Path Planning for AGVs: Balancing Computational Efficiency and Optimality

Automated Guided Vehicles (AGVs) play a crucial role in industrial automation, streamlining material handling and logistics operations. This paper investigates the performance of eleven classical pathfinding algorithms—including Dijkstra, Depth First Search (DFS), Floyd Warshall, Binary Search, and Prim’s MST on an Ackermann steered electric vehicle (EV) based AGV navigating six charging stations in a simulated urban environment. Algorithms are classified into By Station and By Edge , with implementations adapted to meet their specific routing requirements. AGV performance is evaluated using conventional metrics runtime, travel distance, cost, speed, and station visitation alongside a novel Energy Consumption Score (ECS) that balances operational efficiency with computational overhead. Results show that classical algorithms can efficiently handle static, predictable networks, though some prioritize cost over station coverage. ECS provides a unified framework to rank algorithm suitability. The obtained results shows, Based on the tested algorithms: Floyd Warshall achieved the best overall performance in terms of minimal cost and ECS. For maximizing station coverage, TSP and Prim’s MST were the most effective. Edge-based algorithms proved reliable for navigating predefined routes but were generally less efficient in terms of overall cost and energy consumption.Future work will explore heuristic and hybrid approaches, dynamic networks, multi AGV coordination, and energy aware routing for improved urban path planning.