A full-factory simulator as a daily decision-support tool for 300MM wafer fabrication productivity

Bagchi, Sugato; Chen-Ritzo, Ching-Hua; Shikalgar, Sameer T.; Toner, Michael

doi:10.1109/wsc.2008.4736297

Cited by 27 publications

(15 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Instead we determine the Expected Work In Process will be for each tool and time period, based on a simulation of the flow of wafers through the fab. For this we use the IBM Research WIP Simulator, developed specifically for semi-conductor manufacturing (described in (Bagchi et al 2008)). We run 20 replications of the WIP simulator for a whole scheduling horizon, based on a division of the horizon into one hour time buckets.…”

Section: Objectivesmentioning

confidence: 99%

Integrated Maintenance Scheduling for Semiconductor Manufacturing

Davenport

2010

Lecture Notes in Computer Science

View full text Add to dashboard Cite

We present a maintenance scheduling problem arising from semi-conductor manufacturing which is characterized by low resource contention and multiple complex objectives and preferences. Since semi-conductor manufacturing involves a very high degree of uncertainty at the level of detailed operations, such as machine availability, yield, and processing times, generating a maintenance schedule which takes into account production operations is a challenging problem. We have developed an integrated approach, with respect to production operations, which uses simulation to estimate expected levels of work in process in the fab. We present details of our solution approach which is based on goal programming, constraint programming and mixed-integer programming.

show abstract

Section: Objectivesmentioning

confidence: 99%

Integrated Maintenance Scheduling for Semiconductor Manufacturing

Davenport

2010

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…We have therefore developed a specialized simulation model for operational decision support in the fab [16]. In contrast to previous work, our modeling intent is to cover the entire fab, so as to capture the impact of any operational decisions across all product routes and process centers, and to predict operational metrics over very short-time horizons of the order of hours and days.…”

Section: Motivation and Overviewmentioning

confidence: 99%

“…As such, there is no long-term production schedule that can be used to determine the WIP levels for each tool. Therefore, we use the expected WIP for each tool and time period obtained from WIPSim (Section 4; see also [16]) as follows. We use 20 replications of WIPSim for the whole scheduling horizon, based on a division of the horizon into one-hour time buckets, and from the WIPSim output, we obtain the expected WIP level for each tool during each time bucket.…”

Section: Objectivesmentioning

confidence: 99%

Data analytics and stochastic modeling in a semiconductor fab

Bagchi

Baseman

Davenport

et al. 2010

Appl Stoch Models Bus & Ind

Self Cite

View full text Add to dashboard Cite

The scale, scope and complexity of the manufacturing operations in a semiconductor fab lead to some unique challenges in ensuring product quality and production efficiency. We describe the use of various analytical techniques, based on data mining, process trace data analysis, stochastic simulation and production optimization, to address these manufacturing issues, motivated by the following two objectives. The first objective is to identify the sub-optimal process conditions or tool settings that potentially affect the process performance and product quality. The second objective is to improve the overall production efficiency through better planning and resource scheduling, in an environment where the product mix and process flow requirements are complex and constantly changing. Figure 5. Method for rule extraction and generalization ( denotes the empty rule; {Nodei} denotes the condition at the Nodei; the AND operator adds conditions to the rule).Figure 6. TDVs may reflect instabilities in various ways including gradual drifts and abrupt transitions.presumed tool stability issues. Alternatively, if the underlying analytics are designed to reveal TDV mismatches, then the corresponding heat map will highlight presumed tool matching issues. Assuming the effectiveness of the analytics layer, the TRACER chamber/process report then enables the supervising engineer to rapidly review thousands of underlying TDVs in a single view, in which the operationally significant signals are clearly highlighted. By clicking through the cells in the heat map, the associated detail reports for the relevant TDVs can be accessed, as illustrated in Figure 9.Although, as noted above, the TRACER framework typically uses only univariate analysis (since each TDV is analyzed independently), the heat-map representation implicitly reveals some of the data interactions. For example, if the detected problem is related to a particular process step, the

show abstract

“…Decision support, as the name implies, refers to providing information to clinicians, typically at the point of decision making. It comes in a variety of forms [11] and has also been applied to problems related to production, quality, and infrastructure across many fields [12][13][14]. However, many current decision support systems in healthcare rely on expert-or standards-based models, rather than models that adapt population-based guidelines to individual patient characteristics by utilizing existent EHR patient data.…”

Section: Introductionmentioning

confidence: 99%

EHRs connect research and practice: Where predictive modeling, artificial intelligence, and clinical decision support intersect

Bennett

Doub

Selove

2012

Health Policy and Technology

View full text Add to dashboard Cite

Objectives: Electronic health records (EHRs) are only a first step in capturing and utilizing healthrelated data-the challenge is turning that data into useful information. Furthermore, EHRs are increasingly likely to include data relating to patient outcomes, functionality such as clinical decision support, and genetic information as well, and, as such, can be seen as repositories of increasingly valuable information about patients' health conditions and responses to treatment over time.Methods: We describe a case study of 423 patients treated by Centerstone within Tennessee and Indiana in which we utilized electronic health record data to generate predictive algorithms of individual patient treatment response. Multiple models were constructed using predictor variables derived from clinical, financial and geographic data.Results: For the 423 patients, 101 deteriorated, 223 improved and in 99 there was no change in clinical condition. Based on modeling of various clinical indicators at baseline, the highest accuracy in predicting individual patient response ranged from 70% to 72% within the models tested. In terms of individual predictors, the Centerstone Assessment of Recovery Level-Adult (CARLA) baseline score was most significant in predicting outcome over time (odds ratio 4.1 + 2.27). Other variables with consistently significant impact on outcome included payer, diagnostic category, location and provision of case management services.Conclusions: This approach represents a promising avenue toward reducing the current gap between research and practice across healthcare, developing data-driven clinical decision support based on real-world populations, and serving as a component of embedded clinical artificial intelligences that ''learn'' over time.

show abstract

A full-factory simulator as a daily decision-support tool for 300MM wafer fabrication productivity

Cited by 27 publications

References 8 publications

Integrated Maintenance Scheduling for Semiconductor Manufacturing

Integrated Maintenance Scheduling for Semiconductor Manufacturing

Data analytics and stochastic modeling in a semiconductor fab

EHRs connect research and practice: Where predictive modeling, artificial intelligence, and clinical decision support intersect

Contact Info

Product

Resources

About