Using multi-agent simulation and learning to design new business processes

Business process modeling, analysis, and then redesign are the central task in the efforts of re-engineering business processes. Frequently reviewing and promptly changing business processes to adapt to new business environment is the key to maintain agility under the competitive global market. It is imperative to enhance the adaptability of business processes. This paper proposes a multi-agent information system based on Swarm, a multi-agent simulation platform, to simulate business processes and incorporate reinforcement learning to obtain better process adaptability. The resulting system, called BPSLS, is elaborated and evaluated by the order fulfilment process in supply chain networks.