Dynamic task assignment of autonomous AGV system based on multi agent architecture

Increasing number of worldwide researchers and practitioners within manufacturing industry had recognized the need for an organization to implement distributed manufacturing system to be flexible and adaptable to a more demanding and fluctuating market. This paper proposes distributed architecture to control Automated Guided Vehicle (AGV) operation in manufacturing industry based on multi- agent system (MAS). System and agent architectures had been designed to enable control the material handling activities. All agents are equipped with decision making capability to plan and execute their responsibilities autonomously or collectively when needed. Additionally, improvement had been made to FIPA Contract Net Protocol (CNP) to support dynamic attributes of AGV task assignment mechanism.