Process Control Policy Research Articles

Policy Optimization in Dynamic Bayesian Network Hybrid Models of Biomanufacturing Processes

Biopharmaceutical manufacturing is a rapidly growing industry with impact in virtually all branches of medicine. Biomanufacturing processes require close monitoring and control, in the presence of complex bioprocess dynamics with many interdependent factors, as well as extremely limited data due to the high cost of experiments and the novelty of personalized bio-drugs. We develop a new model-based reinforcement learning framework that can achieve human-level control in low-data environments. A dynamic Bayesian network is used to capture causal interdependencies between factors and predict how the effects of different inputs propagate through the pathways of the bioprocess mechanisms. This model is interpretable and enables the design of process control policies that are robust against model risk. We present a computationally efficient, provably convergent stochastic gradient method for optimizing such policies. Validation is conducted on a realistic application with a multidimensional, continuous state variable. History: Accepted by Bruno Tuffin, Area Editor for Simulation. Funding: This work was partially supported by National Institute of Standards and Technology [Grant 70NANB17H002]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/ijoc.2022.1232 .

Editorial 2017 Best Paper Award

Each year the Transactions on Semiconductor Manufacturing recognizes the best paper published during the previous calendar year. This year 9 papers were nominated by our reviewers and members of our Editorial Board and Steering Committee, out of a total of 78 papers published in our pages in 2017. From among these, the members of the Editorial Board and Steering Committee selected the paper entitled “A Novel Dynamic Process Control Scheme to Improve Cycle Time in Fabs” by Dr. Michael Hassoun of the Department of Industrial Engineering and Management of Ariel University in Israel. This paper continues an ongoing line of research examining the impact of process control policies on the tradeoff between cycle time and yield in wafer fabrication. The primary finding is that modest relaxation of process control parameters to reduce time spent adjusting production equipment when it is in high demand can often reduce cycle time with minimal negative impact. A number of fruitful extensions for future work are also discussed, holding out the possibility of improved integration of scheduling and process control techniques. On behalf of the TSM community, I have the pleasure of congratulating Dr. Hassoun on this recognition, and of wishing him continued success in future.

Simulation of Semiconductor Manufacturing Supply-Chain Systems With DEVS, MPC, and KIB

The dynamics of high-volume discrete-part semiconductor manufacturing supply-chain systems can be described using a combination of Discrete EVent System Specification (DEVS) and model predictive control (MPC) modeling approaches. To formulate the interactions between the discrete process model and its controller, another model called Knowledge Interchange Broker (KIB) is used. A robust and scalable testbed supporting DEVS-based manufacturing process modeling, MPC-based controller design, and the KIBDEVS/MPC interaction model is developed. A suite of experiments have been devised and simulated using this testbed. The flexibility of this approach for modeling, simulating, and evaluating stochastic discrete process models under alternative control schemes is detailed. The testbed illustrates the benefits and challenges associated with developing and using realistic manufacturing process models and process control policies. The simulation environment demonstrates the importance of explicitly defining and exposing the interactions between the manufacturing and control subsystems of complex semiconductor supply-chain systems.

Simultaneous optimization of [Xbar] control chart and age-based preventive maintenance policies under an economic objective

The economic design of control charts and the optimization of preventive maintenance policies have separately received a tremendous amount of attention in the quality and reliability literature over the years in an attempt to reduce the costs associated with operating manufacturing processes. Not until recently has the proposal been made to integrate these two fields and utilize the relationship between quality and equipment performance to improve the productivity of a manufacturing process. In this paper, we extend the initial preliminary investigation of this idea of using an chart in conjunction with an age-replacement preventive maintenance policy. We formulate a partially observable, discrete-time Markov decision process in order to obtain the near-optimal combined preventive maintenance/statistical process control policy that minimizes the costs associated with maintenance, sampling, and poor quality. We develop transition probabilities for the various states of the infinite horizon problem and a solution algorithm for finding the best policy in polynomial time by finding a control limit on the sampling policy. We also perform sensitivity analysis on the decision variables for each of the various input parameters. It is shown that in every case a combined PM and SPC policy is the most cost efficient.

Simulation-based optimization of process control policies for inventory management in supply chains

A simulation-based optimization framework involving simultaneous perturbation stochastic approximation (SPSA) is presented as a means for optimally specifying parameters of internal model control (IMC) and model predictive control (MPC)-based decision policies for inventory management in supply chains under conditions involving supply and demand uncertainty. The effective use of the SPSA technique serves to enhance the performance and functionality of this class of decision algorithms and is illustrated with case studies involving the simultaneous optimization of controller tuning parameters and safety stock levels for supply chain networks inspired from semiconductor manufacturing. The results of the case studies demonstrate that safety stock levels can be significantly reduced and financial benefits achieved while maintaining satisfactory operating performance in the supply chain.

Sequential process control under capacity constraints

We consider a sequential process control model under inspection capacity constraints. A 2-state production process observed imperfectly by inspection with general costs and distribution functions is studied. The objective is to develop implementable control policies deciding at each decision epoch whether to produce items without any action, to inspect items, or to repair the process as well as satisfying inspection capacity constraints. The problem is formulated as a Markov decision process using the Bayesian approach. We characterize the features of an unconstrained problem and show that a constrained problem can be transformed to an unconstrained problem. A simplified process control policy is proposed, and numerical examples are presented which indicate that the approximate policy may provide very close results compared with the optimal policy in terms of total operation costs.

Optimal process control policies under a time-varying cost structure

Quality experts have recognized two types of quality costs resulting from failure of control; namely, internal and external failure costs of control. Explicitly considering the difference between internal and external failure costs in designing statistical process control procedures is especially important in today’s business environment because (1) inventory holding time is strategically kept very short because of the just-in-time philosophy, and (2) product values decrease at a very fast rate over time because of advances in technology and keen competition. In this paper, we develop a process control model under a time-varying cost structure, based on internal and external failure costs. Using the model, we also study the process control policies when the value of the product may decrease over time. We consider two forms of product perishability. The first is an exponential product value drop, and the second is a fixed life product life.

EFFECTS OF RANDOM SHIFTS OF TESTING EQUIPMENT ON PROCESS CONTROL SYSTEM DESIGN AND SELECTION OF PROCESS CONTROL POLICIES*

This paper studies issues associated with designing process control systems when the testing equipment is subjected to random shifts. We consider a production process with two states: in control and out of control. The process may shift randomly to the out‐of‐control state over time. The process is monitored by periodically sampling finished items from the process. The equipment used to test sampled items also is assumed to have two states and may shift randomly during the testing process. We formulate a cost model for finding the optimal process control policy that minimizes the expected unit time cost. Numerical results show that shifts of the testing equipment may significantly affect the performance of a process control policy. We also studied the effects of the testing equipment's shifts on the selection of process control policies.

Taguchi's method and an empirical procedure for on-line process control

Taguchi et al. (1989) proposed an effective on-line process control policy by minimizing the expected quality cost per item in a production process. The goal is to continually improve or regulate the process so that the quality characteristic stays as close to the target value as possible. Although Taguchi's method is easy to implement, it produces higher expected quality cost than other methods under certain cost and model structures. We propose an empirical procedure in this paper since the analytical solution of control scheme for stochastic models other than the integrated moving average process, IMA(1,1), is not available. The procedure combines the phase of modeling the measurements of quality characteristic by an ARIMA process, and the phase of obtaining a control scheme based on simulation. The performance of the empirical procedure is examined by comparing favorably its expected quality costs with that of Taguchi's procedure.

An economic model for determining the optimal quality and process control policy in a queue-like production system

SUMMARY This paper interrelates and synthesizes the economic design of quality and process control policy and the management of a production system described in terms of an M/G/X queue. A mathematical model is developed to determine the optimal quality and process control policy when both the quality and quantity issues in the production system are considered. The optimal control policy is defined as the one that maximizes the expected profit per unit time over an infinite horizon. Hypothetical data is used to illustrate the impact of on-line control policies on the operating characteristics of the production process. For a given control policy, it is shown that the expected profit per unit time of the production system can be increased by either improving the reliability of the production system, increasing the arrival rate of incoming jobs, increasing the system's processing rate, or shortening the system's shutdown time. The results also indicate that the larger the performance difference between the i...

An Integrated Cost Model for the Joint Optimization of Process Control and Maintenance

Statistical process control and preventive maintenance have been treated separately by past research. This study presents an economic model that incorporates both process control and maintenance procedures, and simultaneously optimizes their design parameters. It is shown how a pure process-control policy and a pure preventive-maintenance policy constitute special cases of the general model. Numerical examples illustrate the application of the method and suggest that, under the Markovian deterioration assumption, in many cases a pure policy is economically justifiable.