A Framework for Modelling Reconfigurable Manufacturing Systems Using Hybridized Discrete-Event and Agent-based Simulation

Abstract

Important objectives and challenges in today'ns manufacturing environment include the introduction of new products variants and the designing and developing of reconfigurable manufacturing systems. Due to changing customer requirements, products development time is shorter and manufacturing systems'n ability to physically reconfigure is important. The objective of this research is to investigate and support the reconfigurability of a manufacturing system by applying hybridized Agent-based and Discrete-Event simulation modelling technique. Emergent behaviour of the simulation model, when various modifications take place in the system, is examined. The benefits of this framework are decentralized control and collaborative decision making using the UML object-oriented modelling technique, flexible reaction to system changes in terms of product variety and possible system disturbances, and system performance improvement. AnyLogic multi-method simulation modelling platform is utilized to design and create different types of agents. The proposed simulation model results are demonstrated and verified in a case study using the configurable assembly Learning Factory (iFactory) in the Intelligent Manufacturing Systems (IMS) Center at the University of Windsor. The benefits and limitations of the proposed framework are discussed.