Scheduling of non-repetitive lean manufacturing systems under uncertainty using intelligent agent simulation

World-class manufacturing paradigms emerge from specific types of manufacturing systems with which they remain associated until they are obsolete. Since its introduction the lean paradigm is almost exclusively implemented in repetitive manufacturing systems employing flow-shop layout configurations. Due to its inherent complexity and combinatorial nature, scheduling is one application domain whereby the implementation of manufacturing philosophies and best practices is particularly challenging. The study of the limited reported attempts to extend leanness into the scheduling of non-repetitive manufacturing systems with functional shop-floor configurations confirms that these works have adopted a similar approach which aims to transform the system mainly through reconfiguration in order to increase the degree of manufacturing repetitiveness and thus facilitate the adoption of leanness. This research proposes the use of leading edge intelligent agent simulation to extend the lean principles and techniques to the scheduling of non-repetitive production environments with functional layouts and no prior reconfiguration of any form. The simulated system is a dynamic job-shop with stochastic order arrivals and processing times operating under a variety of dispatching rules. The modelled job-shop is subject to uncertainty expressed in the form of high priority orders unexpectedly arriving at the system, order cancellations and machine breakdowns. The effect of the various forms of the stochastic disruptions considered in this study on system performance prior and post the introduction of leanness is analysed in terms of a number of time, due date and work-in-progress related performance metrics.