Communication and cooperation in a distributed automatic guided vehicle system

Abstract

A distributed artificial intelligence (DAI) approach to controlling and scheduling automatic guided vehicles (AGVs) is presented. Each vehicle uses an intelligent agent to handle task scheduling, guidance, fault-recovery, and resource contention. The agents communicate with one another in an attempt to efficiently solve the overall problem. Each agent is a member of a contract net. The agents examine the plans of other agents in the network and determine the best path between two pickup and delivery stations. These plans minimize the probability of delay due to two or more vehicles waiting for a congested area on the shop-floor to clear. The system, composed of a network of these agents and pick-up and delivery stations, demonstrates the added efficiency and reliability of a DAI approach. >