LAUSR

Abstract The paper reports development of the multiagent-based framework for distributed manufacturing process planning. The research builds on previous results in developing intelligent manufacturing planning platform and extends it into agent-based modeling. Both, design and manufacturing tasks are defined as sequences of steps to be performed by highly intelligent autonomous agents. Autonomous agents are classified into two categories: task agents, which are responsible for completing design or manufacturing tasks, and service agents, which provide services for individual steps of those tasks. For multiagent implementation, we have chosen JADE framework and both task agents and service agents have been implemented by extending agents from JADE. Thus, the JADE’s message passing mechanism has been adopted for inter-agent communication and appropriate ontologies for the message passing have been built from planning classes and objects. The brief overview of IMPlanner functionalities has been provided and development and implementation of service agents for several typical manufacturing planning tasks have been explained. The proposed framework has been executed in the context of the simulation model for evaluation of dynamic routing policies.