Prioritized Planning for Large-Scale Multiple-AGV Scheduling Problem in Smart Manufacturing

Robotics and automation is one of crucial trend in smart manufacturing to improve production efficiency. Au-tomated guided vehicles (AGVs) are a type of mobile robot used for material handling and have become widely utilized to achieve transportation automation. The usage of multiple AGVs introduces potential risks, such as traffic conflicts and safety risk. To meet high production demands, numerous shop floors are set up for large-scale manufacturing. Thus, reasonable and efficient AGV scheduling is vital for real-world operations. This paper proposes an efficient and scalable prioritized planning algorithm for large-scale multiple-AGV scheduling problem in manufacturing. The algorithm sequentially addresses two primary sub-problems: job assignment and conflict-free routing. The results of job assignment dictate the routes taken by the AGVs. In job assignment, jobs are allocated sequentially based on their pickup times. In conflict-free routing, AGV priorities are predefined, ensuring that higher priority AGVs maintain their movement while adjustments are made only to lower priority AGV plans when conflicts arise. Simulation is conducted on two real shop floor layouts and demonstrates the effectiveness and high efficiency of proposed algorithm. Even in a large-scale layout with 500 jobs and 20 AGVs, the computation time is only around 21 seconds.