Power grids, flexible manufacturing, cellular communications: interconnectedness has consequences. This remarkable book gives the tools and philosophy you need to build network models detailed enough to capture essential dynamics but simple enough to expose the structure of effective control solutions and to clarify analysis. Core chapters assume only exposure to stochastic processes and linear algebra at the undergraduate level; later chapters are for advanced graduate students and researchers/practitioners. This gradual development bridges classical theory with the state-of-the-art. The workload model that is the basis of traditional analysis of the single queue becomes a foundation for workload relaxations used in the treatment of complex networks. Lyapunov functions and dynamic programming equations lead to the celebrated MaxWeight policy along with many generalizations. Other topics include methods for synthesizing hedging and safety stocks, stability theory for networks, and techniques for accelerated simulation. Examples and figures throughout make ideas concrete. Solutions to end-of-chapter exercises available on a companion website.

http://www.meyn.ece.ufl.edu/archive/spm_files/Presentations/Atlanta05.pdf

Control Techniques for Complex Networks

Variations of the kanban system: Literature review and classification

Since its introduction more than a decade ago, the CONstant Work In Process (CONWIP) production control system has received a great deal of attention from practitioners and researchers. In this paper, different contributions are reviewed according to three issues: operation, application, and comparison of Conwip. Referring to the operation of Conwip, some implementation decisions, such as lot sizing, have been overlooked, most studies being focused on card setting and job sequencing. Besides, no research has been conducted to show the relative importance of the different implementation decisions on the system performance. With respect to the application of Conwip, although it has been successfully implemented in many manufacturing environments, there are no studies to assess the impact of the manufacturing conditions on the performance of the system. Finally, most of the comparisons of Conwip with other production control systems suffer the lack of a unified framework for comparison where each system is t...

The CONWIP production control system: Review and research issues

This paper synthesizes recent work by the author and his colleagues on the performance analysis of manufacturing systems and on real-time scheduling methods. It describes a decentralized scheduling/control policy and proposes a method for analyzing the performance of systems that are operated according to the policy. It presents three views of the production system: surplus-based, time-based, and token-based; and shows how they are equivalent in special cases. While there are no assurances of optimality, experience suggests that the policy has desirable characteristics. The paper also suggests how the physical resources and the control policy can be designed together.

Design and operation of manufacturing systems: the control-point policy

During the past three decades, the success of the Toyota Production System has spurred much research in manufacturing systems engineering. Productivity and quality have been extensively studied, but there is little research in their intersection. The goal of this paper is to analyze how production system design, quality, and productivity are inter-related in small production systems. We develop a new Markov process model for machines with both quality and operational failures, and we identify important differences between types of quality failures. We also develop models for two-machine systems, with infinite buffers, buffers of size zero, and finite buffers. We calculate total production rate, effective production rate (ie, the production rate of good parts), and yield. Numerical studies using these models show that when the first machine has quality failures and the inspection occurs only at the second machine, there are cases in which the effective production rate increases as buffer sizes increase, and there are cases in which the effective production rate decreases for larger buffers. We propose extensions to larger systems.

Integrated quality and quantity modeling of a production line

We study the performance of the kanban, minimal blocking, basestock, CONWIP, and hybrid kanban-CONWIP control policies in a four-machine tandem production line making parts for an automobile assembly line. Cases with both constant and changing demand rates are studied. The main performance measures are the service level and the amount of work-in-progress. We also consider other performance measures such as variability amplification along the line. The results are obtained by extensive simulations. We find that the best parameter choices for the hybrid policy decrease inventories by 10% to 20% over the best kanban policy while maintaining the same service levels. The inventory difference grows as the demands on service level increase. The performance of basestock and CONWIP policies falls between those of the kanban and hybrid policies. The CONWIP and hybrid policies also give significantly better response to changes in the demand rate.

C.E. Couch

Papers

A comparison of production-line control mechanisms