging, 2001. 24th International Spring Seminar on},
     Editor = {Dumitrache, I. and Svasta, P.},
     Address= {Dept. of Electr. Eng., Tech. Univ. Dresden, Germany},
     Pages = {26-31},
     Abstract = {For manufacturers of assembled boards seeking ISO 9000 certification, a number of requirements are placed on measurement instruments and tooling. This standard requires a system to document and maintain calibration and machine capabilities. One of these problems is the measurement of printing accuracies of solder paste printers. The paper presents: short theoretical overview of machine capabilities; methods of measurement of accuracies (x, y and rotation); recommendations for the calibration of printers; indications for the sources of inaccuracies; practical examples and results.},
     Keywords = {certification, ISO standards, calibration, measurement standards, soldering, assembling, printed circuit manufacture, process capability coefficient, machine capability coefficient, solder paste printers, assembled board manufacture, ISO 9000 certificat},
     Year = {2001} }

2000

• G. P. Glinski, C. Bailey, and K. Pericleous, "Simulation of the stencil printing process [solder pastes]," in Electronic Materials and Packaging, 2000. (EMAP 2000). International Symposium on, Centre for Numerical Modelling and Process Analysis, Greenwich Univ., London, UK, 2000, pp. 364-370.
  [Bibtex]

  @inproceedings{
  Glinski00,
     Author = {Glinski, G.P. and Bailey, C. and Pericleous, K.},
     Title = {Simulation of the stencil printing process [solder pastes]},
     BookTitle = {Electronic Materials and Packaging, 2000. (EMAP 2000). International Symposium on},
     Editor = {Kim, J.K. and Teng, A. and Lee, S-W R.},
     Address= {Centre for Numerical Modelling and Process Analysis, Greenwich Univ., London, UK},
     Pages = {364-370},
     Abstract = {This paper describes the application of advanced computational fluid dynamics (CFD) methods to model the stencil printing process at both macroscopic and microscopic length scales. The solder paste flow is simulated on a macroscopic scale to study flow characteristics of the bulk motion of the whole paste roll as it travels ahead of a squeegee blade. The studies are performed using finite volume CFD methods, treating the paste as a continuum. Constitutive relationships are used with experimentally obtained parameters to represent the nonNewtonian properties of solder pastes. Shear rate and pressure distributions are obtained and applied as boundary conditions for microscopic studies, in which lattice-Boltzmann methods are used to couple the simultaneous solution of the solder particle dynamics and flux flow during aperture filling and stencil withdrawal. Discrete element method (DEM) type algorithms are used to control particle collisions. The simulation results are intended to provide insight into the phenomena which occur during the printing of solder paste in fine-pitch applications.An outline of a number of multi-scale modelling methods have been presented, which are being integrated to study macroscopic and microscopic mechanisms of the solder paste stencil printing process. Finite volume methods have been used to obtain shear rate and pressure data to define boundary conditions of microscale simulations in which coupled solutions of flux flow and solder particle transport are solved simultaneously using Lattice-Boltzmann techniques for the particle-fluid interaction, integrated with discrete element type methods to control particle collisions. The microscopic simulations to date are two- dimensional. Such simulations are beneficial to the development of the methods and code as numerous benchmark cases exist for validation purposes. The modelling of flux flow around spherical solder particles in two dimensions is not realistic since the particles are represented as cylindrical objects. Work is now focused on extending methods to three dimensions and will be validated against appropriate benchmark studies of hydrodynamic forces on spheres. Optimum lattice densities need to be established for efficiency/accuracy trade-off. Full three-dimensional simulations of solder particles in various aperture geometries will follow these studies. Experimental measurement and visualization studies are also planned to validate both the macroscopic and microscopic simulation results.},
     Keywords = {computational fluid dynamics, soldering, assembling, printed circuit manufacture, electronic engineering computing, flow simulation, fine-pitch technology, finite volume methods, Boltzmann equation, stencil printing process simulation, computational flu},
     Year = {2000} }

• W. Kim and G. Vachtsevanos, "Hierarchical process control by combining SPC and soft computing methods," in Fuzzy Information Processing Society, 2000. NAFIPS. 19th International Conference of the North American, 2000, pp. 485-489.
  [Bibtex]

  @inproceedings{
  Kim+Vachtsevanos00,
     Author = {Kim, Wonoh and Vachtsevanos, G.},
     Title = {Hierarchical process control by combining SPC and soft computing methods},
     BookTitle = {Fuzzy Information Processing Society, 2000. NAFIPS. 19th International Conference of the North American},
     Editor = {Whalen, T.},
     Pages = {485-489},
     Abstract = {Statistical process control (SPC) provides methods of monitoring a system to improve the quality of the product. However, using SPC alone has limitations since it does not control a system but rather monitors it to remove the root causes. SPC and feedback control are combined in a hierarchical structure to monitor a system and use the information to compensate for deviations of the output of a system. Most industrial systems are composed of several constituent subsystems that are coupled. Large-scale system (LSS) theory and hierarchical control structure are used to not only optimize the recipe for the individual subsystems but also to optimize for the entire system. The proposed algorithm will be implemented in the controlled atmosphere brazing (CAB) furnace to braze aluminum heat exchange components.},
     Keywords = {statistical process control, computerised monitoring, feedback, quality control, aluminium, electric furnaces, brazing, fuzzy control, hierarchical process control, SPC, soft computing methods, statistical process control, system monitoring, feedback co},
     Year = {2000} }

• T. O. Kvalseth, "On exponential entropies," in Systems, Man, and Cybernetics, 2000 IEEE International Conference on, 2000, pp. 2822-2826 vol.4.
  [Bibtex]

  @inproceedings{
  Kvalseth00,
     Author = {Kvalseth, T.O.},
     Title = {On exponential entropies},
     BookTitle = {Systems, Man, and Cybernetics, 2000 IEEE International Conference on},
     Volume = {4},
     Pages = {2822-2826 vol.4},
     Abstract = {As a number of information, an entropy has been defined as the weighted mean of a set of exponential functions involving the probabilities of a set of random events. The exponential entropy is claimed to have certain advantages over the classical Shannon entropy (C.E. Shannon, 1948). The article proposes two different generalizations of the exponential entropy, each of which represents a one-parameter generalization. Shannon's entropy is shown to be a particular member of one of these two new families of information measures. Some of the important properties of the new measures are discussed},
     Keywords = {computational complexity, entropy, probability},
     Year = {2000} }

• N. Morad, H. K. Yii, M. S. Hitam, and C. P. Lim, "Development of an intelligent system for the solder paste printing process," in TENCON 2000. Proceedings, Sch. of Ind. Technol., Univ. Sains Malaysia, Penang, Malaysia, 2000, pp. 479-483.
  [Bibtex]

  @inproceedings{
  Morad00,
     Author = {Morad, N. and Yii, Hee Kim and Hitam, M.S. and Lim, Chee Peng},
     Title = {Development of an intelligent system for the solder paste printing process},
     BookTitle = {TENCON 2000. Proceedings},
     Address= {Sch. of Ind. Technol., Univ. Sains Malaysia, Penang, Malaysia},
     Volume = {3},
     Pages = {479-483},
     Abstract = {A neurogenetic-based hybrid framework is developed where the main components within the framework are artificial neural networks (ANNs) and genetic algorithms (GAs). The investigation covers a mode of combination or hybridisation between the two components that is called task hybridisation. The combination between ANNs and GAs using task hybridisation leads to the development of a hybrid multilayer feedforward network, trained using supervised learning. This paper discusses the GA method used to optimize the process parameters, using the ANN developed as the process mode, in a solder paste printing process, which is part of the process in the surface mount technology (SMT) method. The results obtained showed that the GA-based optimization method works well under various optimization criteria.},
     Keywords = {printing, surface mount technology, feedforward neural nets, genetic algorithms, learning (artificial intelligence), electronic engineering computing, process control, neurocontrollers, optimal control, intelligent system, solder paste printing process,},
     Year = {2000} }

• T. A. Nguty, S. Budiman, D. Rajkumar, R. Solomon, N. N. Ekere, and M. A. Currie, "Understanding the process window for printing lead-free solder pastes," in Electronic Components and Technology Conference, 2000. 2000 Proceedings. 50th, Electron. Manuf. Eng. Res. Group, Salford Univ., UK, 2000, pp. 1426-1435.
  [Bibtex]

  @inproceedings{
  Nguty00,
     Author = {Nguty, T. A. and Budiman, S. and Rajkumar, D. and Solomon, R. and Ekere, N. N. and Currie, M. A.},
     Title = {Understanding the process window for printing lead-free solder pastes},
     BookTitle = {Electronic Components and Technology Conference, 2000. 2000 Proceedings. 50th},
     Address= {Electron. Manuf. Eng. Res. Group, Salford Univ., UK},
     Pages = {1426-1435},
     Abstract = {Solder paste is primarily used as a bonding medium for surface mount assemblies (SMA) in the electronics industry. The solder paste is typically deposited using the stencil printing process. The stencil printing of solder paste is a very important and critical stage in the reflow soldering of surface mount devices. A high proportion of all SMA defects is related to the stencil printing process. This is likely to continue with the drive towards the introduction of lead-free solder pastes. Much work is continuing on the metallurgical properties of these lead-free solders, including solder joint strength and material compatibility. However, the initial challenge for the new Pb-free formulations is in achieving repeatable solder deposit from print to print and from pad to pad. To meet this challenge, new formulations in flux medium are now being developed. For a smooth transition to Pb-free soldering formulations, a proper understanding of the printing performance of the new solder pastes is necessary. The key parameters that affect solder paste printing have been identified and the subject of numerous studies. In lead-free solder paste, the replacement of lead with other elements (including Bismuth, Copper) changes the density of this dense suspension (solder paste). In this paper, we investigate the effects of the printer parameters, i.e. squeegee speed and pressure (defined as the process window) on the printing performance of a variety of lead-free solder pastes. A three-level design of experiment on these factors (pressure and speed) was used. Comparisons will be presented with lead-rich solder pastes. The metal content of the lead-free solders had a significant effect of the on the process window.},
     Keywords = {Reflow soldering, Environmental factors, Surface mount technology, Design of experiments, process window, lead-free solder paste, electronic assembly, stencil printing, reflow soldering, surface mount device, design of experiments, defects related SPP},
     Year = {2000} }

• H. L. J. Pang, K. H. Ang, X. Q. Shi, and Z. P. Wang, "Methodology for a highly accelerated solder joint reliability test," in Electronics Packaging Technology Conference, 2000. (EPTC 2000). Proceedings of 3rd, 2000, pp. 385-390.
  [Bibtex]

  @inproceedings{
  Pang00,
     Author = {Pang, H.L.J. and Ang, K.H. and Shi, X.Q. and Wang, Z.P.},
     Title = {Methodology for a highly accelerated solder joint reliability test},
     BookTitle = {Electronics Packaging Technology Conference, 2000. (EPTC 2000). Proceedings of 3rd},
     Editor = {Beng, L.T. and Lee, C. and Chuan, T.K.},
     Pages = {385-390},
     Abstract = {A thermo-mechanical deflection system (TMDS) test method has been developed for evaluating the fatigue performance of solder joints in printed circuit board (PCB) assemblies. The TMDS test imparts cyclic twisting deflections on an assembled PCB test vehicle which is tested under controlled isothermal conditions in a thermal chamber. Tests were conducted at various angles of twist and isothermal conditions at 100/spl deg/C and 25/spl deg/C. Failure analysis using SEM showed that the solder joint failure for the TMDS test is comparable to the failure mechanisms for accelerated thermal cycling (ATC) test failures for solder joints. Weibull failure distribution plots for TMDS and ATC tests were compared and a scale factor (SF) was used to correlate the test results.},
     Keywords = {circuit reliability, failure analysis, printed circuit testing, soldering, assembling, life testing, fatigue, Weibull distribution, highly accelerated solder joint reliability test, thermo-mechanical deflection system, TMDS test method, fatigue performa},
     Year = {2000} }

• D. Rajkumar, T. A. Nguty, and N. N. Ekere, "Optimising process parameters for flip chip stencil printing using Taguchi's method," in Electronics Manufacturing Technology Symposium, 2000. Twenty-Sixth IEEE/CPMT International, Sch. of Aeronaut. Mech. & Manuf. Eng., Salford Univ., UK, 2000, pp. 382-388.
  [Bibtex]

  @inproceedings{
  Rajkumar00,
     Author = {Rajkumar, D. and Nguty, T. A. and Ekere, N. N.},
     Title = {Optimising process parameters for flip chip stencil printing using Taguchi's method},
     BookTitle = {Electronics Manufacturing Technology Symposium, 2000. Twenty-Sixth IEEE/CPMT International},
     Address= {Sch. of Aeronaut. Mech. & Manuf. Eng., Salford Univ., UK},
     Pages = {382-388},
     Abstract = {Solder paste printing is an important process in surface mount device assembly using the reflow soldering technique. There is wide agreement in the industry that the paste printing process accounts for the majority of assembly defects, and most defects originate from poor understanding of the effect of printing process parameters on printing performance and the nature of their interactions. The key solder paste printing process parameters considered in this study are the squeegee pressure, squeegee speed, stencil-substrate separation speed and squeegee print direction. Previous work shows that these process parameters affect printing process performance. As the current product miniaturisation trend continues for hand-held consumer products, area array type package solutions such as chip scale packages and flip chip are now being designed into products. Assembly of these devices requires the printing of very small solder paste deposits consistently from pad to pad, and from board to board. This paper concerns the determination of the solder paste printing process window for flip chip assembly applications. Five different solder paste formulations, specially formulated for flip-chip assembly, were evaluated as part of a broader study on low cost solder bumped flip-chip assembly. The results were also used for establishing guidelines for printing solder pastes for both solder bumping and the flip-chip assembly process. The experimental design for the study was based on the Taguchi method. A 2-level and 4-factor orthogonal array was used to investigate the main effects.},
     Keywords = {reflow soldering, Taguchi methods, design of experiments, flip-chip devices, microassembling, integrated circuit packaging, integrated circuit interconnections, optimisation, process parameter optimization, flip chip stencil printing, Taguchi method, so},
     Year = {2000} }

• M. S. Ramkumar and J. Lemery, "High speed stencil printing of solder paste a DOE full factorial experiment," in SMTA International, Maui, Hawaii, 2000.
  [Bibtex]

  @inproceedings{
  Ramkumar00,
     Author = {Ramkumar, S. Manian and Lemery, Jason},
     Title = {High speed stencil printing of solder paste a DOE full factorial experiment},
     BookTitle = {SMTA International},
     Address= {Maui, Hawaii},
     Abstract = {This paper is aimed at presenting the results of a full factorial DOE on High Speed Stencil Printing of Solder Paste. The experiment was funded by the TRW Foundation and was conducted at RIT’s Center for Electronics Manufacturing. The various factors that were toggled as part of the experimental design were stencil construction, stencil thickness, paste formulation, and print speed. The levels for these factors included: two stencil constructions, two stencil thicknesses, two no-clean paste formulations, and two print speeds. The response that will be used in the analysis is print volume. In order to analyze the print characteristics for a wide range of component pitches, test boards were designed to accommodate 50, 20 and 16 mil pitch components. The experiment will be conducted for various combinations of factors and levels to obtain a valid full factorial experiment and the measured responses will be analyzed using MiniTAB software. The analysis is expected to provide information pertaining to the most influential factors. The experiment details, test board design and results are presented in detail in this paper.The analysis technique would involve the plotting of a Half Normal and Pareto chart, to identify the most influential factors. Mean solder paste volume, with respect to each of the levels of the factor, will also be plotted, in order to determine the nature of influence of the factors at various levels. Interaction plots would be plotted to analyze the combined effects of the different factors and levels. The analysis would also include the development of a prediction equation using regression analysis techniques and the determination of optimum setpoints for obtaining the required solder paste volume. Since the experiment is still underway we are not able to provide the actual analysis results in the paper. The results will be presented in detail at the conference.      This experiment will enable us to identify the factors and setpoints that will provide good process capability for the three different pitches being studied. Even though all of the process variables have not been included in this experiment, it will still provide a good understanding of the high-speed stencil printing process. Once optimum setpoints are determined, this experiment can be taken to a more elaborate level, by keeping those setpoints as constant parameters and including new factors.},
        Year = {2000} }

• V. Torra, "Learning weights for weighted OWA operators," in Industrial Electronics Society, 2000. IECON 2000. 26th Annual Conference of the IEEE, 2000, pp. 2530-2535.
  [Bibtex]

  @inproceedings{
  Torra00,
     Author = {Torra, V.},
     Title = {Learning weights for weighted OWA operators},
     BookTitle = {Industrial Electronics Society, 2000. IECON 2000. 26th Annual Conference of the IEEE},
     Volume = {4},
     Pages = {2530-2535},
     Abstract = {Weighted OWA (ordered weighted aggregation) operators were introduced as a generalization of the weighted mean (WM) and the OWA operators, so that the advantages of both could be used in a single data fusion function. In this work, we study the determination of their parameters when a set of examples is at our disposal. The approach presented in this paper is of interest in data mining when a certain variable has to be expressed in terms of some other ones. In this case, the learning of weights corresponds to fitting the model, and the weights correspond to the importance of the variables and of their values},
     Keywords = {data mining, learning by example, mathematical operators, sensor fusion},
     Year = {2000} }

• N. Van-Hop, M. T. Tabucanon, and D. Q. Minh, "PCB assembly sequence and feeder assignment problem for the case of Tchebyshev robot arm motion. I. Basic problem," in Management of Innovation and Technology, 2000. ICMIT 2000. Proceedings of the 2000 IEEE International Conference on, 2000, pp. 919-924.
  [Bibtex]

  @inproceedings{
  Van-Hop00a,
     Author = {Van-Hop, Nguyen and Tabucanon, M.T. and Minh, Do Quang},
     Title = {PCB assembly sequence and feeder assignment problem for the case of Tchebyshev robot arm motion. I. Basic problem},
     BookTitle = {Management of Innovation and Technology, 2000. ICMIT 2000. Proceedings of the 2000 IEEE International Conference on},
     Volume = {2},
     Pages = {919-924},
     Abstract = {This paper considers a basic problem in printed circuit board (PCB) assembly planning for the case of Tchebyshev robot arm motion in which it will be used for the other problems. In this problem, the assembly sequence is determined when feeder rack travel with infinite speed and the board is fixed. The problem is analyzed and converted into a shortest path problem of a corresponding interval graph that is proved to be NP-complete. Therefore, a heuristic procedure is developed based on network theory and interval graph. This heuristic is tested and found to give the closed optimal solution.},
     Keywords = {printed circuit manufacture, industrial robots, computational complexity, optimisation, assembly planning, PCB assembly sequence, feeder assignment problem, Tchebyshev robot arm motion, PCB assembly planning, feeder rack travel, infinite speed, shortest},
     Year = {2000} }

• N. Van-Hop, M. T. Tabucanon, and D. Q. Minh, "PCB assembly sequence and feeder assignment problem for the case of Tchebyshev robot arm motion. II. Models development," in Management of Innovation and Technology, 2000. ICMIT 2000. Proceedings of the 2000 IEEE International Conference on, 2000, pp. 925-930.
  [Bibtex]

  @inproceedings{
  Van-Hop00b,
     Author = {Van-Hop, Nguyen and Tabucanon, M.T. and Minh, Do Quang},
     Title = {PCB assembly sequence and feeder assignment problem for the case of Tchebyshev robot arm motion. II. Models development},
     BookTitle = {Management of Innovation and Technology, 2000. ICMIT 2000. Proceedings of the 2000 IEEE International Conference on},
     Volume = {2},
     Pages = {925-930},
     Abstract = {This paper continues part I to develop models and solutions for printed circuit board (PCB) assembly sequence and feeder assignment problems in the case of Tchebyshev robot arm motion. The problem is decomposed into sub-problems of feeder assignment and assembly sequence separately and the combined one under different scenarios of machine settings. The analysis of the problem leads to a number of nonlinear multiple objective decision making (MODM) models. These models are linearized and solved by heuristics using the solution procedure in part I and some standard techniques.},
     Keywords = {printed circuit manufacture, assembly planning, industrial robots, decision theory, PCB assembly sequence, feeder assignment, Tchebyshev robot arm motion, models development, sub-problems, machine settings, nonlinear multiple objective decision making m},
     Year = {2000} }

• H. Zhen, Q. Ershi, and L. Zixian, "Quality improvement through SPC/DOE in SMT manufacturing," in Management of Innovation and Technology, 2000. ICMIT 2000. Proceedings of the 2000 IEEE International Conference on, Sch. of Manage., Tianjin Univ., China, 2000, pp. 855-858.
  [Bibtex]

  @inproceedings{
  Zhen00,
     Author = {Zhen, He and Ershi, Qi and Zixian, Liu},
     Title = {Quality improvement through SPC/DOE in SMT manufacturing},
     BookTitle = {Management of Innovation and Technology, 2000. ICMIT 2000. Proceedings of the 2000 IEEE International Conference on},
     Address= {Sch. of Manage., Tianjin Univ., China},
     Volume = {2},
     Pages = {855-858},
     Abstract = {The paper addresses how to analyze and improve process quality in surface mount technology (SMT) manufacturing using multi-vari analysis (MVA), measurement system analysis (MSA), and design of experiment (DOE). The key process parameter (solder paste height) in the printing process has been optimized and defects caused by poor soldering have been reduced. The paper provides a good solution for increasing solderability of surface mount technology manufacturing through joint use of statistical process control (SPC) and DOE.},
     Keywords = {surface mount technology, statistical process control, quality control, design of experiments, soldering, SMT manufacturing, process quality improvement, surface mount technology, multi-vari analysis, measurement system analysis, design of experiment, p},
     Year = {2000} }

1999

• L. G. Barajas, "Real Time Kernel Controller for Oil Fields with Heterogeneous Variable Frequency Drives," Occidental Petroleum Corp., Oil & Gas Division, Telecommunications & SCADA Department, Engineering Report, Versions 1-3, , , 1999.
  [Bibtex]

  @techreport{
  oxy99,
     Author = {Barajas, Leandro G.},
     Title = {Real Time Kernel Controller for Oil Fields with Heterogeneous Variable Frequency Drives},
     Institution = {Occidental Petroleum Corp., Oil & Gas Division, Telecommunications & SCADA Department},
     Type = {Engineering Report, Versions 1-3,},
     Year = {1999} }

• M. Howarth and A. Lotfi, "Adaptive fuzzy control of solder paste printing: the identification of deposit defects," in Electronics Manufacturing Technology Symposium, 1999. Twenty-Fourth IEEE/CPMT, Dept. of Mech. & Manuf. Eng., Nottingham Trent Univ., UK, 1999, pp. 102-107.
  [Bibtex]

  @inproceedings{
  Howarth+Lotfi99,
     Author = {Howarth, M. and Lotfi, A.},
     Title = {Adaptive fuzzy control of solder paste printing: the identification of deposit defects},
     BookTitle = {Electronics Manufacturing Technology Symposium, 1999. Twenty-Fourth IEEE/CPMT},
     Address= {Dept. of Mech. & Manuf. Eng., Nottingham Trent Univ., UK},
     Pages = {102-107},
     Abstract = {This paper presents the use of fuzzy logic to implement feature extraction and defect classification of an adaptive fuzzy control algorithm for the control of the solder paste stencil printing stage of surface mount printed circuit board (PCB) assembly. Experimental results are presented to illustrate the capability of the algorithm.},
     Keywords = {surface mount technology, printed circuit manufacture, assembling, soldering, adaptive control, fuzzy control, fuzzy logic, circuit reliability, fault diagnosis, feature extraction, process control, adaptive fuzzy control, solder paste printing, deposit},
     Year = {1999} }

• M. Howarth, S. A. Silvester, M. Lacey, and K. Sivaygonathan, "An investigation into the printing characteristics and mechanical dynamics of advanced squeegee mechanisms," in Electronics Manufacturing Technology Symposium, 1999. Twenty-Fourth IEEE/CPMT, Dept. of Mech. & Manuf. Eng., Nottingham Trent Univ., UK, 1999, pp. 178-184.
  [Bibtex]

  @inproceedings{
  Howarth+Silvester99,
     Author = {Howarth, M. and Silvester, S.A. and Lacey, M. and Sivaygonathan, K.},
     Title = {An investigation into the printing characteristics and mechanical dynamics of advanced squeegee mechanisms},
     BookTitle = {Electronics Manufacturing Technology Symposium, 1999. Twenty-Fourth IEEE/CPMT},
     Address= {Dept. of Mech. & Manuf. Eng., Nottingham Trent Univ., UK},
     Pages = {178-184},
     Abstract = {The new breed of solder paste deposition systems, such as the DEK ProFlow, incorporate a contained solder paste chamber enabling independent control of machine parameters, pressure, down force and print speed. Previous work by Ekere et al. (1994) and Monaghan et al. (1991-2) have modelled conventional squeegee systems, but only part of this work is applicable to the new mechanism. Practical experimentation has been employed, both to gain an appreciation of the printing characteristics in general and to serve as reference for purely analytical models. Flow visualisation experiments have provided useful indications of flow dynamics within the print head, highlighting dominant flow patterns, velocity profiles and other phenomena such as evidence of air entrapment or cavitation. The FIDAP CFD package is used to perform detailed analysis, providing measures of data not practically obtainable through physical measurement. Pressure, shear and temperature profiles are obtained for a range of operating conditions and their influence on ink/solder paste transfer can be evaluated, leading to the identification of operational limits. Results from the FIDAP model compare very well with the experimental work and have identified the critical areas of the squeegee mechanism. This analysis is now being used to refine the process, to extend its application to lower viscosity materials and to incorporate simpler control systems than have been found to be necessary with conventional squeegee mechanisms (Lotfi and Howarth, 1998; Zhuang et al., 1997).},
     Keywords = {soldering, assembling, printed circuit manufacture, flow visualisation, computational fluid dynamics, electronic engineering computing, viscosity, cavitation, printing characteristics, mechanical dynamics, squeegee mechanisms, solder paste deposition sy},
     Year = {1999} }

• T. A. Nguty, N. N. Ekere, and A. Adebayo, "Correlating solder paste composition with stencil printing performance," in Electronics Manufacturing Technology Symposium, 1999. Twenty-Fourth IEEE/CPMT, Dept. of Aeronaut. Mech. & Manuf. Eng., Salford Univ., UK, 1999, pp. 304-312.
  [Bibtex]

  @inproceedings{
  Nguty99,
     Author = {Nguty, T. A. and Ekere, N. N. and Adebayo, A.},
     Title = {Correlating solder paste composition with stencil printing performance},
     BookTitle = {Electronics Manufacturing Technology Symposium, 1999. Twenty-Fourth IEEE/CPMT},
     Address= {Dept. of Aeronaut. Mech. & Manuf. Eng., Salford Univ., UK},
     Pages = {304-312},
     Abstract = {Soldering technologies continue to evolve, particularly in the area of solder pastes used in SMD reflow soldering. Solder pastes typically consist of solder alloy powder, flux, viscosity control agents, and a solvent system. By varying solder particle size, distribution and shape, as well as the other constituent materials, solder paste rheology and printing performance can be controlled. As PCB pad and component lead sizes reduce to meet ultra-fine pitch and flip-chip assembly requirements, most paste suppliers have opted for smaller powder particle sizes and new rheology modifier formulations. Successful assembly at these small-scale geometries requires deposition of small and consistent paste deposits from pad to pad, and from board to board. Two primary mechanisms dominate paste printing at this chip-scale geometry: paste transfer into stencil apertures, and paste release from the apertures on to PCB pads, i.e. paste flow behaviour is crucial in defining printing performance. Solder paste flow properties primarily depend on their chemical and physical structure, and hence composition. In this paper, we study the rheological profile of various paste formulations, and how this can be used to provide better understanding of composition effects on stencil printing performance. Pastes for ultra-fine pitch and flip chip applications were studied via a wide range of rheological tests, including the creep recovery test to determine paste slump characteristics and the oscillatory test to determine visco-elastic behaviour required for two critical sub-processes: aperture filling and paste withdrawal.},
     Keywords = {reflow soldering, surface mount technology, printed circuit manufacture, rheology, viscosity, particle size, recovery-creep, viscoelasticity, printed circuit testing, assembling, flip-chip devices, fine-pitch technology, integrated circuit packaging, cr},
     Year = {1999} }

• J. Pan, G. L. Tonkay, R. H. Storer, R. M. Sallade, and D. J. Leandri, "Critical variables of solder paste stencil printing for micro-BGA and fine pitch QFP," in Electronics Manufacturing Technology Symposium, 1999. Twenty-Fourth IEEE/CPMT, Dept. of Ind. & Manuf. Syst. Eng., Lehigh Univ., Bethlehem, PA, USA, 1999, pp. 94-101.
  [Bibtex]

  @inproceedings{
  Pan99,
     Author = {Pan, Jianbiao and Tonkay, G.L. and Storer, R.H. and Sallade, R.M. and Leandri, D.J.},
     Title = {Critical variables of solder paste stencil printing for micro-BGA and fine pitch QFP},
     BookTitle = {Electronics Manufacturing Technology Symposium, 1999. Twenty-Fourth IEEE/CPMT},
     Address= {Dept. of Ind. & Manuf. Syst. Eng., Lehigh Univ., Bethlehem, PA, USA},
     Pages = {94-101},
     Abstract = {Stencil printing continues to be the dominant method of solder deposition in high volume surface mount assembly. Control of the amount of solder paste deposited is critical for fine pitch and ultra-fine pitch SMT assembly. The process is still not well understood as indicated by the fact that industry reports 52-71% of SMT defects are related to the solder paste stencil printing process. The purpose of this paper is to identify the critical variables that influence the deposited solder paste volume, area, and height. An experiment was conducted to investigate the effects of relevant process parameters on the amount of solder paste deposited for BGAs and QFPs of 5 different pitches, ranging from 0.76 mm (30 mil) to 0.3 mm (12 mil). The effects of aperture size and shape, board finish, stencil thickness, solder type, and print speed were examined. The deposited solder paste was measured by an in-line fully automatic laser-based 3D triangulation solder paste inspection system. Analysis of variance (ANOVA) shows that aperture size and stencil thickness are the two most critical variables. A linear relationship between transfer ratio (defined as the ratio of deposited paste volume to stencil aperture volume) and area ratio (defined as the ratio of the area of the aperture opening to the area of the aperture wall) is proposed. Analysis indicates that proper stencil thickness selection is the key to controlling the amount of solder paste deposited and that the selection of maximum stencil thickness should be based on the area ratio. The experimental results are shown to be consistent with a theoretical model, which is also described.},
     Keywords = {integrated circuit packaging, ball grid arrays, soldering, fine-pitch technology, surface mount technology, assembling, automatic optical inspection, integrated circuit interconnections, solder paste stencil printing, micro-BGA, fine pitch QFP, solder d},
     Year = {1999} }

• O. Stan and E. W. Kamen, "New block recursive MLP training algorithms using the Levenberg-Marquardt algorithm," in Neural Networks, 1999. IJCNN '99. International Joint Conference on, Sch. of Electr. & Comput. Eng., Georgia Inst. of Technol., Atlanta, GA, USA, 1999, pp. 1672-1677.
  [Bibtex]

  @inproceedings{
  Sta+Kam99,
     Author = {Stan, O. and Kamen, Edward W.},
     Title = {New block recursive MLP training algorithms using the Levenberg-Marquardt algorithm},
     BookTitle = {Neural Networks, 1999. IJCNN '99. International Joint Conference on},
     Address= {Sch. of Electr. & Comput. Eng., Georgia Inst. of Technol., Atlanta, GA, USA},
     Volume = {3},
     Pages = {1672-1677},
     Abstract = {A block formulation of the Levenberg-Marquardt algorithm to train feedforward MLPs is designed to track time-varying nonlinear functions. The resulting algorithm is called the block Levenberg-Marquardt algorithm. There are two varieties of the algorithm: the overlapping and the non-overlapping block Levenberg-Marquardt. The two algorithms are developed in terms of a block presentation of the input/output training set. The tracking problem can be viewed as one of solving a sequence of nonlinear identification problems. With the persistent excitation and slowly-varying system conditions satisfied, the Levenberg-Marquardt algorithm can be shown to have a uniform rate of convergence over the entire sequence of problems. The block Levenberg-Marquardt algorithms are tested on a nonlinear time-varying function tracking problem. The algorithms show performance that is superior to the performance of existing algorithms like the global extended Kalman filter algorithm with state noise in the system equations.},
     Keywords = {multilayer perceptrons, learning (artificial intelligence), nonlinear functions, tracking, convergence, Levenberg-Marquardt algorithm, time-varying nonlinear functions, block recursive learning, multilayer perceptrons, tracking, convergence},
     Year = {1999} }

1998

• D. Burr, "Printing Guidelines for BGA and CSP Assemblies," in Surface Mount International, 1998, pp. 417-424.
  [Bibtex]

  @inproceedings{
  Burr98,
     Author = {Burr, Donald},
     Title = {Printing Guidelines for BGA and CSP Assemblies},
     BookTitle = {Surface Mount International},
     Pages = {417-424},
     Abstract = {BACKGROUNDER -- Solder Paste Deposition Inspection:The First Step to Quality
  Why is Solder Paste Application So Important?Table: AOI  Measurement  - Possible Cause  - Action Considering that repair after solder-paste deposition is at least 10 times less expensive than after reflow, and atleast 50 times less than after in-circuit test, process engineers realize that the time has come for post-print solder-paste inspection.},
        Year = {1998} }

• N. N. Ekere, D. He, and M. H. A. Riedlin, "The viscoelastic characteristics of solder paste under high frequency oscillatory shear," in Electronics Manufacturing Technology Symposium, 1998. Twenty-Third IEEE/CPMT, Dept. of Aeronaut. Mech. & Manuf. Eng., Salford Univ., UK, 1998, pp. 373-376.
  [Bibtex]

  @inproceedings{
  Ekere98,
     Author = {Ekere, N. N. and He, D. and Riedlin, M.H.A.},
     Title = {The viscoelastic characteristics of solder paste under high frequency oscillatory shear},
     BookTitle = {Electronics Manufacturing Technology Symposium, 1998. Twenty-Third IEEE/CPMT},
     Address= {Dept. of Aeronaut. Mech. & Manuf. Eng., Salford Univ., UK},
     Pages = {373-376},
     Abstract = {In this paper, the viscoelastic characteristics of solder paste under high frequency (50 Hz-90 Hz) oscillatory shear were investigated by oscillatory shear measurements. Measurement results show that the dynamic viscosity of solder paste decreases as the oscillatory frequency increases, which means the oscillation can help to enhance solder paste fluidity. This proves that in a stencil printing process, the application of a vibrating squeegee can help the paste to fill the apertures. Within the oscillatory frequency range, results show that the solder paste loss modulus is much higher than its storage modulus. With shear stress of 500 Pa, the phase shift (/spl zeta/) is around 700, and with shear stress of 1000 Pa it is around 750, so the behaviour of solder paste under oscillatory shear is more like a viscous liquid than a solid. The storage modulus with low shear stress, corresponding to small shear strain, is higher than that with high shear stress, corresponding to large shear strain. This means that the solder paste storage modulus is not only dependent on frequency but is also dependent on shear strain. High oscillatory shear strain can reduce the solder paste deformation recoverability. It suggests that when passing over the apertures, the vibrating squeegee can help to obtain uniform and high solder particle packing density inside the apertures.},
     Keywords = {elastic moduli, soldering, assembling, printed circuit manufacture, stress analysis, viscosity, viscoelasticity, shear deformation, vibrations, surface mount technology, viscoelastic characteristics, solder paste, high frequency oscillatory shear, oscil},
     Year = {1998} }

• C. Jia, "Stochastic single machine scheduling with a general objective function," in Decision and Control, 1998. Proceedings of the 37th IEEE Conference on, 1998, pp. 2166-2170.
  [Bibtex]

  @inproceedings{
  Jia98,
     Author = {Jia, Chunfu},
     Title = {Stochastic single machine scheduling with a general objective function},
     BookTitle = {Decision and Control, 1998. Proceedings of the 37th IEEE Conference on},
     Volume = {2},
     Pages = {2166-2170},
     Abstract = {We consider the problem of scheduling n jobs on a single machine which is subject to stochastic breakdowns to minimize the expectation of the linear combination of three functions of job completion times: (i) the weighted sum of the squares, (ii) the square of the weighted mean and (iii) the weighted mean. Many regular and irregular objective functions are the particular cases of this general objective function. The deterministic equivalent objective function of this problem is derived when the counting process N(t) describing the number of the machine breakdowns is a generalized Poisson process. For the two cases: (a) the processing times of the jobs are equal and (b) the weights of the jobs are proportional to their processing times, several properties of the optimal sequences of the problem are developed},
     Keywords = {production control, stochastic processes},
     Year = {1998} }

• L. Li, S. Wiegele, P. Thompson, and R. Lee, "Stencil printing process development for low cost flip chip interconnect," in Electronic Components and Technology Conference, 1998. 48th IEEE, Semicond. Products Sector, Motorola Inc., Tempe, AZ, USA, 1998, pp. 421-426.
  [Bibtex]

  @inproceedings{
  Li98,
     Author = {Li, Li and Wiegele, S. and Thompson, P. and Lee, R.},
     Title = {Stencil printing process development for low cost flip chip interconnect},
     BookTitle = {Electronic Components and Technology Conference, 1998. 48th IEEE},
     Address= {Semicond. Products Sector, Motorola Inc., Tempe, AZ, USA},
     Pages = {421-426},
     Abstract = {Traditional methods for forming flip chip interconnection include evaporation and electroplating. While both options have proven performance records, their costs are too high for many of today's cost-sensitive applications. Motorola AISL (Advanced Interconnect Systems Laboratory) developed a low cost flip chip interconnect alternative, which electroless plates Ni/Au as the under bump metallurgy (UBM), deposits solder paste using a stencil or photoresist mask, and reflows the paste to form solder bumps. This paper is focused on the stencil print process development for wafer solder bumping. Solder paste selection, characterization, stencil design, and process parameter optimization are critical factors for successful fine pitch stencil printing. Fine mesh eutectic solder pastes (Type 5: -500+635, Type 6: -635) with different flux vehicles were evaluated for printability and reflow studies. Paste and flux modification and co-development with vendors for this specific application were conducted. An optimal paste, with specified viscosity and desired print, reflow and cleaning behavior, was developed. Wafer stencil design rules were established to deposit the right amount of solder paste to form the required reflowed solder bump height without bridges between pads. Printing and reflow design of experiments were performed to establish the baseline and optimum process parameters. Solder bump characterizations, including bump height and uniformity, composition, shear force, and SEM bump profile and cross section microstructure analysis, were conducted. Development and characterization results are presented.},
     Keywords = {printing, flip-chip devices, integrated circuit interconnections, reflow soldering, fine-pitch technology, surface cleaning, integrated circuit packaging, stencil printing process development, flip chip interconnect, Motorola AISL, Advanced Interconnect},
     Year = {1998} }

• A. Lotfi and M. Howarth, "An intelligent closed-loop control of solder paste stencil printing," in Electronics Manufacturing Technology Symposium, 1998. Twenty-Third IEEE/CPMT, Dept. of Mech. & Manuf. Eng., Nottingham Trent Univ., UK, 1998, pp. 87-91.
  [Bibtex]

  @inproceedings{
  Lotfi98,
     Author = {Lotfi, A. and Howarth, M.},
     Title = {An intelligent closed-loop control of solder paste stencil printing},
     BookTitle = {Electronics Manufacturing Technology Symposium, 1998. Twenty-Third IEEE/CPMT},
     Address= {Dept. of Mech. & Manuf. Eng., Nottingham Trent Univ., UK},
     Pages = {87-91},
     Abstract = {We have introduced an automated learning closed-loop control scheme for the solder paste stencil printing stage of SMT. It reduces the number of board failures due to inadequate regulation of the printing parameters. Learning is used to compensate for a lack of a priori design information by exploiting empirical information that is gained experimentally.},
     Keywords = {closed loop systems, process control, intelligent control, learning (artificial intelligence), soldering, surface mount technology, printed circuit manufacture, assembling, circuit reliability, intelligent closed-loop control, solder paste stencil print},
     Year = {1998} }

• T. A. Nguty, M. H. A. Riedlin, and N. N. Ekere, "Evaluation of process parameters for flip chip stencil printing," in Electronics Manufacturing Technology Symposium, 1998. Twenty-Third IEEE/CPMT, Dept. of Aeronaut. Mech. & Manuf. Eng., Salford Univ., UK, 1998, pp. 206-216.
  [Bibtex]

  @inproceedings{
  Nguty98,
     Author = {Nguty, T. A. and Riedlin, M. H. A. and Ekere, N. N.},
     Title = {Evaluation of process parameters for flip chip stencil printing},
     BookTitle = {Electronics Manufacturing Technology Symposium, 1998. Twenty-Third IEEE/CPMT},
     Address= {Dept. of Aeronaut. Mech. & Manuf. Eng., Salford Univ., UK},
     Pages = {206-216},
     Abstract = {There is a great deal of interest in solder paste printing in flip chip assembly. The advantages of this process are low cost and high throughput. However, to meet this challenge there is a need to evaluate the stencil printing process parameters. These parameters can be divided into printer parameters, stencil parameters, environmental parameters, and solder paste parameters. The parameters evaluated in this paper are the stencil and solder paste parameters. Several different stencils with various shapes, sizes and pitches were investigated; these dimensions determine the type of solder paste that can be printed. In this paper, we address the need for new paste formulation characterisation, and present a procedure for evaluating solder pastes developed for flip chip application via the stencil printing process. Rheological measurements are used to correlate solder paste behaviour to stencil printing process performance. Fundamental procedures used in industry for characterisation in shear conditions include rheograms (flow curves), thixotropic index and single point viscosity measurements. We show how viscosity changes can be quantified. Viscosity measurements were carried out on a controlled stress-strain Reologica StressTech rheometer with parallel plate geometry. We also report on solder paste creep/recovery properties and how they can be correlated to slump. To correlate solder paste rheological properties to stencil printing behaviour, several printing tests were carried out. These tests can help define the process window for any given flip chip application, in terms of limiting pitch size, pad size and minimum pad gap.},
     Keywords = {flip-chip devices, microassembling, soldering, rheology, viscosity, particle size, integrated circuit packaging, integrated circuit testing, design of experiments, creep, recovery, flip chip stencil printing process parameters, solder paste printing, fl},
     Year = {1998} }

• E. Sahinci and E. W. Kamen, "Detection of discrete faults in electronics assembly," in Control Applications, 1998. Proceedings of the 1998 IEEE International Conference on, Sch. of Electr. & Comput. Eng., Georgia Inst. of Technol., Atlanta, GA, USA, 1998, pp. 776-780.
  [Bibtex]

  @inproceedings{
  Sahinci98,
     Author = {Sahinci, E. and Kamen, Edward W.},
     Title = {Detection of discrete faults in electronics assembly},
     BookTitle = {Control Applications, 1998. Proceedings of the 1998 IEEE International Conference on},
     Address= {Sch. of Electr. & Comput. Eng., Georgia Inst. of Technol., Atlanta, GA, USA},
     Volume = {2},
     Pages = {776-780},
     Abstract = {We present an optimal estimation approach to detecting discrete faults in electronics assembly. The algorithm utilizes a nonlinear fault model whose state is a vector containing the probabilities of the process in the normal operating state and all possible fault states at a given time. The model describes the time behaviour of the process state and how the process state is related to the process measurements. An extended Kalman filter is used to obtain an estimate of the process state. This approach is applied to an electronics assembly line located at the Center for Board Assembly Research at Georgia Tech.},
     Keywords = {assembling, production control, fault diagnosis, state estimation, radial basis function networks, hidden Markov models, probability, Kalman filters, electronics assembly, optimal estimation, discrete fault detection, nonlinear fault model, probability,},
     Year = {1998} }

1997

• D. Burr, "Solder paste inspection: process control for defect reduction," in Test Conference, 1997. Proceedings., International, CyberOpt. Corp., USA, 1997, pp. 1036.
  [Bibtex]

  @inproceedings{
  Burr97,
     Author = {Burr, D.},
     Title = {Solder paste inspection: process control for defect reduction},
     BookTitle = {Test Conference, 1997. Proceedings., International},
     Address= {CyberOpt. Corp., USA},
     Pages = {1036},
     Abstract = {As quality control for SMT electronics assembly shifts from final inspection/final test to production line monitoring and process control, there is a greater emphasis on post-print solder paste inspection as a means of reducing the number of defects appearing on finished boards. Automatic post-print inspection systems provide operators with the continuous information required to implement in-line continuous process control, identifying out-of-spec trends so that printing problems can be corrected before defective boards are produced.The later a defect is caught, the more expensive it is to repair, so catching a defect early in the process is inherently cheaper. Correcting a defect after reflow produces a more brittle joint and increases risk of fieldfailure. Therefore, finding a defect before reflow is an opportunity to save money and increase reliability. A defect found early in the process should not be viewed as a penalty, but rather an opportunity for savings The “1OX Rule” is a commonly-accepted quick formula for estimating the cost of a defect or failure at each stage in the process:The later a defect is caught, the more expensive it isPrint Failure      $    .50After Reflow      $   5.00In-Circuit Test   $  50.00Field Failure     $ 500.00},
     Keywords = {printed circuit manufacture, surface mount technology, printed circuit testing, process control, soldering, inspection, economics, quality control, quality control, SMT electronics assembly, defect reduction, process control, production line monitoring,},
     Year = {1997} }

• R. C. Crida, A. J. Stoddart, and J. Illingworth, "Using PCA to model shape for process control," in 3-D Digital Imaging and Modeling, 1997. Proceedings., International Conference on Recent Advances in, Centre for Vision, Speech & Signal Process., Surrey Univ., Guildford, UK, 1997, pp. 318-325.
  [Bibtex]

  @inproceedings{
  Crida97,
     Author = {Crida, R.C. and Stoddart, A.J. and Illingworth, J.},
     Title = {Using PCA to model shape for process control},
     BookTitle = {3-D Digital Imaging and Modeling, 1997. Proceedings., International Conference on Recent Advances in},
     Address= {Centre for Vision, Speech & Signal Process., Surrey Univ., Guildford, UK},
     Pages = {318-325},
     Abstract = {Before surface mount components can be placed on a circuit board, it is necessary to print solder paste onto pads. The paste is then melted to make an electrical connection (reflow). A screen printing process is used to print the solder paste onto the board. This is a complicated process with a large number of input parameters. Some of these parameters can be controlled and it is the purpose of this work to investigate control of the process based on measurement of the output shape of the printed paste. The shape is measured using a laser range scanner Principal Component Analysis (PCA) is proposed as a tool for describing solder paste shape with a small number of parameters. This paper discusses the use of PCA for shape analysis in range images as well as explaining how such a description can be incorporated into a process control loop.},
     Keywords = {process control, solid modelling, laser ranging, surface mount technology, printed circuit manufacture, process control, shape modelling, surface mount components, solder paste, electrical connection, screen printing process, laser range scanner princip},
     Year = {1997} }

• J. M. Hong, S. H. Lee, and B. H. Lee, "A hierarchical optimization method in the PCB assembly for surface mounting machines," in Industrial Electronics, 1997. ISIE '97., Proceedings of the IEEE International Symposium on, 1997, pp. 129-134.
  [Bibtex]

  @inproceedings{
  Hong97,
     Author = {Hong, Jee Min and Lee, Sung Han and Lee, Beom Hee},
     Title = {A hierarchical optimization method in the PCB assembly for surface mounting machines},
     BookTitle = {Industrial Electronics, 1997. ISIE '97., Proceedings of the IEEE International Symposium on},
     Volume = {1},
     Pages = {129-134},
     Abstract = {An hierarchical optimization method to improve efficiency of the surface mounting machines (SMM) in PCB assembly is presented. To minimize the setup time, an automatic CAD data conversion system and a database system of components are proposed, which automatically generate the data for pick and place sequences from the CADSTAR output files. For maximization of the efficiency of operations, an effective assignment algorithm for heads and nozzles and an effective arrangement algorithm for feeders are proposed. Finally, a modified effective heuristic algorithm for traveling path of heads is also proposed, which is based on the traveling salesman problem (TSP) solution technique. The proposed algorithms are tested through simulations on the assembly tasks of PCBs actually manufactured. It is shown that the proposed algorithms reduce the time required to perform the assembly sequences by up to 40 percent compared to the randomly generated sequences.},
     Keywords = {printed circuit manufacture, surface mounting machines, PCB assembly tasks, hierarchical optimization method, automatic CAD data conversion system, database system, pick and place sequences, CADSTAR output files, CAD/CAM, assignment algorithm, arrangeme},
     Year = {1997} }

• T. Jiang, S. S. Ahmad, C. Jesse, and W. Moden, "Optimizing stencil printing parameters for organic materials," in Electronics Manufacturing Technology Symposium, 1997., Twenty-First IEEE/CPMT International, Micron Technol. Inc., Boise, ID, USA, 1997, pp. 68-72.
  [Bibtex]

  @inproceedings{
  Jiang97,
     Author = {Jiang, Tongbi and Ahmad, S.S. and Jesse, C. and Moden, W.},
     Title = {Optimizing stencil printing parameters for organic materials},
     BookTitle = {Electronics Manufacturing Technology Symposium, 1997., Twenty-First IEEE/CPMT International},
     Address= {Micron Technol. Inc., Boise, ID, USA},
     Pages = {68-72},
     Abstract = {Effect of printing speed and squeegee pressure were studied for the quality of stencil printed organic material using a commercially available statistical application, JMP(R). Prints were evaluated by measuring their actual widths compared to the stencil opening in both parallel and perpendicular directions relative to squeegee travel; by print wall angle; film thickness and the extent of material bleed.JMP(R) software gave a visual picture of the effectiveness of specific machine parameters. Output of stencil printing can be predicted quantitatively for various factors. Running behavior of material (flow beyond stencil aperture boundary) was dissimilar in parallel and perpendicular directions. The optimum print quality can be achieved by varying the process parameters.},
     Keywords = {design of experiments, statistical analysis, quality control, printing, assembling, process control, printed circuit manufacture, production engineering computing, stencil printing parameters optimisation, organic materials, printing speed, squeegee pre},
     Year = {1997} }

• A. Lotfi, M. Howarth, and P. D. Thomas, "Orthogonal fuzzy model of the solder paste printing stage of surface mount technology," in Fuzzy Systems, 1997., Proceedings of the Sixth IEEE International Conference on, Dept. of Mech. & Manuf. Eng., Nottingham Trent Univ., UK, 1997, pp. 1433-1437.
  [Bibtex]

  @inproceedings{
  Lotfi97,
     Author = {Lotfi, A. and Howarth, M. and Thomas, P.D.},
     Title = {Orthogonal fuzzy model of the solder paste printing stage of surface mount technology},
     BookTitle = {Fuzzy Systems, 1997., Proceedings of the Sixth IEEE International Conference on},
     Address= {Dept. of Mech. & Manuf. Eng., Nottingham Trent Univ., UK},
     Publisher = {Application},
     Volume = {3},
     Pages = {1433-1437},
     Abstract = {The novel idea of an orthogonal fuzzy rule-based system is introduced in this paper. It is shown that in a fuzzy rule-based system where rules are orthogonal, each rule induces its maximum effect on the final decision. Using 18 orthogonal fuzzy rules, a model for the solder paste printing stage of SMT is created. The predicted percentage of deposit volume from the orthogonal and non-orthogonal models are compared with actual values. The orthogonal model shows less error in 85% of the data points.},
     Keywords = {fuzzy systems, fuzzy set theory, knowledge based systems, soldering, surface mount technology, modelling, printed circuit manufacture, orthogonal fuzzy model, solder paste printing, surface mount technology, fuzzy rule-based system, orthogonal rules, fu},
     Year = {1997} }

• S. Venkateswaran, K. Srihari, J. H. Adriance, and G. R. Westby, "A realtime process control system for solder paste stencil printing," in Electronics Manufacturing Technology Symposium, 1997., Twenty-First IEEE/CPMT International, Dept. of Syst. Sci. & Ind. Eng., State Univ. of New York, Binghamton, NY, USA, 1997, pp. 62-67.
  [Bibtex]

  @inproceedings{
  Venka97,
     Author = {Venkateswaran, S. and Srihari, K. and Adriance, J.H. and Westby, G.R.},
     Title = {A realtime process control system for solder paste stencil printing},
     BookTitle = {Electronics Manufacturing Technology Symposium, 1997., Twenty-First IEEE/CPMT International},
     Address= {Dept. of Syst. Sci. & Ind. Eng., State Univ. of New York, Binghamton, NY, USA},
     Pages = {62-67},
     Abstract = {The goal of stencil printing solder paste in surface mount printed circuit board (PCB) assembly is to apply an accurate and repeatable volume of solder paste at precise locations. The causes for a substantial proportion of the problems associated with PCB assembly can be traced back to the solder paste printing process. Control of the stencil printing process has become significantly more important over the years due to the introduction of ultra fine pitch technologies. Three important issues have to be considered in the stencil printing process to obtain a good yield. They are: (i) setting up the process in an efficient way, (ii) monitoring and controlling the process when necessary and (iii) troubleshooting the process when a defect occurs. This research focused on the design and development of software systems that could perform the above-mentioned functions. Three systems were developed in the research discussed in this paper. They are the process advisor, the intelligent control system (or the ICS) and the diagnosis system. The process advisor (the first system) provides for the setup of the stencil printing process. It helps the user to set up a new as well as an existing application. The process advisor also helps the user in estimating the volume of solder paste that would be deposited for a given stencil thickness and aperture dimensions. The ICS (the second system) was developed to control the stencil printing process in real-time. The diagnosis system (the third system) is used to troubleshoot the process, if needed.},
     Keywords = {soldering, fine-pitch technology, process control, printed circuit manufacture, real-time systems, assembling, surface mount technology, printing, monitoring, intelligent control, electronic engineering computing, realtime process control system, solder},
     Year = {1997} }

1996

• R. L. Mahajan, "Statistical neural network modeling for stencil printing," in Surface Mount International, 1996.
  [Bibtex]

  @inproceedings{
  Mahajan96,
     Author = {Mahajan, Roop L.},
     Title = {Statistical neural network modeling for stencil printing},
     BookTitle = {Surface Mount International},
     Abstract = {Stencil printing is the first and perhaps the most critical step affecting the quality and yield of a fine pitch surface mount assembly. This process must be optimized for higher assembly yield. In this paper, a statistical design of experiment (DOE)-based neural network approach is presented that serves to economically model the process. It is shown that a neural network model based on a Taguchi L27 orthogonal array does an excellent job of determining the optimum settings for minimum solder paste height variation. The effect of optimum settings on the fatigue life of solder joints is presented. Finally, techniques are proposed to transfer a neural network process model from one machine to another without the need for extensive experimental data.},
     Keywords = {defects related SPP},
     Year = {1996} }

• V. Torra, "Weighted OWA operators for synthesis of information," in Fuzzy Systems, 1996., Proceedings of the Fifth IEEE International Conference on, 1996, pp. 966-971.
  [Bibtex]

  @inproceedings{
  Torra96,
     Author = {Torra, V.},
     Title = {Weighted OWA operators for synthesis of information},
     BookTitle = {Fuzzy Systems, 1996., Proceedings of the Fifth IEEE International Conference on},
     Volume = {2},
     Pages = {966-971},
     Abstract = {One of the properties that the OWA operator satisfies is commutativity. This condition, that is not satisfied by the weighted mean, stands far equal reliability of all the information sources that supply the data. In this paper we define a new combination function, the WOWA (Weighted OWA), that combines the advantages of the OWA operator and the ones of the weighted mean},
     Keywords = {fuzzy logic, sensor fusion},
     Year = {1996} }

1995

• N. N. Ekere, S. H. Mannan, and M. A. Currie, "Solder paste printing process modelling map," in Electronic Manufacturing Technology Symposium, 1995, Proceedings of 1995 Japan International, 18th IEEE/CPMT International, Dept. of Aeronaut., Mech. & Manuf. Eng., Salford Univ., UK, 1995, pp. 137-141.
  [Bibtex]

  @inproceedings{
  Ekere95,
     Author = {Ekere, N. N. and Mannan, S.H. and Currie, M.A.},
     Title = {Solder paste printing process modelling map},
     BookTitle = {Electronic Manufacturing Technology Symposium, 1995, Proceedings of 1995 Japan International, 18th IEEE/CPMT International},
     Address= {Dept. of Aeronaut., Mech. & Manuf. Eng., Salford Univ., UK},
     Pages = {137-141},
     Abstract = {This paper presents work done in the modelling of solder paste printing, a crucial process in the reflow soldering of surface mounted electronic components. The map summarises the results of the experimental observations and theoretical models developed. In the mapping approach developed, information on the process models and modelling technique (including assumptions used), process and equipment parameters, physical sub-processes, process variables, and the process performance in terms of quality and/or defects are represented. The solder paste printing process is modelled as six different interacting sub-processes in the physical sub-processes layer, namely: pre-print paste treatment, squeegee deformation, paste roll in front of the squeegee, aperture filling, aperture emptying and paste slump. The sub-processes are linked together by the properties of the solder paste such as its flow history or theology in terms of the stress/strain history and the viscosity/time graph.},
     Keywords = {printing, solder paste stencil printing process, modelling map, reflow soldering, surface mounted electronic component, paste treatment, squeegee deformation, paste roll, aperture filling, aperture emptying, paste slump, Critical Variables},
     Year = {1995} }

• T. Okura, M. Kanai, S. Ogata, T. Takei, and H. Takakusagi, "Optimization of solder paste printability with laser inspection technique," in Electronics Manufacturing Technology Symposium, 1995. 'Manufacturing Technologies - Present and Future', Seventeenth IEEE/CPMT International, Piscataway, NJ, 1995, pp. 361-365.
  [Bibtex]

  @inproceedings{
  Okura95,
     Author = {Okura, T. and Kanai, M. and Ogata, S. and Takei, T. and Takakusagi, H.},
     Title = {Optimization of solder paste printability with laser inspection technique},
     BookTitle = {Electronics Manufacturing Technology Symposium, 1995. 'Manufacturing Technologies - Present and Future', Seventeenth IEEE/CPMT International},
     Address= {Piscataway, NJ},
     Publisher = {IEEE},
     Pages = {361-365},
     Abstract = {This paper describes an evaluation technique for solder paste printability using a solder paste inspection system and optimization of printing parameters. In a Surface Mount Assembly (SMA), it is very important to evaluate printability of solder paste precisely, because printability of solder paste directly influences reflow soldering quality. We established quite reliable inspection criteria for print quantity of solder paste. So it was found that about 90% of reflow soldering defects could be reduced by prevention of printing defects. Furthermore, since package types have been varied in a SMA, high soldering quality demands stable printability for all patterns. This paper presents a printability evaluation adopted procedure based on the quality engineering (TAGUCHI method). As a result of optimizing squeegee shape and printer condition by this evaluation method, printability was improved better than 20% in the production process. And optimization of solder paste and stencil added to printer condition resulted in good printability for 0.3 mm pitch patterns on the optimum condition.},
     Keywords = {reflow soldering, inspection, laser beam applications, printing, surface mount technology, solder paste printability, laser inspection technique, printing parameters, surface mount assembly, reflow soldering quality, inspection criteria, package types, },
     Year = {1995} }

1994

• H. K. Brown, L. A. Martin-Vega, W. H. Shaw, J. G. Madry, and C. L. Taylor, "Utilization of sensory feedback during SMT assembly," in Southcon/94. Conference Record, Florida Inst. of Technol., Melbourne, FL, USA, 1994, pp. 349-352.
  [Bibtex]

  @inproceedings{
  Brown94,
     Author = {Brown, H.K. and Martin-Vega, L.A. and Shaw, W.H. and Madry, J.G. and Taylor, C.L.},
     Title = {Utilization of sensory feedback during SMT assembly},
     BookTitle = {Southcon/94. Conference Record},
     Address= {Florida Inst. of Technol., Melbourne, FL, USA},
     Pages = {349-352},
     Abstract = {This paper investigates the insertion of inline sensors into Surface Mount Technology (SMT) production lines for the purpose of improving yield and reducing production costs by elevating the quality level of the product to the six sigma level. Each process step in the assembly process, starting with board loading, through deposition, part placement, and reflow will be discussed for monitoring in real time for deviations. When deviations are detected, feedback can be sent either to operators or the equipment for corrective action thus providing a closed loop SMT assembly process. A study of an SMT assembly line indicates that video surveillance at the beginning of the line can be used to inspect the quality of the Printed Wire Boards (PWBs) before additional processing actions are taken on them. Similar surveillance techniques employed in the deposition stage to ensure that only PWBs with proper solder paste deposition are allowed to continue will be discussed. The parts placement stage provides many opportunities for the placement of inline sensors ranging from inspection of the parts for defects and orientation errors before they are placed to determining their exact position on the PWB after they have been placed. The potential of similar placement monitoring techniques before and after reflow for monitoring this process are explored.},
     Keywords = {surface mount technology, assembling, circuit optimisation, integrated circuit yield, quality control, reflow soldering, closed loop systems, sensory feedback, SMT assembly, inline sensors, production costs, yield, quality level, six sigma level, proces},
     Year = {1994} }

• M. A. Currie, N. N. Ekere, S. H. Mannan, and I. Ismail, "Solder paste characteristics and their effect on fine pitch printing," in Factory 2000 - Advanced Factory Automation, Fourth International Conference on (Conf. Publ. No. 398), University Of Salford, Salford, UK, 1994, pp. 552-558.
  [Bibtex]

  @inproceedings{
  Currie94,
     Author = {Currie, M. A. and Ekere, N. N. and Mannan, S. H. and Ismail, I.},
     Title = {Solder paste characteristics and their effect on fine pitch printing},
     BookTitle = {Factory 2000 - Advanced Factory Automation, Fourth International Conference on (Conf. Publ. No. 398)},
     Address= {University Of Salford, Salford, UK},
     Pages = {552-558},
        Year = {1994} }

• D. Filev and R. R. Yager, "Learning OWA operator weights from data," in Fuzzy Systems, 1994. IEEE World Congress on Computational Intelligence., Proceedings of the Third IEEE Conference on, 1994, pp. 468-473.
  [Bibtex]

  @inproceedings{
  Filev94,
     Author = {Filev, D. and Yager, R.R.},
     Title = {Learning OWA operator weights from data},
     BookTitle = {Fuzzy Systems, 1994. IEEE World Congress on Computational Intelligence., Proceedings of the Third IEEE Conference on},
     Volume = {1},
     Pages = {468-473},
     Abstract = {We investigate the issue of obtaining the weights associated with the ordered weighted operator (OWA) aggregation in the situation when we have observed data on the arguments and the aggregated value. This provides a procedure to learn the decision maker's level of confidence from previous decisions. We next introduce a family of OWA operators called exponential OWA operators},
     Keywords = {aggregated value
  confidence level
  constrained optimisation
  decision making
  decision theory
  fuzzy set theory
  learning
  learning (artificial intelligence)
  operator weights
  optimisation
  ordered weighted operator},
     Year = {1994} }

• S. F. Kirmani and G. D. Foote, "Solder paste critical parameter measurement in surface mount boards using model-based vision processing techniques," in SME Technical Paper (Series) MS, Minneapolis, MN, USA, 1994, pp. 1-8.
  [Bibtex]

  @inproceedings{
  Kirmani94,
     Author = {Kirmani, Shahzad F.  and Foote, Gerald D.},
     Title = {Solder paste critical parameter measurement in surface mount boards using model-based vision processing techniques},
     BookTitle = {SME Technical Paper (Series) MS},
     Address= {Minneapolis, MN, USA},
     Pages = {1-8},
     Abstract = {As the functionality and complexity of printed circuit boards increases, manufacturers are moving to surface mount technology with component lead pitches as low as 0.2 mm. Solder paste is deposited at widths of 100 microns and more, and has to be monitored for irregularities like bridges or insufficient solder deposition. Dimensional and volumetric measurements are also of interest and aid in providing a closed-loop control process for solder printing. Image processing techniques can be employed to measure critical solder paste parameters like height, volume, coverage and footprint area. This paper will discuss model-based image processing techniques for extracting solder paste information from images containing several solder pads. Methods of finding and processing solder pads, and other features of interest, will be discussed and sample images will be presented in the paper. (Author abstract) 7 Refs.},
     Keywords = {Computer vision,  Surface mount technology,  Printed circuit manufacture,  Soldering alloys,  Inspection,  Process control,  Image processing,  Optics,  Integrated circuits,  Monitoring,  Size determination,  Closed loop control process,  Model based im},
     Year = {1994} }

1993

• D. Braunstein and H. Asada, "Real Time Process Monitoring of Solder Paste Stencil Printing," Motorola Inc. and MPM Corp., Progress Report: August-December, , 1993.
  [Bibtex]

  @techreport{
  Braunstein93,
     Author = {Braunstein, Daniel and Asada, Haruhiko},
     Title = {Real Time Process Monitoring of Solder Paste Stencil Printing},
     Institution = {Motorola Inc. and MPM Corp.},
     Number = {August-December},
     Type = {Progress Report:},
     Keywords = {SPP Model Identification, equations, RLS LS, exponential forgetting},
     Year = {1993} }

• K. Feldmann and J. Sturm, "Yield improvement in SMT production by integrated process monitoring and testing," in Electronic Manufacturing Technology Symposium, 1993, Fifteenth IEEE/CHMT International, Inst. for Manuf. Autom. & Production Syst., Erlangen Univ., Germany, 1993, pp. 229-234.
  [Bibtex]

  @inproceedings{
  Feldmann93,
     Author = {Feldmann, K. and Sturm, J.},
     Title = {Yield improvement in SMT production by integrated process monitoring and testing},
     BookTitle = {Electronic Manufacturing Technology Symposium, 1993, Fifteenth IEEE/CHMT International},
     Address= {Inst. for Manuf. Autom. & Production Syst., Erlangen Univ., Germany},
     Pages = {229-234},
     Abstract = {Almost 30-50% of fabrication costs in electronics production are caused by testing and repair operations. A strategy for yield improvement requires strong dedication to causal relationships and specification of process and product quality. Critical process parameters and their causal effects on product quality must be analyzed. Once critical process parameters are addressed, control strategies can be developed to assure high first pass yields. An overview of ongoing work that has been performed to establish a surface mount technology (SMT) production line with closed loop quality control via direct process monitoring capabilities and integrated inspection operations is given. For different inspection tasks, 3-D laser and X-ray inspection is applied. Process parameter as well as test and inspection results are gathered online. Thus, interrelationships between process steps beginning from material inspection, solder paste application, components placement, reflow soldering and final inspection can be analyzed.},
     Keywords = {surface mount technology, printed circuit manufacture, process control, automatic optical inspection, closed loop systems, reflow soldering, printed circuit testing, economics, quality control, design of experiments, statistical process control, CAD/CAM},
     Year = {1993} }

• S. R. Kulkarni and C. Horn, "Convergence of the Robbins-Monro algorithm under arbitrary disturbances," in Decision and Control, 1993., Proceedings of the 32nd IEEE Conference on, Dept. of Electr. Eng., Princeton Univ., NJ, USA, 1993, pp. 537-538.
  [Bibtex]

  @inproceedings{
  Kulkarni93,
     Author = {Kulkarni, S.R. and Horn, C.},
     Title = {Convergence of the Robbins-Monro algorithm under arbitrary disturbances},
     BookTitle = {Decision and Control, 1993., Proceedings of the 32nd IEEE Conference on},
     Address= {Dept. of Electr. Eng., Princeton Univ., NJ, USA},
     Volume = {1},
     Pages = {537-538},
     Abstract = {The Robbins-Monro algorithm under arbitrary deterministic disturbances is studied and necessary and sufficient conditions on the noise sequence are obtained for convergence of the algorithm. We introduce a notion of persistently disturbing noise sequences, and show that it characterises convergence of the algorithm under each fixed noise sequence. The results obtained are stronger than previous conditions and the proof techniques are simpler, involving only basic notions of convergence.},
     Keywords = {convergence of numerical methods, identification, noise, function approximation, stochastic approximation, Robbins-Monro algorithm, arbitrary deterministic disturbances, sufficient conditions, necessary conditions, noise sequence, convergence},
     Year = {1993} }

1991

• S. Fujiuchi, "Fundamental study on solder paste for fine pitch soldering," in Electronics Manufacturing Technology Symposium, 1991., Eleventh IEEE/CHMT International, IBM Japan, Shiga, Japan, 1991, pp. 163-165.
  [Bibtex]

  @inproceedings{
  Fuji91,
     Author = {Fujiuchi, S.},
     Title = {Fundamental study on solder paste for fine pitch soldering},
     BookTitle = {Electronics Manufacturing Technology Symposium, 1991., Eleventh IEEE/CHMT International},
     Address= {IBM Japan, Shiga, Japan},
     Pages = {163-165},
     Abstract = {For fine pitch QFP soldering, solder amount and locations applied on pad must be precisely controlled. Solder paste has been printed individually on each pad for most card assemblies, but this printing method is confronted with process control difficulties. As a solution of these potential problems, a new type of printing method, in which many pads are covered with one bar (or belt) shaped paste pattern, is under evaluation. The author describes some experimental studies of solder paste suitable for this printing method and the feasibility for fine pitch soldering. An experiment on three kinds of solder pastes showed that low-melting solder containing SN-Pb-Bi eutectic solder was suitable for this method because of its good solderability and even solder amount on each pad after reflow soldering. These advantages seemed to result from partial melting of low-melting solder and bonding between solder particles during preheating in the reflow process. This was confirmed by scanning electron microscopy. With this printing method, soldering of fine pitch (0.5/0.3 mm) leads was successfully demonstrated.},
     Keywords = {solder paste, fine pitch soldering, QFP soldering, printing method, low-melting solder, partial melting, preheating, reflow process, scanning electron microscopy, SnPbBi solder, packaging, printed circuit manufacture, soldering, surface mount technology},
     Year = {1991} }

• S. Kumar, "Survey of various statistical process control methods," in Electronics Manufacturing Technology Symposium, 1991., Eleventh IEEE/CHMT International, Motorola Inc., Austin, TX, USA, 1991, pp. 387-390.
  [Bibtex]

  @inproceedings{
  Kumar91,
     Author = {Kumar, S.},
     Title = {Survey of various statistical process control methods},
     BookTitle = {Electronics Manufacturing Technology Symposium, 1991., Eleventh IEEE/CHMT International},
     Address= {Motorola Inc., Austin, TX, USA},
     Pages = {387-390},
     Abstract = {The author discusses the various statistical process control (SPC) techniques currently being used by different industries. He also introduces a few nonconventional SPC charts which were published recently and could be used for better results. He also describes the Motorola method of performing characterization of processes.},
     Keywords = {statistical process control methods, SPC charts, Motorola method, characterization, statistical process control, DOE},
     Year = {1991} }

• Y. Maeda and Y. Kanata, "An algorithm for a least-square approximation problem of unknown systems," in Industrial Electronics, Control and Instrumentation, 1991. Proceedings. IECON '91., 1991 International Conference on, Dept. of Electr. Eng., Kansai Univ., Osaka, Japan, 1991, pp. 1881-1886.
  [Bibtex]

  @inproceedings{
  Maeda91,
     Author = {Maeda, Y. and Kanata, Y.},
     Title = {An algorithm for a least-square approximation problem of unknown systems},
     BookTitle = {Industrial Electronics, Control and Instrumentation, 1991. Proceedings. IECON '91., 1991 International Conference on},
     Address= {Dept. of Electr. Eng., Kansai Univ., Osaka, Japan},
     Volume = {3},
     Pages = {1881-1886},
     Abstract = {The authors consider a problem of finding a least-squares approximation parameter that minimizes the output error of unknown systems. When the dimension of the output is equal to the dimension of the input, one can apply the stochastic approximation algorithm. On the other hand, if the dimension of the output is greater than the dimension of the input, one cannot use stochastic approximation. The authors propose an algorithm that is applicable to this problem. This algorithm is an extension of the Robbins-Monro stochastic approximation procedure. A convergence theorem for this proposed procedure is demonstrated.},
     Keywords = {unknown systems, least-squares approximation, output error, Robbins-Monro stochastic approximation, convergence theorem, convergence of numerical methods, least squares approximations},
     Year = {1991} }

• M. A. Zohdy and B. Adamczyk, "Least squares approach to constrained global optimization," in Decision and Control, 1991., Proceedings of the 30th IEEE Conference on, Dept. of Electr. & Syst. Eng., Oakland Univ., Rochester, MI, USA, 1991, pp. 945-946.
  [Bibtex]

  @inproceedings{
  Zoh+Bog91,
     Author = {Zohdy, M.A. and Adamczyk, B.},
     Title = {Least squares approach to constrained global optimization},
     BookTitle = {Decision and Control, 1991., Proceedings of the 30th IEEE Conference on},
     Address= {Dept. of Electr. & Syst. Eng., Oakland Univ., Rochester, MI, USA},
     Pages = {945-946},
     Abstract = {The authors present a stochastic least squares approach to the problem of determining the global extremum of multivariable nonlinear objective functions subject to constraints. The approximate value of the global extremum is found by using a special transformation followed by least squares estimation. The corresponding optimal coordinates are derived by a neural network.},
     Keywords = {least squares approximations, neural nets, optimisation, constrained global optimization, stochastic least squares approach, global extremum, multivariable nonlinear objective functions, neural network},
     Year = {1991} }

1990

• H. R. Berenji, Y. -Y. Chen, and R. R. Yager, "Using new aggregation operators in rule-based intelligent control," in Decision and Control, 1990., Proceedings of the 29th IEEE Conference on, 1990, pp. 2198-2203.
  [Bibtex]

  @inproceedings{
  Berenji90,
     Author = {Berenji, H.R. and Chen, Y.-Y. and Yager, R.R.},
     Title = {Using new aggregation operators in rule-based intelligent control},
     BookTitle = {Decision and Control, 1990., Proceedings of the 29th IEEE Conference on},
     Volume = {4},
     Pages = {2198-2203},
     Abstract = {A new aggregation operator is applied in the design of an approximate reasoning-based controller. The ordered weighted averaging (OWA) operator has the property of lying between the AND function and the OR function used in previous fuzzy set reasoning systems. It is shown that, by applying OWA operators, more generalized types of control rules, which may include linguistic quantifiers such as MANY and MOST, can be developed. The new aggregation operators, as tested in a cart-pole balancing control problem, illustrate improved performance when compared with existing fuzzy control aggregation schemes},
     Keywords = {control system synthesis, fuzzy logic, inference mechanisms, knowledge based systems},
     Year = {1990} }

• M. O'Hagan, "A Fuzzy Neuron Based on Maximum Entropy Ordered Weighted Averaging," in Signals, Systems and Computers, 1990. 1990 Conference Record Twenty-Fourth Asilomar Conference on, 1990, pp. 618.
  [Bibtex]

  @inproceedings{
  OHagan90,
     Author = {O'Hagan, M.},
     Title = {A Fuzzy Neuron Based on Maximum Entropy Ordered Weighted Averaging},
     BookTitle = {Signals, Systems and Computers, 1990. 1990 Conference Record Twenty-Fourth Asilomar Conference on},
     Volume = {2},
     Pages = {618},
        Year = {1990} }

1988

• M. O'Hagan, "Aggregating template or rule antecedents in real-time expert systems with fuzzy set Logic," in Signals, Systems and Computers, 1988. Twenty-Second Asilomar Conference on, 1988, pp. 681-689.
  [Bibtex]

  @inproceedings{
  OHagan88,
     Author = {O'Hagan, M.},
     Title = {Aggregating template or rule antecedents in real-time expert systems with fuzzy set Logic},
     BookTitle = {Signals, Systems and Computers, 1988. Twenty-Second Asilomar Conference on},
     Volume = {2},
     Pages = {681-689},
        Year = {1988} }

2010

• J. A. Alcazar and L. G. Barajas, Dexterous Robotic Hand Grasping Method for Automotive Parts, 2010.
  [Bibtex]

  @misc{
  Alcazar10,
     Author = {Alcazar, Javier A.  and Barajas, Leandro G.},
     Title = {Dexterous Robotic Hand Grasping Method for Automotive Parts},
     Pages = {6},
     Month = {August 21-24},
     Abstract = {Kitting processes are fundamental enablers in flexible manufacturing environments derived from minomi principles. A typical kitting application will sort parts without packaging into a tray or “kit” and then place them near the point of assembly for easy reach from assembly workers. In order to prepare, sort and sequence the kits, it is necessary to have adaptable robots and automation able to pick-and-place a variety of parts at line production rate. Nowadays, this translates into being able to assemble kits of about 10 different parts within 60 seconds. This presents a fundamental challenge for kitting automation given that it requires several parts of different shapes, which are presented in either random or semi-structured fashion, to be grasped and placed into the kit. A methodology that builds a set of hand grasping shapes based on the capabilities and mechanical constraints of the robotic hand is proposed herein. Pre-grasp shapes for a Barrett Hand are studied and defined using finger spread and flexion. In addition, a robust and efficient vision algorithm is used to servo the robot in the pick-and-place task of 9 different vehicle door parts. Finally, experimental results are presented which rank the ability of the robotic hand to grasp both flexible and rigid parts.},
           Year = {2010} }

• L. G. Barajas, S. R. Biller, F. Gu, and C. Yuan, Virtual Launch & Validation of Manufacturing Automation Controls, 2010.
  [Bibtex]

  @misc{
  VLVofMAC,
     Author = {Barajas, Leandro G. and Biller, Stephan R.  and Gu, Fangming and Yuan, Chengyin},
     Title = {Virtual Launch & Validation of Manufacturing Automation Controls},
     Pages = {8},
     Month = {August 21-24},
     Abstract = {In the automotive industry an integrated set of manufacturing engineering activities lie on the critical path of the Vehicle Development Process (VDP). After the product integration design is finalized, the Production Tool Design and Build (PTDB) processes becomes the VDP critical path. It is in the PTDB where virtual launch and validation activities are performed. To increase throughput of the engineering factory and reduce product development cost, we propose a research initiative in the area of Virtual Launch & Validation (VLV) of Manufacturing Automation Controls (MAC). VLV of MAC is composed of a series of PTDB launch and pre-launch activities in a virtual environment using physical, emulated, and simulated hardware and software.  It encompasses the testing of MAC, its interactions with MAC analysis tools, and its interfaces with IT systems and related business processes. VLV of MAC also supports the development and testing of new IT systems and business processes as well as its hardware and software plant floor systems constituents. The mission of VLV of MAC is to minimize the duration of the launch and ramp up activities by reducing the implementation and execution times for plant floor systems and equipments via a priori system emulation, validation and testing. Based on our findings, we conclude that the most pressing needs for VLV of MAC lie in the thrust areas of object-oriented standard logic, auto-generation of virtual models, logic validation and system wide emulation. In summary, we conservatively estimate that a successful VLV of MAC implementation will reduce the VDP duration by several weeks hence providing automakers with savings in the order of hundreds of millions of dollars per year. },
           Year = {2010} }

2008

• L. G. Barajas and N. Srinivasa, Real-Time Diagnostics, Prognostics & Health Management for Large-Scale Manufacturing Systems, 2008.
  [Bibtex]

  @misc{
  BarajasMSEC08,
     Author = {Barajas, Leandro G. and Srinivasa, Narayan},
     Title = {Real-Time Diagnostics, Prognostics & Health Management for Large-Scale Manufacturing Systems},
     Month = {October 7-10},
     Abstract = {Traditional technologies emphasize either experience or model-based approaches to the Diagnostics, Prognostics & Health Management (DPHM) problem. However, most of these methodologies often apply only to the narrow type of machines that they were developed for, and only support strategic level assessments as opposed to real-time tactical decisions. By enabling widespread integration of diagnostics and prognostics into our manufacturing business processes, we have reduced spacio-temporal uncertainties associated with future states and system performance and therefore enabled more informed and effective decisions on manufacturing activities. For large-scale systems, the usual approach is to aggregate multidimensional data into a single-dimensional stream. These methods are generally adequate to extract key performance indicators. However, they only point to observable effects of a failure and not to their root causes. An integrated framework for DPHM requires the availability of bidirectional cause-effect relationships that enable system-wide health management rather than just predicting what its future state would be. This paper summarizes best practices, benchmarks, and lessons learned from the design, development, deployment, and execution of DPHM systems into real-life applications in the automotive industry. },
     Keywords = {DPHM},
        Year = {2008} }

• N. Srinivasa, Q. Jiang, and L. G. Barajas, High-Impact Event Prediction by Temporal Data Mining through Genetic Algorithms, 2008.
  [Bibtex]

  @misc{
  BarajasICNC08,
     Author = {Srinivasa, Narayan and Jiang, Qin and Barajas, Leandro G.},
     Title = {High-Impact Event Prediction by Temporal Data Mining through Genetic Algorithms},
     Volume = {1},
     Pages = {614 - 620},
     Month = {October 18-20 },
     Abstract = {This paper describes a genetic algorithm based approach to detect and predict high-impact events. While, these events occur infrequently, they are quite costly, meaning that they have a high-impact on the system key performance indicators. This approach is based on mining for these events and subsequences that are predictive of these high-impact events from historical data and then classifying these predictive patterns. The resulting mined patterns are subsequently used to make future prediction of occurrences. The approach uses a genetic algorithm (GA) for estimating the parameters for the mining process and for the prediction. This makes our approach robust as the parameters are optimized for best accuracy in classification. This approach was tested on high-impact events that occur in automotive manufacturing lines and it was found to be robust, highly accurate and with low probability of false alarms for prediction of future occurrences of such events.},
     Keywords = {DPHM, Event Detection, Genetic Algorithms, Manufacturing, Prediction, Prognostics   },
     ISBN = {978-0-7695-3304-9},
     Year = {2008} }

2005

• Mathworks, Optimization Toolbox User's Guide: for use with MATLAB, version 3.0.3, Natick, MA: MathWorks, 2005.
  [Bibtex]

  @book{
  MLOptimTool05,
     Author = {Mathworks},
     Title = {Optimization Toolbox User's Guide: for use with MATLAB, version 3.0.3},
     Publisher = {MathWorks},
     Address = {Natick, MA},
     url = {http://www.mathworks.com/access/helpdesk/help/toolbox/optim/},
        Year = {2005} }

2003

• L. G. Barajas, "Process control in high-noise environments using a limited number of measurements," Ph.D. dissertation PhD Thesis, 2003.
  [Bibtex]

  @phdthesis{
  Barajas03PhD,
     Author = {Barajas, Leandro G.},
     Title = {Process control in high-noise environments using a limited number of measurements},
     School = {Georgia Institute of Technology},
     Type = {Ph.D. dissertation},
     Abstract = {The topic of this dissertation is the derivation, development, and evaluation of novel hybrid algorithms for process control that use a limited number of measurements and that are suitable to operate in the presence of large amounts of process noise. As an initial step, affine and neural network statistical process models are developed in order to simulate the steady-state system behavior. Such models are vitally important in the evaluation, testing, and improvement of all other process controllers referred to in this work. Afterwards, fuzzy logic controller rules are assimilated into a mathematical characterization of a model that includes the modes and mode transition rules that define a hybrid hierarchical process control. The main processing entity in such framework is a closed-loop control algorithm that performs global and then local optimizations in order to asymptotically reach minimum bias error; this is done while requiring a minimum number of iterations in order to promptly reach a desired operational window. The results of this research are applied to surface mount technology manufacturinglines yield optimization. This work achieves a practical degree of control over the solderpaste volume deposition in the Stencil Printing Process (SPP). Results show that it is possible to change the operating point of the process by modifying certain machine parameters and even compensate for the difference in height due to change in print direction.},
     Keywords = {LS, hybrid control, noise, least squares, affine estimator, conjugated gradient, process control, statistical methods, fuzzy systems, industrial noise, Electrical and computer engineering Theses 2003},
     Year = {2003} }

• L. G. Barajas, A. Kansal, A. Saxena, M. B. Egerstedt, A. Goldstein, and E. W. Kamen, "Modeling and control of SMT manufacturing lines using hybrid dynamic systems." Prague, The Czech Republic: Springer-Verlag, 2003.
  [Bibtex]

  @incollection{
  Barajas+Hybrid03,
     Author = {Barajas, Leandro G. and Kansal, A. and Saxena, A. and Egerstedt, Magnus B. and Goldstein, Alex and Kamen, Edward W.},
     Title = {Modeling and control of SMT manufacturing lines using hybrid dynamic systems},
     BookTitle = {Hybrid Systems: Computation and Control},
     Publisher = {Springer-Verlag},
     Address = {Prague, The Czech Republic},
     Abstract = {In this paper we show how hybrid control and modeling techniques can be put to work for solving a problem of industrial relevance in Surface Mount Technology (SMT) manufacturing. In particular, by closing the loop over the stencil printing process, we obtain a robust system that can recover from faulty initial settings, adapt to environmental changes and unscheduled interrupts, and remove discrepancies associated with bidirectional printing machines. Moreover, a timed Petri net argument is invoked for bounding the control effort in such a way that the throughput of the system is unaffected by the introduction of the closed-loop controller. The soundness of the approach is verified on a real SMT manufacturing line.},
        Year = {2003} }

• P. M. Lee, Information on the History of the Normal LawHTTP, 2003.
  [Bibtex]

  @misc{
  Lee03,
     Author = {Lee, Peter M.},
     Title = {Information on the History of the Normal Law},
     Publisher = {HTTP},
     Volume = {2004},
     Number = {June  4},
     url = {http://www.york.ac.uk/depts/maths/histstat/normal_history.pdf},
     Keywords = {Poisson integral},
        Year = {2003} }

• Minitab, One-Sample t-TestMiniTab, 2003.
  [Bibtex]

  @misc{
  DescStatsMinitab,
     Author = {Minitab},
     Title = {One-Sample t-Test},
     Publisher = {MiniTab},
     Volume = {2006},
     Number = {August},
     url = {http://www.minitab.com/training/t-Test.pdf},
     Keywords = {Descriptive Statistics},
        Year = {2003} }

2001

• H. J. Kushner and P. Dupuis, Numerical methods for stochastic control problems in continuous time, 2nd ed., Springer, 2001.
  [Bibtex]

  @book{
  Kushner01,
     Author = {Kushner, Harold J. and Dupuis, Paul},
     Title = {Numerical methods for stochastic control problems in continuous time},
     Publisher = {Springer},
     Edition = {2nd},
     Keywords = {Stochastic control theory, Markov processes, Numerical analysis},
     Year = {2001} }

• O. Tani, Method and apparatus for cleaning screen used in screen printing machineTani Denkikogyo Co. Ltd, 2001.
  [Bibtex]

  @misc{
  Tani01,
     Author = {Tani, Okie},
     Title = {Method and apparatus for cleaning screen used in screen printing machine},
     Publisher = {Tani Denkikogyo Co. Ltd},
     Month = {2001/05/22},
     Abstract = {The invention encompasses method and apparatus for cleaning screen used in screen printing machine for printing of solder paste onto a surface of a printed circuit board. The apparatus includes two upper and lower cleaning chambers 4, 5 which are provided on upper and lower sides of the screen 2 in opposed relationship with each other in such a manner that the two chambers are in watertight contact with the upper and lower surfaces 2A, 2B, of screen 2, respectively and are communicated with each other through openings 2C in the screen 2. The upper and lower cleaning chambers are essential components of the apparatus for carrying out cleaning the solder paste residue adhered to inner peripheral surfaces of the openings 2C in the screen 2 as well as the solder paste residue adhered to the upper and lower surfaces 2A, 2B from the screen 2. A wash liquid supply means is provided for supplying wash liquid into each of the chambers 4, 5 to wash away the past residues adhered to the screen 2.},
     Keywords = {USPAT SPP},
     ISBN = {101-424-000 101-425-000 134-093-000 134-199-000},
     Year = {2001} }

2000

• D. P. Bertsekas, Dynamic programming and optimal control, 2nd ed., Belmont, Mass.: Athena Scientific, 2000.
  [Bibtex]

  @book{
  Bertsekas00,
     Author = {Bertsekas, Dimitri P.},
     Title = {Dynamic programming and optimal control},
     Publisher = {Athena Scientific},
     Address = {Belmont, Mass.},
     Edition = {2nd},
     Series = {Athena Scientific optimization and computation series.},
     Keywords = {Dynamic programming, Control theory},
     Year = {2000} }

• H. Demuth and M. Beale, Neural Network Toolbox User's Guide: for use with MATLAB, version 4, Natick, MA: MathWorks, 2000.
  [Bibtex]

  @book{
  Demuth00,
     Author = {Demuth, Howard and Beale, Mark},
     Title = {Neural Network Toolbox User's Guide: for use with MATLAB, version 4},
     Publisher = {MathWorks},
     Address = {Natick, MA},
     url = {http://www.mathworks.com/products/neuralnet/},
        Year = {2000} }

• M. Detyniecki, "Mathematical Aggregation Operators and their Application to Video Querying," Ph.D. dissertation PhD Thesis, 2000.
  [Bibtex]

  @phdthesis{
  Detyniecki00,
     Author = {Detyniecki, Marcin},
     Title = {Mathematical Aggregation Operators and their Application to Video Querying},
     School = {University of Paris 6. University Pierre and Marie Curie},
     Type = {Ph.D. dissertation},
     Abstract = {In this manuscript we present the mathematical aggregation operators and their application to the video querying. This work is divided in three parts. The first one offers the definition of mathematical aggregation operators and some properties, followed by an extensive overview of the existing operators. The second part is dedicated to the study of the aggregation under uncertainty. We present a deep study on t-norms and t-conorms, pursued by a study on aggregation of truth and falsity values in non-phrase calculus way. We also introduce a non-axiomatic way, based on the metaphor of a balance, which in the one hand allows the visualization of the global behavior and of the sensitivity of an operator and in the other hand offers a guide for the construction of additive generated operators.  The third part is devoted to the illustration of the theoretical results in the framework of video querying. We expound two complementary approaches. The first one based on "computing with words" explains how to browse a video with temporal queries. The second one makes obvious how to aggregate criteria pointing to the same conclusion. We prove the feasibility of the approach with real search engine and we expound the used technology (Java, XML, etc.)},
     Keywords = {Aggregation, Multimedia, Fuzzy logic, Data Fusion, Video, Truth and Falsity, Mathematical Operators, Query, Balance Metaphor},
     Year = {2000} }

• P. T. Jahromi, "Performance analysis of packet-switched networks with tree topology," Ph.D. dissertation PhD Thesis, 2000.
  [Bibtex]

  @phdthesis{
  Jahromi00,
     Author = {Jahromi, Payam Torab},
     Title = {Performance analysis of packet-switched networks with tree topology},
     School = {Georgia Institute of Technology},
     Type = {Ph.D. dissertation},
     Abstract = {Generic packet-switched networks with tree topology are the focus of this study. We present a systematic approach to model and analyze these networks from a performance standpoint, with emphasis on network topology as the common element in all network analysis problems. In particular, we study network  load and delay analysis problems. Given the network topology and traffic volume among all network nodes, load analysis is defined as computing the amount of traffic or load on all network nodes. Similarly, delay analysis generates end-to-end delay measures by combining the delay times at individual network nodes based on the network topology. Almost all developed algorithms have an iterative nature. They start with the network graph model, coupled with an appropriate performance matrix such as the traffic matrix. The idea is to do some simple processing on the performance matrix and trim the graph model to obtain a reduced equivalent network. The procedure is repeated until the network is reduced to a single node, which is usually the backbone switch in the network. Using these analysis techniques, instead of relying on common conceptions and legacy design rules for particular network topologies, the designer has the freedom of evaluating various topologies and selecting the one that results in the best performance for a particular application.},
     Keywords = {Pocket switching (Data transmission), Computer networks Mathematical models},
     Year = {2000} }

• J. Pan, "Modeling and process optimization of solder paste stencil printing for micro-BGA and fine pitch surface mount assembly," Ph.D. dissertation PhD Thesis, 2000.
  [Bibtex]

  @phdthesis{
  Pan02,
     Author = {Pan, Jianbiao},
     Title = {Modeling and process optimization of solder paste stencil printing for micro-BGA and fine pitch surface mount assembly},
     School = {Lehigh University},
     Type = {Ph.D. dissertation},
     Abstract = {The need for higher pin count, higher performance, smaller size and lighter weight has driven the development of advanced packages such as Quad Flat Package (QFP), Ball Grid Array (BGA), Chip Scale Package (CSP), and Flip Chip. These high-density electronic packages have had a tremendous impact on board-level assembly. One of the main challenges continues to be the solder paste deposition process and, specifically, controlling the amount of solder paste deposited. Stencil printing is one of the most cost-effective processes for solder paste deposition and it has been widely used in traditional surface mount assembly. However, the solder paste stencil printing process is still not completely understood as indicated by the fact that industry reports 52-71% of fine pitch and ultra-fine pitch SMT defects are related to the solder paste stencil printing process. The objective of this research was to understand and optimize the solder paste stencil printing process. An analytical model was developed using fluid mechanics theory. The paste flow pattern in front of the squeegee was studied and the volume of paste through the aperture was calculated. The relationship between transfer ratio and area ratio was derived. This model explains two experimental phenomena. One is that increasing stencil thickness does not necessarily lead to thicker deposits. The other is that perpendicular apertures print thicker than parallel apertures. An experiment was designed and conducted to validate the model and to investigate the effects of relevant process parameters on the amount of solder paste deposited. The factors selected in the experiment are aperture size, aperture shape, board finish, stencil thickness, solder type, and print speed. The deposited solder paste was measured by an automatic laser-based 3-D triangulation solder paste inspection system. The gauge repeatability & reproducibility (R&R) of the inspection system was evaluated. The experimental data were analyzed by Analysis of Variance (ANOVA). The critical variables were identified and interactions between process variables were determined. The experimental results are shown to be consistent with the theoretical model. This study will help to accelerate the development and utilization of BGA, CSP and flip chip packages.},
     Keywords = {I},
     Year = {2000} }

• A. Wachi, S. Yanachi, N. Kakishima, and M. Matsumoto, Screen printing method and screen printing apparatusMatsushita Electric Industrial Co., Ltd., 2000.
  [Bibtex]

  @misc{
  Wachi+etal00,
     Author = {Wachi, Akihiko and Yanachi, Seishiro and Kakishima, Nobuyuki  and Matsumoto, Masaya},
     Title = {Screen printing method and screen printing apparatus},
     Publisher = {Matsushita Electric Industrial Co., Ltd.},
     Month = {2000/10/17},
     Abstract = {PCT No. PCT/JP97/02027 Sec. 371 Date Dec. 4, 1998 Sec. 102(e) Date Dec. 4, 1998 PCT Filed Jun. 12, 1997 PCT Pub. No. WO97/48258 PCT Pub. Date Dec. 18, 1997  Coordinate positions of openings and lands on a screen and a first circuit board are recognized by a recognition camera and a display device. A correction amount for movement of the stage is calculated by a control device only for the first circuit board. Then, the correction of movement of the stage is executed for the second and following circuit boards by utilizing the correction amount.},
     Keywords = {USPAT SPP},
     ISBN = {101-129-000 101-126-000 101-DIG-0 36 033-620-000},
     Year = {2000} }

1999

• W. H. Fleming, W. M. McEneaney, G. Yin, and Q. Zhang, Stochastic analysis, control, optimization, and applications : a volume in honor of W.H. Fleming, Boston: Birkhäuser, 1999.
  [Bibtex]

  @book{
  Fleming99,
     Author = {Fleming, Wendell Helms and McEneaney, William M. and Yin, George and Zhang, Qing},
     Title = {Stochastic analysis, control, optimization, and applications : a volume in honor of W.H. Fleming},
     Publisher = {Birkhäuser},
     Address = {Boston},
     url = {William M. McEneaney, G. George Yin, and Qing Zhang, editors.
  ill. ; 25 cm.
  W.H. Fleming's curriculum vitae -- Representations for functionals of Hilbert space valued diffusions / A. Budhiraja and P. Dupuis -- Risk-sensitive, minimax, and mixed risk-neutral/minimax control of Markov decision processes / S.P. Coraluppi and S.I. Marcus -- Partially observed control problems with multiplicative cost / D. Hernandez-Hernandez -- Nonlinear semigroups for partially observed risk-sensitive control and minimax games / M.R. James -- Nonlinear, dissipative, infinite dimensional systems / M. Kocan and P. Soravia -- Singular limits of Bellman equations of ergodic type related to risk-sensitve control / H. Nagai -- Game approach to risk sensitive control for stochastic evolution systems / M. Nisio -- On the solutions of the equation arising from the singular limit of some Eigen problems / S.-J. Sheu and A.D. Wentzell.
  (continued) Nonlinear H[superscript exponent infinity] controller design via viscosity supersolutions of the Isaacs equation / M. Xiao and T. Basar -- Singularities of semiconcave functions in Banach spaces / P. Albano and P. Cannarsa -- Invariant sets for controlled degenerate diffusions: a viscosity solutions approach / M. Bardi and P. Goatin -- Remarks on the Dirichlet problem for quasilinear elliptic and parabolic equations / G. Barles, E. Rouy, and P.E. Souganidis -- A generalized Hamilton-Jacobi-Bellman equation for deterministic optimal control problems / L.D. Berkovitz -- Regular solutions of stochastic Burgers equation / P.L. Chow -- Piecewise-deterministic processes and viscosity solutions / M.H.A. Davis and M. Farid -- Mathematical approaches to the problem of noise-induced exit / M.V. Day.
  (continued) An approximation scheme for evolutive Hamilton-Jacobi equations / M. Falcone and T. Giorgi -- Homogenization of the Cauchy problem for Hamilton-Jacobi equations / H. Ishii -- The critical exponent for a stochastic PDE to hit zero / C. Mueller and E. Pardoux -- Robustness of Zakai's equation via Feynman-Kac representations / R. Atar, F. Viens, and O. Zeitouni -- Estimation of probability distributions for individual parameters using aggregate population data / H.T. Banks ... [et al.] -- Solvable infinite time horizon stochastic control problems in noncompact symmetric spaces / T.E. Duncan -- Exact finite dimensional filters for exponential functionals of the state / R.J. Elliott and V. Krishnamurthy -- A Lyapunov theory of nonlinear observers / A.J. Krener -- Existence of optimal controls for variance control / H.J. Kushner.
  (continued) On optimal ergodic control of diffusions with jumps / J.-L. Menaldi and M. Robin -- Markov marginal problems and their applications to Markov optimal control / T. Mikami -- Entropy inequalities and entropy dynamics in nonlinear filtering of diffusion processes / D. Ocone -- Identification for linear stochastic distributed parameter systems with boundary/point control / B. Pasik-Duncan -- Monte Carlo estimation of diffusion distributions at inter-sampling times / C.J. Wypasek, J.V. Butera, and B.G. Fitzpatrick -- Option pricing in a market with frictions / A. Bensoussan and H. Julien -- Pathwise comparison of arithmetic Brownian motions and log-normal processes / G. Ferreyra and P. Sundar -- Critical power for asymptotic connectivity in wireless networks / P. Gupta and P.R. Kumar -- Pricing models with transaction fees / J.E. Hodder and T. Zariphopoulou.
  (continued) A verification theorem in general equilibrium model of asset prices / C.-F. Huang, M. Taksar, and S.H. Zhu -- Optimal portfolio management with partial observations and power utility function / R. Rishel -- Hierarchical production controls for a stochastic manufacturing system with long-run average cost: asymptotic optimality / S.P. Sethi and H. Zhang.},
     Keywords = {Stochastic analysis, Control theory},
     Year = {1999} }

• E. W. Kamen and J. Su, Introduction to optimal estimation, London ; New York: Springer, 1999.
  [Bibtex]

  @book{
  Kam+Su99,
     Author = {Kamen, Edward W. and Su, Jonathan},
     Title = {Introduction to optimal estimation},
     Publisher = {Springer},
     Address = {London ; New York},
     Series = {Advanced textbooks in control and signal processing},
     Keywords = {Signal processing., Estimation theory., Mathematical optimization.},
     Year = {1999} }

• G. Monari and G. Dreyfus, Local Linear Least Squares: Performing Leave-one-out Without Leaving Anything OutHTTP, 1999.
  [Bibtex]

  @misc{
  Mon+Dre99,
     Author = {Monari, G.  and Dreyfus, G.},
     Title = {Local Linear Least Squares: Performing Leave-one-out Without Leaving Anything Out},
     Publisher = {HTTP},
     Volume = {2002},
     Number = {May 22},
     Abstract = {Based on a local linear expansion, in parameter space, of the least squares solution, we show that the effect of removing an example from the training set can be accurately predicted, and that, in addition, the geometrical interpretation of this expansion leads to a method of detecting models which are likely to exhibit overfitting. As a consequence, the computationally expensive leave-one-out procedure can be replaced by a prediction thereof: instead of training N different models on the data sets with N-1 examples, one has to train a single model on the whole data set. Moreover, we demonstrate that selecting models on the basis of the performance evaluation after leave-one-out training may lead to overfitted model even if the jacobian matrix is shown to be non-singular.The price to be paid for predicting the leave-one result is the computation of the diagonal elements of the orthogonal projection matrix onto the local solution subspace, in the neighborhood of the parameter vector obtained by training; this computational overhead is still very slight since this computation must be performed after training only. We have performed a detailed comparison between standard leave-one-out and our method, on a teacher-student problem, for performance evaluation and model selection. The method is by no means specific to neural networks: it can be applied to any nonlinear regression method, provided the jacobian matrix of the model is easily available.},
     Keywords = {"Learning" Conference, Snowbird, 1999},
        Year = {1999} }

• T. A. Nguty and N. N. Ekere, "The rheological properties of solder and solar pastes and the effect on stencil printing," , Springer-Verlag, Ed., Berlin Heidelberg: Springer-Verlag, 1999, vol. 39, Issue 6, pp. 607-612.
  [Bibtex]

  @incollection{
  Nguty+Ekere99,
     Author = {Nguty, T. A. and Ekere, Ndy N.},
     Title = {The rheological properties of solder and solar pastes and the effect on stencil printing},
     BookTitle = {Rheologica Acta},
     Editor = {Springer-Verlag},
     Publisher = {Springer-Verlag},
     Address = {Berlin Heidelberg},
     Volume = {39, Issue 6},
     Pages = {607-612},
     Abstract = {Solar and solder pastes are widely used in the electronics industry. Solder paste is the principal joining medium in the assembly of surface mount components, whilst solar paste is used in the manufacture of semiconductor solar cells in the photo-voltaic industry. The stencil printing of both solder and solar pastes is a very important and critical stage in the assembly process. With miniaturisation of components, this is likely to continue. The challenge in stencil printing at such dimensions is in achieving repeatable deposition of both solar and solder pastes from print to print. To meet this challenge requires an understanding of the flow behaviour of both solar and solder pastes. The rheological properties of solar and solder pastes have been evaluated through three different types of experiments. Existing models were applied to compare their rheological behaviour under these schemes. One striking difference was that solar paste showed a higher viscosity than solder paste. Both solar and solder pastes were found to be non-Newtonian materials, showing a decrease in viscosity with increasing shear rates. In this paper we investigate the rheological properties of both solder and solar paste under steady shear and creep-recovery tests.},
     Keywords = {Solar paste, Solder paste, Rheology, Creep recovery, Viscosity, Solar cells, Steady shear},
     Year = {1999} }

• S. Octavian, "New recursive algorithms for training feedforward multilayer perceptrons," Ph.D. dissertation PhD Thesis, 1999.
  [Bibtex]

  @phdthesis{
  Stan99,
     Author = {Octavian, Stan},
     Title = {New recursive algorithms for training feedforward multilayer perceptrons},
     School = {Georgia Institute of Technology},
     Type = {Ph.D. dissertation},
     Abstract = {The topic of this dissertation is the design of high performance least-squares training algorithms for a particular neural network architecture, the feedforward multilayer perceptron (MLP). Three new algorithms are presented. They address two important needs in MLP training: the need for high-efficiency algorithms to train the weights of high-complexity structures and the need for training algorithms with good tracking ability for modeling time-varying nonlinear functions. The first algorithm, called the local linearized least-squares (LLLS), addresses the issue of efficiency. LLLS is a computationally efficient approximation of the global extended Kalman filter (GEKF) algorithm. It is based on solving local system identification subproblems at the neuron level. The LLLS is shown to give better convergence results for two benchmark problems in comparison with existing local algorithms, and to be a good approximation for the GEKF algorithm, used as a benchmark. The second algorithm, called the block Levenberg Marquardt algorithm, addresses the issue of modeling nonlinear time-varying problems. We identify and analyze two important variations of the algorithm: the overlapping and the non-overlapping block Levenberg Marquardt algorithms. It is shown that the block Levenberg Marquardt outperforms other algorithms like the backpropagation algorithm and the global EKF with system noise for time-varying problems. The third algorithm is called block linearized least squares (BLLS), an alteration of the GEKF algorithm with a finite data window. Experiments show that the BLLS performs better than the standard GEKF for a benchmark time-varying problem, but both the GEKF with state noise and the block Levenberg-Marquardt algorithm exceed its performance. At last, the new algorithms are applied for neural network modeling of a process of fine pitch stencil printing for solder paste deposition in surface mount technology (SMT). A feedforward multilayer perceptron is trained to model the nonlinear relationship between the input variables to the process and the solder paste height/variance. The extension of this work to an on-line neural network control of the process is outlined as a topic of future research.},
     Keywords = {Operations research, Algorithms, Recursive functions, Perceptrons, Error-correcting codes, Neural networks (Computer science)},
     Year = {1999} }

• M. Tomomatsu, T. Sakaue, and M. Murakami, Solder paste screen printing apparatus and solder paste screen printing methodMatsushita Electric Industrial Co., Ltd., 1999.
  [Bibtex]

  @misc{
  Tomomatsu99,
     Author = {Tomomatsu, Michinori and Sakaue, Takaaki and Murakami, Minoru},
     Title = {Solder paste screen printing apparatus and solder paste screen printing method},
     Publisher = {Matsushita Electric Industrial Co., Ltd.},
     Month = {1999/12/07},
     Abstract = {There is disclosed a solder paste screen printing apparatus and a solder paste screen printing method, in which a force of pressing of a squeegee can be adjusted delicately and accurately so as to effect the printing of solder paste with the optimum pressing force. A squeegee is connected to a lower end of a rod of a cylinder, and a load cell is provided above the rod. A first pressure and a second pressure are applied respectively to an upper chamber and a lower chamber of the cylinder. A load, applied to the load cell, is measured while varying the level of the first pressure, and the pressure measurement for the pressing force is effected in accordance with the measured load. In accordance with this result, the instruction pressure for the cylinder, which is necessary for obtaining the optimum pressing force, is determined, and then the printing of solder paste onto a substrate is started with this instruction pressure.},
     Keywords = {USPAT SPP},
     ISBN = {101-123-000 101-124-000},
     Year = {1999} }

1998

• L. G. Barajas, "Neuro-fuzzy embedded control system," B.S. Honored Thesis PhD Thesis, 1998.
  [Bibtex]

  @phdthesis{
  Barajas98,
     Author = {Barajas, Leandro G.},
     Title = {Neuro-fuzzy embedded control system},
     School = {Universidad Distrital Francisco José de Caldas},
     Type = {B.S. Honored Thesis},
     Abstract = {This work presents a parallel hybrid control system that uses as main tools Fuzzy Logic and Neural Networks, showing the practical applications that these technologies have in the industry. Neural-Fuzzy Embedded Control System (SCINEF) was applied to the control of a 1MW-power generator to optimize its set point, diminishing in that way its fuel consumption up to in a 20% and decreasing the generator and refrigeration system wear. SCINEF includes RS232 and RS485 serial communications modules with speeds up to 115.2 Kbps, programmable digital inputs/outputs, relay outputs, parallel printer port, AD/DA converters to the analog interface with the fuzzy micro-controller and how general purpose analog input/output, analog voice play/recording module and Smart Cards interface. As main computing module it uses a Z80180 processor, which can be programmed in C and assembler languages, in which It was implemented, besides the conventional digital control, the neural network. The fuzzy logic processing module use the Programmable Analog IC (PAICTM) AL220 which can use up to 4 analog inputs and 4 analog outputs, 111 fuzzy variables, 50 rules that can be evaluated up to 500K rules/sec. The neural network used is a feed forward with 2 inner layers and back propagation was used as learning algorithm. To the hardware develop the PC/104 standard was used for physical dimensions and control signals, giving the possibility of integrate any kind and quantity of peripheral drivers that the system could require. The main advantages of this control system a re its high response velocity, proved efficiency, hardware versatility, software flexibility, high learning capabilities and the fact that it is a system easy to adapt to different industrial processes without any kind of hardware modifications.},
     Keywords = {Fuzzy Logic, Fuzzy Hardware, AL220, Neural Networks, feed-forward, back-propagation, Hybrid Control System, parallel processing, PC/104, Smart Card, Analog Voice Recording, Power Generation Control.},
     Year = {1998} }

• K. G. Hoebener, E. M. Hubacher, and J. P. Partridge, Print circuit board product with stencil controlled fine pitch solder formation for fine and coarse pitch component attachmentInternational Business Machines Corporation , Motorola, Inc., 1998.
  [Bibtex]

  @misc{
  Hoebener98,
     Author = {Hoebener, Karl Grant  and Hubacher, Eric Max   and Partridge, Julian Peter},
     Title = {Print circuit board product with stencil controlled fine pitch solder formation for fine and coarse pitch component attachment},
     Publisher = {International Business Machines Corporation , Motorola, Inc.},
     Month = {1998/10/20},
     Abstract = {A product created through the reflow of low melting point solder on select contacts of a printed circuit board. In one form, the printed circuit board has fine pitch devices, including flip-chip integrated circuits, connected to a board having conventional coarse pitch surface. The fine pitch contacts of the board are exposed through holes in a stencil characterized in its ability to withstand solder reflow temperatures, not be wettable by solder, and have a coefficient of thermal expansion relatively matching the printed circuit board. Low temperature solder paste is screen deposited into the stencil openings. With the stencil fixedly positioned on the board, the solder paste retained by the stencil pattern is reflowed to selectively form on the underlying contacts of the printed circuit board. Thereafter, the stencil is removed from the board and the board is subject to previously practiced depositions of flux and paste in preparation for fine and coarse pitch component placement and ensuing solder reflow. The fine contact pitch and coarse contact pitch differ by nominal factor of two or more.},
     Keywords = {USPAT SPP},
     ISBN = {361-777-000 174-261-000 228-180-021 228-180-022 361-760-000 361-774-000},
     Year = {1998} }

• A. Krauss, "Control of run-by-run processes with applications to large-area material deposition," Ph.D. dissertation PhD Thesis, 1998.
  [Bibtex]

  @phdthesis{
  Krauss98,
     Author = {Krauss, Alan},
     Title = {Control of run-by-run processes with applications to large-area material deposition},
     School = {Georgia Institute of Technology},
     Type = {Ph.D. dissertation},
     Abstract = {In this work, the run-by-run control problem as it applies to large-area material deposition in electronics packaging is considered. The first part of the work presents a discussion of constructibility and controllability of linear run-by-run processes, and a new matrix Kalman filtering technique for identification of coefficients in a linear process. The next part of the work focuses on nonlinear run-by-run processes for which a new multiple-point block-form linearization technique that may be used to control a nonlinear process is derived. Then, in the third part of the work, identification and nonlinear control techniques are applied to meniscus coating and stencil printing. Methods for identification and control discussed in this work are applied to experimental data. In the case of meniscus coating, control techniques are able to generate process parameters that keep the process output closer to a target value than when the process is run open-loop.},
     Keywords = {Electronic packaging, Manufacturing processes, Production engineering.},
     Year = {1998} }

• A. Lotfi and M. Howarth, An intelligent closed-loop control of solder paste stencil printing stage of surface mount technology, Nottingham: Nottingham Trent University. Department of Mechanical and Manufacturing Engineering, 1998.
  [Bibtex]

  @book{
  Lotfi+Howarth98,
     Author = {Lotfi, Ahmad and Howarth, Martin},
     Title = {An intelligent closed-loop control of solder paste stencil printing stage of surface mount technology},
     Publisher = {Nottingham Trent University. Department of Mechanical and Manufacturing Engineering},
     Address = {Nottingham},
     Keywords = {Surface mount technology, Solder pastes, Printed circuits},
     Year = {1998} }

• R. A. Marcoux and M. Curtin, Cartridge squeegee head with engagement locking mechanismTransition Automation, Inc., 1998.
  [Bibtex]

  @misc{
  Marcoux98,
     Author = {Marcoux, Richard A.  and Curtin, Mark},
     Title = {Cartridge squeegee head with engagement locking mechanism},
     Publisher = {Transition Automation, Inc.},
     Month = {1998/08/11},
     Abstract = {A cartridge squeegee head for solder paste machines. A squeegee blade is fixed to a coupling block. The coupling block is characterized by a coupling slot formed therein. The coupling block pivots and puts the squeegee blade into a print stroke position. The squeegee blade is then driven across the stencil foil surface.},
     Keywords = {USPAT SPP},
     ISBN = {101-123-000 101-127-000},
     Year = {1998} }

1997

• H. J. Kushner and G. Yin, Stochastic approximation algorithms and applications, Springer, 1997.
  [Bibtex]

  @book{
  Kushner97,
     Author = {Kushner, Harold J. and Yin, George},
     Title = {Stochastic approximation algorithms and applications},
     Publisher = {Springer},
     Keywords = {Stochastic approximation.},
     Year = {1997} }

• R. P. Prasad, Surface mount technology : principles and practice, 2nd ed., New York: Chapman & Hall, 1997.
  [Bibtex]

  @book{
  Prasad97,
     Author = {Prasad, Ray P.},
     Title = {Surface mount technology : principles and practice},
     Publisher = {Chapman & Hall},
     Address = {New York},
     Edition = {2nd},
     Keywords = {Surface mount technology.},
     Year = {1997} }

• P. Worhach, "Integration of environmental factors in manufacturing modeling for product planning and design, with applications to printed circuit board assembly," Ph.D. dissertation PhD Thesis, 1997.
  [Bibtex]

  @phdthesis{
  Worh97,
     Author = {Worhach, Paul},
     Title = {Integration of environmental factors in manufacturing modeling for product planning and design, with applications to printed circuit board assembly},
     School = {University of California},
     Type = {Ph.D. dissertation},
     Abstract = {This dissertation develops a planning methodology for integrating manufacturing-related environmental factors into product, process and facility design and operational decisions, with applications to printed circuit board assembly. Central to this methodology is the definition of manufacturing models for unit-level and batch processes that relate process parameters and component design features to waste streams, energy consumption, yield, and process time. Models for surface mount (stencil application and reflow soldering), through-hole component assembly (wave soldering), and board cleaning are developed and validated with data from several production lines. The process models are then used in production models of facility operations that include data from a range of product types and processes modeling manufacturing activities at the plant level. Metrics for evaluating hazard potential and costs associated with environmental impact are developed to guide planning and design decisions. Detailed case studies are presented for a single printed circuit board and for several boards produced over a planning horizon of one week in a single facility. Specific operational issues can be addressed through optimization formulations with relevant environmental objectives. For printed circuit board assembly, a model is formulated for the assignment of boards to production lines and workers to production lines to minimize overall waste generation and to minimize the maximum potential workplace hazard at any line due to these wastes. This work also contributes to the fields of life-cycle analysis (LCA) and design-for-environment (DfE) by developing an analysis methodology based on process models, rather than on aggregate industry or plant level data. When the process models are used in conjunction with the production models, the effects of the timing and location of wastes, which are typically not accounted for in LCA, can be included in the analysis. The LCA and DfE literature is comprehensively reviewed to provide the context for this contribution. The key contributions of this dissertation are: (1) The collection of detailed process waste data (solder paste, solvents, water, solder dross, and volatile organic compounds) and energy utilization data (thermal and mechanical) for printed circuit board assembly from several industrial sites for VOC clean, aqueous clean, and no-clean processes. (2) The development of process models for solder paste stenciling, reflow soldering, VOC and aqueous clean, and wave soldering. The models express waste and energy flows on a per-unit and per-batch basis. Parameters from the models are estimated from the process data, and the models are validated with aggregate waste and energy flow data. (3) The specification of mappings from the domain of the process data and the process models to decisions at the product and process levels of design. (4) The formulation of production optimization models for the operational management of hazard profiles and total solder paste waste generation in a production facility for printed circuit board assembly. The models plan production schedules and worker assignments; an integrated production-worker assignment problem is formulated as well. (5) The implementation of a World Wide Web interface for the process and production optimization models.},
        Year = {1997} }

1996

• G. H. Golub and C. F. Van Loan, Matrix computations, 3rd ed., Baltimore, MD.: Johns Hopkins University Press, 1996.
  [Bibtex]

  @book{
  Gol+VL96,
     Author = {Golub, Gene H. and Van Loan, Charles F.},
     Title = {Matrix computations},
     Publisher = {Johns Hopkins University Press},
     Address = {Baltimore, MD.},
     Edition = {3rd},
     Series = {Johns Hopkins studies in the mathematical sciences},
     Keywords = {Matrices Data processing.},
     Year = {1996} }

• Y. Li, "Yield Improvement, Reliability Modeling and Design Optimization for Solder Interconnection (Joints)," Ph.D. dissertation PhD Thesis, 1996.
  [Bibtex]

  @phdthesis{
  Li96,
     Author = {Li, Yuan},
     Title = {Yield Improvement, Reliability Modeling and Design Optimization for Solder Interconnection (Joints)},
     School = {University of Colorado at Boulder},
     Type = {Ph.D. dissertation},
     Abstract = {This research studies the effect of the surface mount assembly process and solder joint design on the quality and reliability of solder joints using a judicious combination of design of experiment, neural network and stress/strain finite element analysis techniques. A statistical-neural network modeling approach is applied to optimize stencil printing process. The objective is to determine the settings of the machinery parameters that result in minimum solder paste height variation. First, a Taguchi array, L27, is designed to conduct the experiment. Then, neural network models are developed to relate the desired quality characteristics to the parameters. A modular approach is used to select the appropriate architecture for these models. These models, in conjunction with the gradient descent algorithm, are used to determine the optimum settings. Confirming experiments on the production line validate the optimum settings predicted by the model. The effect of stencil printing optimization on the reliability of the ceramic and plastic ball grid arrays (BGA) is also investigated. The procedures to calculate the solder fillet shape for ceramic BGAs using mathematical functions, and to predict the solder joint shape for plastic BGAs using finite element solutions are described. Stress/strain finite element analysis and the Coffin-Manson relationship are applied to calculate the mean fatigue lives of the solder joints. The results reveal that an optimized stencil printing process significantly reduces variation in the life of ceramic BGAs, and the maximum strain region shifts from the card-side eutectic solder to the module side as the card-side solder volume increases. This shift in maximum strain suggests that there exists an optimum ratio between the card-side and the module-side solder volume for the reliability of a given ceramic BGA. The implications of this for the package developers and users are discussed. On the other hand, the calculations indicate that the life of plastic BGAs is practically insensitive to the card-side solder volume. In addition to stencil printing process, the solder joint design has a significant effect on its reliability. A ceramic BGA is selected as the demonstration vehicle. A central composite design is employed to investigate the effect of the design parameters such as solder volume, pad size and solder ball size on the reliability. A response surface is built and used to find the optimum design values. The results reveal that, all three parameters are significant. In particular, it is found that the optimum value of the card-side solder volume is close to the module-side solder volume, the fatigue life does not vary monotonically with the solder ball size--the interplay of stress concentration and the compliancy of a solder joint decides its effect on the reliability, and beyond some value of the card pad size, its effect on fatigue life is marginal. (Abstract shortened by UMI.)},
        Year = {1996} }

• S. Venkateswaran, "Comprehensive Process Control for Solder Paste Stencil Printing," Ph.D. dissertation PhD Thesis, 1996.
  [Bibtex]

  @phdthesis{
  Venka96,
     Author = {Venkateswaran, S.},
     Title = {Comprehensive Process Control for Solder Paste Stencil Printing},
     School = {State University of New York at Binghamton},
     Type = {Ph.D. dissertation},
        Year = {1996} }

1994

• T. Nanzai, Method for screen printing of pasteTani Denkikogyo Co., Ltd., 1994.
  [Bibtex]

  @misc{
  Nanzai94,
     Author = {Nanzai, Takashi},
     Title = {Method for screen printing of paste},
     Publisher = {Tani Denkikogyo Co., Ltd.},
     Month = {1994/05/10},
     Abstract = {A method for printing solder paste onto a surface of a printed circuit board through the medium of a screen having openings. The method includes feeding the paste into a chamber of a dispenser, dispensing the paste contained in the chamber onto a first side of the screen and onto the surface of the board from the dispenser through a slit-like aperture, removing paste residue from the first side of the screen by a blade extending from a first cleaning apparatus, removing paste residue from a second side of the screen by another blade extending from a second cleaning apparatus, and sucking and transporting the paste residue adhered to inner peripheral surfaces of openings in the screen as well as the paste residue being removed from the second side of the screen by a vacuum device connected to the second cleaning apparatus. The slit-like aperture of the dispenser is formed by a fixed front wall and a moveable rear wall selectively moving toward and away from the fixed front wall.},
     Keywords = {USPAT SPP},
     ISBN = {101-425-000 101-129-000},
     Year = {1994} }

• C. -Y. Yang and J. -C. Lin, Use of radius weighted mean to cluster two-class data, 1994.
  [Bibtex]

  @misc{
  Yang+Lin94,
     Author = {Yang, C.-Y. and Lin, J.-C.},
     Title = {Use of radius weighted mean to cluster two-class data},
     Volume = {30},
     Number = {10},
     Pages = {757-759},
     Abstract = {A new method using the radius weighted mean to cluster two-class data is proposed. Experiments show that the clustering results are good, the computation is fast, and the method is easy to implement. The method can be applied to block truncation coding, codebook generation, and decision tree construction},
     Keywords = {block codes, image coding, pattern recognition, vector quantisation},
        Year = {1994} }

1993

• C. Chung, "Closed Loop Process Control for Solder Paste Stencil Printing," Ph.D. dissertation PhD Thesis, 1993.
  [Bibtex]

  @phdthesis{
  Chung93,
     Author = {Chung, Chao-Wen},
     Title = {Closed Loop Process Control for Solder Paste Stencil Printing},
     School = {State University of New York at Binghamton},
     Type = {Ph.D. dissertation},
     Abstract = {The importance of solder paste has increased along with the move towards fine and ultra- fine pitch Surface Mount Components (SMCs). Increasing component lead counts and decreasing lead pitches mandate the need to thoroughly understand surface mount technology and solder paste related issues. For fine pitch applications, solder paste is best applied to the PCB by using a stencil, a stencil printer, and a squeegee. The print quality depends on several domain related parameters such as the printing equipment used, the PCB characteristics, the stencil, and the solder paste used. These factors are often inter-related. Studying the general relationship among the domain parameters and their effects on a print will help achieve good print quality. Knowledge based expert systems can be used to provide decision support, design closed loop control systems, diagnose defects, and monitor processes in every facet of the surface mount PCB assembly domain. The heuristic nature of this domain and the lack of widespread domain knowledge makes it an ideal field for expert system development and use.The global objective of this research was to design and develop a prototype real-time closed-loop process control system to monitor the solder paste stencil printing process for the assembly of surface mount PCBS. This prototype knowledge based stencil printing process control system was implemented on a personal computer and connected to the stencil printer and a laser based solder paste inspection system to form a closed-loop. Validation of this system reveals that this prototype is capable of recommending the expected printing parameters for either a 'new' PCB application or an 'existing' PCB application. Furthermore, this stencil printing process control system can communicate with the DEK 265 stencil printer to transmit the print commands and to precisely adjust the print parameter values. This prototype process control system achieves 'optimal' print conditions through an iterative approach.},
        Year = {1993} }

• M. Csörgö and L. Horváth, Weighted approximations in probability and statistics, Chichester, England ; New York: John Wiley & Sons, 1993.
  [Bibtex]

  @book{
  Csorgo93,
     Author = {Csörgö, M. and Horváth, Lajos},
     Title = {Weighted approximations in probability and statistics},
     Publisher = {John Wiley & Sons},
     Address = {Chichester, England ; New York},
     Series = {Wiley series in probability and mathematical statistics. Probability and mathematical statistics.},
     Keywords = {Stochastic approximation.},
     Year = {1993} }

1992

• H. J. Kushner and P. Dupuis, Numerical methods for stochastic control problems in continuous time, Springer-Verlag, 1992.
  [Bibtex]

  @book{
  Kushner92,
     Author = {Kushner, Harold J. and Dupuis, Paul},
     Title = {Numerical methods for stochastic control problems in continuous time},
     Publisher = {Springer-Verlag},
     Keywords = {Stochastic control theory, Markov processes Numerical analysis},
     Year = {1992} }

1991

• J. D. Lambert, Numerical methods for ordinary differential systems : the initial value problem, Wiley, 1991.
  [Bibtex]

  @book{
  Lambert91,
     Author = {Lambert, J. D.},
     Title = {Numerical methods for ordinary differential systems : the initial value problem},
     Publisher = {Wiley},
     Keywords = {Initial value problems Numerical solutions},
     Year = {1991} }

• A. Papoulis, Probability, random variables, and stochastic processes, 3rd ed., New York: McGraw-Hill, 1991.
  [Bibtex]

  @book{
  Pap91,
     Author = {Papoulis, Athanasios},
     Title = {Probability, random variables, and stochastic processes},
     Publisher = {McGraw-Hill},
     Address = {New York},
     Edition = {3rd},
     Series = {McGraw-Hill series in electrical engineering. Communications and signal processing},
     Keywords = {Probabilities., Random variables., Stochastic processes.},
     Year = {1991} }

• H. Tore, Process control in practice, Lund : Studentlitteratur ; [Bromley, Kent], Chartwell-Bratt, 1991.
  [Bibtex]

  @book{
  Hagglund91,
     Author = {Tore, Hägglund},
     Title = {Process control in practice},
     Publisher = {Lund : Studentlitteratur ; [Bromley, Kent], Chartwell-Bratt},
     Keywords = {Industries, Automatic control systems},
     Year = {1991} }

1990

• A. Benveniste, M. Métivier, and P. Priouret, Adaptive algorithms and stochastic approximations, Springer-Verlag, 1990.
  [Bibtex]

  @book{
  Benveniste90,
     Author = {Benveniste, Albert and Métivier, Michel and Priouret, P.},
     Title = {Adaptive algorithms and stochastic approximations},
     Publisher = {Springer-Verlag},
     Keywords = {Stochastic approximation., Sequential analysis., Algorithms.},
     Year = {1990} }

• A. Hald, A History of Probability and Statistics and Their Application Before 1750, New York, NY: John Wiley & Sons, 1990.
  [Bibtex]

  @book{
  Hald90,
     Author = {Hald, Anders},
     Title = {A History of Probability and Statistics and Their Application Before 1750},
     Publisher = {John Wiley & Sons},
     Address = {New York, NY},
     Series = {Wiley Series in Probability and Statistics},
     Abstract = {An account of the early development of probability theory and games of chance, statistics in astronomy and demography, and life insurance mathematics.},
        Year = {1990} }

1989

• K. E. Atkinson, An introduction to numerical analysis, 2nd ed., New York: Wiley, 1989.
  [Bibtex]

  @book{
  Atk89,
     Author = {Atkinson, Kendall E.},
     Title = {An introduction to numerical analysis},
     Publisher = {Wiley},
     Address = {New York},
     Edition = {2nd},
     Keywords = {Numerical analysis},
     Year = {1989} }

1987

• D. B. Metzger, G. S. Johnson, and M. D. Bramel, Method and apparatus for screen printing solder paste onto a substrate with device premounted thereonGeneral Motors Corporation, 1987.
  [Bibtex]

  @misc{
  Metzger87,
     Author = {Metzger, David B.  and Johnson, Glenn S. and Bramel, Michael D.},
     Title = {Method and apparatus for screen printing solder paste onto a substrate with device premounted thereon},
     Publisher = {General Motors Corporation},
     Month = {1987/07/07},
     Abstract = {A printing screen for screen printing fabrication substances such as solder paste has an aperture formed therein so that the screen may be placed on a substrate surface with a pre-mounted device extending through the aperture. A metal cap is bonded to the screen over the pre-mounted device to protect the device against damage during a subsequent screen printing operation. Different elements may, therefore, be bonded to a substrate in completely independent steps without the need for fluxing operations in later bonding operations.},
     Keywords = {USPAT SPP},
     ISBN = {156-250-000 118-504-000 156-252-000 427-096-000 427-282-000},
     Year = {1987} }

1984

• H. J. Kushner, Approximation and weak convergence methods for random processes, with applications to stochastic systems theory, MIT Press, 1984.
  [Bibtex]

  @book{
  Kushner84,
     Author = {Kushner, Harold J.},
     Title = {Approximation and weak convergence methods for random processes, with applications to stochastic systems theory},
     Publisher = {MIT Press},
     Keywords = {Stochastic processes, Approximation theory, Convergence, Stochastic systems},
     Year = {1984} }

1978

• I. A. Ibragimov and I. U. A. Rozanov, Gaussian random processes, New York: Springer-Verlag, 1978.
  [Bibtex]

  @book{
  Ibragimov78,
     Author = {Ibragimov, I. A. and Rozanov, I. U. A.},
     Title = {Gaussian random processes},
     Publisher = {Springer-Verlag},
     Address = {New York},
     Keywords = {Stochastic processes.},
     Year = {1978} }

• H. J. Kushner and D. S. Clark, Stochastic approximation methods for constrained and unconstrained systems, Springer-Verlag, 1978.
  [Bibtex]

  @book{
  Kushner78,
     Author = {Kushner, Harold J. and Clark, Dean S.},
     Title = {Stochastic approximation methods for constrained and unconstrained systems},
     Publisher = {Springer-Verlag},
     Keywords = {Stochastic approximation, Estimation theory, Convergence},
     Year = {1978} }

1959

• R. G. Brown, Statistical Forecasting for Inventory Control, New York: McGraw-Hill, 1959.
  [Bibtex]

  @book{
  Brown59,
     Author = {Brown, R.G.},
     Title = {Statistical Forecasting for Inventory Control},
     Publisher = {McGraw-Hill},
     Address = {New York},
        Year = {1959} }