Capacity Management in Reconfigurable Manufacturing Systems With Stochastic Market Demand

Abstract

An optimal solution, based on Markov Decision Theory, is presented for the capacity management problem in Reconfigurable Manufacturing Systems with stochastic market demand with a time delay between the time capacity change is ordered and the time it is delivered. The optimal policy in this paper is presented as optimal boundaries representing the optimal capacity expansion and reduction levels. The effects of change in the cost function parameters and the delay time on the optimal boundaries are presented for a capacity management scenario. The major differences between this research and the ones in inventory control lie in two folds. One is the fact that unlike inventory, capacity levels can be reduced according to the market demand. The other one is the novel approach presented in this paper to solve the delay problem which unlike the inventory control does not account for the cumulative unmet demand as a decision factor.

Optimal reconfiguration policy to react to product changes

“…This represents the unit cost to maintain the current level of system functionality available in period k. (3) represents the management cost for expanding or reducing the system functionality. A similar function has been proposed by Asl and Ulsoy (2002a) in their capacity expansion and reduction problem. The following equation provides a formal definition.…”

Multi-Period Production Capacity Planning for Integrated Product and Production System Design

“…Ref. [9] presents an optimal solution to the capacity management problem in Reconfigurable Machining Systems [10] using stochastic market demand with time delay between the time capacity change is ordered and the time it is delivered, based on the Markov Decision Theory. Ref.…”

Performance Issues in Reconfigurable Manufacturing System

“…Some researchers are of the opinion that it is much more important to decide a suitable policy of when, how, and how much to reconfigure the production system, under unreliable market demand. Further, they argued that it is not always beneficial to opt for high degree of reconfiguration very often as it would not always have economic benefits on ramp up time and associated costs (Asl & Ulsoy, 2002a, 2002b. It can be concluded that to have successful application of RMS, work must be done to develop strategies and techniques which would decrease this delay and improve the ramp up time of new configurations to improve system's performance (Dief & ElMaraghy, 2006b).…”

“…Dynamic approach towards capacity scalability modelling based on feedback control for RMS was proposed by Asl and Ulsoy (2002a). ElMaraghy (2006b, 2007b) also developed a dynamic model for capacity scalability for RMS.…”

“…The RMS throughput is significantly affected by several factors such as, reconfiguration ramp-up time, the reliability of a single component or equipment and process; and the buffer allocations. Throughput model based on these parameters was formulated by Chao et al (2007) for performance evaluation of manufacturing systems with reconfigurable configurations. Production rate in terms of low throughput time values was also used as a performance measure for studying different scheduling rules for manufacturing systems (Goyal et al, 1995).…”