Planning, scheduling and dispatching play critical roles in the operations of a supply chain. The... more Planning, scheduling and dispatching play critical roles in the operations of a supply chain. Their definitions are clearly given through a unified view in this paper. The distinction between planning and scheduling is analyzed from the view point of microeconomics and queueing theory. The distinction between scheduling and dispatching is analyzed from the view point of computational complexity and hierarchical decomposition. Based on the elasticity of price and capacity, planning can be separated into demand planning or capacity planning. Scheduling period is the time horizon where price and average production cost are insensitive to the production rate. The critical roles of the master production schedule and move targets in job scheduling have been explained through the concept of hierarchical decomposition. Dispatching is the last layer of job scheduling in the hierarchical decomposition. The advantage of pull and push systems has been compared and analyzed systematically.
With advances in information and telecommunication technologies and data-enabled decision making,... more With advances in information and telecommunication technologies and data-enabled decision making, smart manufacturing can be an essential component of sustainable development. In the era of the smart world, semiconductor industry is one of the few global industries that are in a growth mode to smartness, due to worldwide demand. The important opportunities that can boost the cost reduction of productivity and improve quality in wafer fabrication are based on the simulations of actual environment in Cyber-Physical Space and integrate them with decentralized decision-making systems. However, this integration faced the industry with novel unique challenges. The stream of the data from sensors, robots, and Cyber-Physical Space can aid to make the manufacturing smart. Therefore, it would be an increased need for modeling, optimization, and simulation for the value delivery from manufacturing data. This paper aims to review the success story of smart manufacturing in semiconductor industry with the focus on data-enabled decision making and optimization applications based on operations research and data science perspective. In addition, we will discuss future research directions and new challenges for this industry.
2017 IEEE/ACIS 16th International Conference on Computer and Information Science (ICIS)
This paper presents an efficient algorithm for diffusion scheduling in a semiconductor fab. The d... more This paper presents an efficient algorithm for diffusion scheduling in a semiconductor fab. The diffusion area commonly creates long queue time in the entire process flow. Due to the complex constraints, such as parallel batching and time windows, and large solution space, it is difficult to find a feasible schedule in a timely manner. A greedy randomized procedure forms the batches. A heuristic method is introduced to handle the time window constraints. Two important properties of the problem are identified and applied to improve the quality of the solution. Simulated annealing is used as a local search procedure. Compared with the real schedule in the fab, the proposed algorithm can increase the effective moves significantly without violating queue time constraints.
Proceedings of the 2006 Winter Simulation Conference, 2006
Contemporary technology for Product Lifecycle Management (PLM) integrates computer aided design (... more Contemporary technology for Product Lifecycle Management (PLM) integrates computer aided design (CAD) and engineering analysis (CAE) to support rapid, distributed, team-oriented product data development and management, including high fidelity simulation on demand.. This technology potentially provides a platform for creating a new generation of factory design tools which enable "on demand" simulation and analytic model results to be used by factory designers. This paper describes the opportunity, and provides an illustration in the context of semiconductor wafer fab design.
Performance evaluation plays a key role in manufacturing system design and productivity improveme... more Performance evaluation plays a key role in manufacturing system design and productivity improvement. Characterizing performance objectively is the first step. Inspired by the underlying structure of tandem queues, we have derived an approximate model to characterize the system performance. The model decomposes system queue time and variability into bottleneck and non-bottleneck parts while capturing the dependence among workstations. Compared the new model with prior approaches, the new model not only is more accurate but also requires less information. The property of manufacturing system performance is given based on the insight from the model.
Tandem queues with finite buffer capacity commonly exist in practical applications. By viewing a ... more Tandem queues with finite buffer capacity commonly exist in practical applications. By viewing a tandem queue as an integrated system, an innovative approach has been developed to analyze its performance through the insight from reduction method. In our approach, the starvation at the bottleneck caused by service time randomness is modeled and captured by interruptions. Fundamental properties of tandem queues with finite buffer capacity are examined. We show that in general system service rate of a dual tandem queue with finite buffer capacity is equal or smaller than its bottleneck service rate, and virtual interruptions, which are the extra idle period at the bottleneck caused by the non-bottlenecks, depend on arrival rates. Hence, system service rate is a function of arrival rate when the buffer capacity of a tandem queue is finite. Approximation for the mean queue time of a dual tandem queue has been developed through the concept of virtual interruptions.
Due to the analytical intractability of general tandem queues, we develop innovative methods to q... more Due to the analytical intractability of general tandem queues, we develop innovative methods to quantify the dependence among stations through simulation. Dependence is defined by the contribution queue time at each station, and contribution factors are developed based on the insight from Friedman's reduction method and Jackson networks. In a tandem queue, the dependence among stations can be either diffusion or blocking, and their impact depends on the positions relative to the bottlenecks. Based on these results, we show that improving the performance of the system bottleneck may not be the most effective place to reduce system cycle time. Rather than making independence assumptions, the proposed method points out a promising direction and sheds light on the insights of the dependence in practical systems.
Dependence among single stations in series and its applications in productivity improvement
European Journal of Operational Research
ABSTRACT Theory of constraints has been commonly used in production systems to improve productivi... more ABSTRACT Theory of constraints has been commonly used in production systems to improve productivity. Since the improvement on an upstream workstation may have impact on its downstream servers, finding the true bottleneck is not trivial in a stochastic production line. Due to the analytical intractability of general tandem queues, we develop methods to quantify the dependence among stations through simulation. Dependence is defined by the contribution queue time at each station, and contribution factors are developed based on the insight from Friedman's reduction method and Jackson networks. In a tandem queue, the dependence among stations can be either diffusion or blocking, and their impact depends on the positions relative to the bottlenecks. Based on these results, we show that improving the performance of the system bottleneck may not be the most effective place to reduce system cycle time. Rather than making independence assumptions, the proposed method points out a promising direction and sheds light on the insights of the dependence in practical systems.
The objective of this dissertation is to enhance the overall understanding of practical manufactu... more The objective of this dissertation is to enhance the overall understanding of practical manufacturing systems by using rigorous academic approaches, primarily queueing theory. The scope spans from the performance of a single manufacturing process to the performance of a manufacturing system. Queueing models are commonly used to evaluate the performance of manufacturing systems. Exact M/M/1 or approximations of G/G/1 models are usually adopted to describe the behavior of a single machine system. However, when applying queueing models to a single machine, some practical issues are encountered. A real machine is subject to different types of interruptions, such as breakdowns, setups and routine maintenance. The proper queueing models under interruptions are presented. The behavior of manufacturing systems is explored by first investigating the underlying structure of tandem queues. We introduce two properties describing the dependence among servers in tandem queues, namely the intrinsi...
2007 IEEE International Conference on Automation Science and Engineering, 2007
Queueing theory is a powerful method to evaluate the performance of a manufacturing system. Howev... more Queueing theory is a powerful method to evaluate the performance of a manufacturing system. However, when applying it to a real system, both practical and theoretical issues arise. The practical issues are how to determine the parameter values for queuing models from the data available in real manufacturing systems. We explore this issue by comparing the data needed by queueing models to SEMI standards definitions. The theoretical issues are the selection of a specific queuing model and its proper use. In order to illustrate this, we design a specific situation for the applications of an M/M/1/Unreliable Machine queue, and compare the performances of two other approaches with it. The results show that, depending on the approach chosen, the differences in approximation error for cycle time can be over 30% at lower utilization levels.
Planning, scheduling and dispatching play critical roles in the operations of a supply chain. The... more Planning, scheduling and dispatching play critical roles in the operations of a supply chain. Their definitions are clearly given through a unified view in this paper. The distinction between planning and scheduling is analyzed from the view point of microeconomics and queueing theory. The distinction between scheduling and dispatching is analyzed from the view point of computational complexity and hierarchical decomposition. Based on the elasticity of price and capacity, planning can be separated into demand planning or capacity planning. Scheduling period is the time horizon where price and average production cost are insensitive to the production rate. The critical roles of the master production schedule and move targets in job scheduling have been explained through the concept of hierarchical decomposition. Dispatching is the last layer of job scheduling in the hierarchical decomposition. The advantage of pull and push systems has been compared and analyzed systematically.
With advances in information and telecommunication technologies and data-enabled decision making,... more With advances in information and telecommunication technologies and data-enabled decision making, smart manufacturing can be an essential component of sustainable development. In the era of the smart world, semiconductor industry is one of the few global industries that are in a growth mode to smartness, due to worldwide demand. The important opportunities that can boost the cost reduction of productivity and improve quality in wafer fabrication are based on the simulations of actual environment in Cyber-Physical Space and integrate them with decentralized decision-making systems. However, this integration faced the industry with novel unique challenges. The stream of the data from sensors, robots, and Cyber-Physical Space can aid to make the manufacturing smart. Therefore, it would be an increased need for modeling, optimization, and simulation for the value delivery from manufacturing data. This paper aims to review the success story of smart manufacturing in semiconductor industry with the focus on data-enabled decision making and optimization applications based on operations research and data science perspective. In addition, we will discuss future research directions and new challenges for this industry.
2017 IEEE/ACIS 16th International Conference on Computer and Information Science (ICIS)
This paper presents an efficient algorithm for diffusion scheduling in a semiconductor fab. The d... more This paper presents an efficient algorithm for diffusion scheduling in a semiconductor fab. The diffusion area commonly creates long queue time in the entire process flow. Due to the complex constraints, such as parallel batching and time windows, and large solution space, it is difficult to find a feasible schedule in a timely manner. A greedy randomized procedure forms the batches. A heuristic method is introduced to handle the time window constraints. Two important properties of the problem are identified and applied to improve the quality of the solution. Simulated annealing is used as a local search procedure. Compared with the real schedule in the fab, the proposed algorithm can increase the effective moves significantly without violating queue time constraints.
Proceedings of the 2006 Winter Simulation Conference, 2006
Contemporary technology for Product Lifecycle Management (PLM) integrates computer aided design (... more Contemporary technology for Product Lifecycle Management (PLM) integrates computer aided design (CAD) and engineering analysis (CAE) to support rapid, distributed, team-oriented product data development and management, including high fidelity simulation on demand.. This technology potentially provides a platform for creating a new generation of factory design tools which enable "on demand" simulation and analytic model results to be used by factory designers. This paper describes the opportunity, and provides an illustration in the context of semiconductor wafer fab design.
Performance evaluation plays a key role in manufacturing system design and productivity improveme... more Performance evaluation plays a key role in manufacturing system design and productivity improvement. Characterizing performance objectively is the first step. Inspired by the underlying structure of tandem queues, we have derived an approximate model to characterize the system performance. The model decomposes system queue time and variability into bottleneck and non-bottleneck parts while capturing the dependence among workstations. Compared the new model with prior approaches, the new model not only is more accurate but also requires less information. The property of manufacturing system performance is given based on the insight from the model.
Tandem queues with finite buffer capacity commonly exist in practical applications. By viewing a ... more Tandem queues with finite buffer capacity commonly exist in practical applications. By viewing a tandem queue as an integrated system, an innovative approach has been developed to analyze its performance through the insight from reduction method. In our approach, the starvation at the bottleneck caused by service time randomness is modeled and captured by interruptions. Fundamental properties of tandem queues with finite buffer capacity are examined. We show that in general system service rate of a dual tandem queue with finite buffer capacity is equal or smaller than its bottleneck service rate, and virtual interruptions, which are the extra idle period at the bottleneck caused by the non-bottlenecks, depend on arrival rates. Hence, system service rate is a function of arrival rate when the buffer capacity of a tandem queue is finite. Approximation for the mean queue time of a dual tandem queue has been developed through the concept of virtual interruptions.
Due to the analytical intractability of general tandem queues, we develop innovative methods to q... more Due to the analytical intractability of general tandem queues, we develop innovative methods to quantify the dependence among stations through simulation. Dependence is defined by the contribution queue time at each station, and contribution factors are developed based on the insight from Friedman's reduction method and Jackson networks. In a tandem queue, the dependence among stations can be either diffusion or blocking, and their impact depends on the positions relative to the bottlenecks. Based on these results, we show that improving the performance of the system bottleneck may not be the most effective place to reduce system cycle time. Rather than making independence assumptions, the proposed method points out a promising direction and sheds light on the insights of the dependence in practical systems.
Dependence among single stations in series and its applications in productivity improvement
European Journal of Operational Research
ABSTRACT Theory of constraints has been commonly used in production systems to improve productivi... more ABSTRACT Theory of constraints has been commonly used in production systems to improve productivity. Since the improvement on an upstream workstation may have impact on its downstream servers, finding the true bottleneck is not trivial in a stochastic production line. Due to the analytical intractability of general tandem queues, we develop methods to quantify the dependence among stations through simulation. Dependence is defined by the contribution queue time at each station, and contribution factors are developed based on the insight from Friedman's reduction method and Jackson networks. In a tandem queue, the dependence among stations can be either diffusion or blocking, and their impact depends on the positions relative to the bottlenecks. Based on these results, we show that improving the performance of the system bottleneck may not be the most effective place to reduce system cycle time. Rather than making independence assumptions, the proposed method points out a promising direction and sheds light on the insights of the dependence in practical systems.
The objective of this dissertation is to enhance the overall understanding of practical manufactu... more The objective of this dissertation is to enhance the overall understanding of practical manufacturing systems by using rigorous academic approaches, primarily queueing theory. The scope spans from the performance of a single manufacturing process to the performance of a manufacturing system. Queueing models are commonly used to evaluate the performance of manufacturing systems. Exact M/M/1 or approximations of G/G/1 models are usually adopted to describe the behavior of a single machine system. However, when applying queueing models to a single machine, some practical issues are encountered. A real machine is subject to different types of interruptions, such as breakdowns, setups and routine maintenance. The proper queueing models under interruptions are presented. The behavior of manufacturing systems is explored by first investigating the underlying structure of tandem queues. We introduce two properties describing the dependence among servers in tandem queues, namely the intrinsi...
2007 IEEE International Conference on Automation Science and Engineering, 2007
Queueing theory is a powerful method to evaluate the performance of a manufacturing system. Howev... more Queueing theory is a powerful method to evaluate the performance of a manufacturing system. However, when applying it to a real system, both practical and theoretical issues arise. The practical issues are how to determine the parameter values for queuing models from the data available in real manufacturing systems. We explore this issue by comparing the data needed by queueing models to SEMI standards definitions. The theoretical issues are the selection of a specific queuing model and its proper use. In order to illustrate this, we design a specific situation for the applications of an M/M/1/Unreliable Machine queue, and compare the performances of two other approaches with it. The results show that, depending on the approach chosen, the differences in approximation error for cycle time can be over 30% at lower utilization levels.
Uploads
Papers by Kan Wu