A review and prospects of path planning methods for automated guided vehicles

Automated guided vehicles (AGVs), as key components of intelligent manufacturing and logistics systems, have their path planning technology directly impacting the efficiency and reliability of the system. This paper systematically reviews current AGV path planning methodologies, categorizing them into six paradigms: rule-based, sampling-based, search-based, optimization-based, learning-based, and local obstacle avoidance approaches. Each method is analyzed for its computational efficiency, solution quality, and applicability across scenarios. The study finds that balancing dynamic environmental adaptability, multi-objective collaborative optimization, and computational efficiency remain core challenges. Future directions emphasize real-time multi-layer planning, adaptive multi-objective optimization, cross-scenario generalization, and edge computing integration. This review provides a foundation for advancing AGV path planning technologies, enabling safer, more efficient, and more scalable autonomous operations in industrial systems.