A new decision making approach for the formation of holonic agent-based manufacturing networks

The research paper presents the methodology used to develop decision algorithms for order holons. Two algorithms are presented in this paper. The algorithms model an actual job allocation system and a conceptual one for the formation of networks of manufacturers. The first algorithm is derived from an industrial case study. The second algorithm exploits the use of multiple flexibility parameters to improve the performance of the job allocation system on quality, cost and delivery time. Multi-agent system (MAS) and discrete event simulation constitute the methodology which allows the two algorithms to be run in a verified and validated environment. The results of the simulation show that the second algorithm outperforms the first algorithm on network conformance reliability, production cost and order delivery times. Eventually, the methodology will be developed to test the algorithms in the market conditions of a volatile environment.