Multi-Agent Simulation for Scheduling and Path Planning of Autonomous Intelligent Vehicles

Autonomous and Guided Vehicles (AGVs) have long been
employed in material handling but necessitate significant investments, such as designating specific movement areas. As an alternative, Autonomous and Intelligent Vehicles (AIVs) have gained traction due to their adaptability, intelligence, and capability to handle unexpected obstacles. Yet, challenges like optimizing scheduling and path planning, and managing routing conflicts persist. This study introduces a simulator tailored for AIV scheduling and path planning in various production systems. The simulator supports both predictive, where paths are predetermined, and dynamic scheduling, with real-time optimization. Paths are determined using Dijkstra’s method, ensuring AIVs use the shortest route. When path-sharing conflicts arise, a multi-criteria priority system comes into play, and its impact on the makespan is assessed. Experimental results highlight the advantage of AIVs over AGVs in most scenarios and the simulator’s efficiency in generating effective schedules, incorporating the priority management system.