An Intelligent Process Planning Concept Using an Agent-based Approach

A concept of a cooperative multi-agent architecture is proposed to develop an intelligent process planning system. Traditional process planning is often suffers from limitations such as lack of integration between design and manufacturing operations, limited geometry to which parts can be processed, non-availability of important functional modules like tool and fixture selection, inability to generate appropriate sequence of operations for a specific part, and the rigidity, complexity, and high expense of the hierarchical structure of the systems as the systems increase in size. Using cooperative distributed problem-solving techniques, our proposed system architecture is based on cooperating planning agents, decentralized according to the organizational model of a typical process planning department. In the model the scope of decision making is kept as localized as possible, and the level of details of the decisions has been balanced with the level in the organization. The architecture comprises of three types of cooperating agents – the job processing planning manager agent, the work center manager agents, and the work center knowledge expert agents. As explained in the scheme of interactions among the agents, it is expected that our proposed architecture can improve the process planning performance in a dynamic environment.