An Efficient Simulation Framework for <scp>2D LiDAR</scp> ‐Based Navigation and Collision Detection of Autonomous Ground Vehicles to Enhance Safety in Warehouses

ABSTRACT The increasing use of autonomous ground vehicles (AGVs) in warehouse automation necessitates effective and reliable collision detection systems to guarantee operational safety. The conventional real‐world evaluation of collision avoidance systems is resource‐demanding and hazardous, highlighting the necessity for dependable simulation tools. This article introduces an efficient simulation framework specifically developed for 2D LiDAR‐based collision detection, aimed at improving the safety performance of AGVs in warehouse settings. The proposed system simulates a realistic warehouse floor environment with multiple passages separated by obstacles. The framework's foundation leverages the cost‐effectiveness and simplicity of 2D LiDAR, employs ray‐casting methods and spatial filtering algorithms to accurately replicate 2D LiDAR scans and identify probable collisions with minimal processing expense. The simulation results indicate obstacle detection accuracy is very precise, using 2D LiDAR. This article offers a simulation environment for the immediate design and testing of 2D LiDAR‐based collision detection, significantly reducing the need for physical trials during the first development phases. The framework serves as a crucial tool for researchers and engineers aiming to improve AGV safety protocols in warehouse automation.