Knowledge-based Multi-site Capacity Planning using Agents

Because of current strong globalization trend, manufacturers are more willing to go to regional or globalism by setting-up plants closer to customers in order to reduce operation costs. Accordingly, mid-term capacity planning paradigm has shifted from algorithm-based solutions for single production site to collaboration-oriented planning in multiple production sites. This paper presents an agent-based multi-site capacity planning framework for autonomous plants within an organization by exploiting the features of agents (autonomy, social ability, re-, and pro-activeness). The creation of a "community of practice", where explicit knowledge in knowledge base and tacit knowledge are shared among systems and people, will facilitate the knowledge flow across department walls and plants' boundaries, supporting the generation of "global optimal" capacity plans. This paper also demonstrates quantitative benefits in reduction of tardy jobs in multi-site capacity planning through a simulation.