A Study on Object Recognition for Autonomous Mobile Robot’s Safety Using Multiple Sensors

Autonomous Mobile Robots (AMRs) play an important role in Industry 4.0, especially after the COVID-19 pandemic. AMR systems provide an efficient means of material transportation within assembly lines, leading to a reduction in reliance on human labor, heightened production efficiency, and an enhanced safety environment for employees. Consequently, ensuring the safety of AMRs during their operations becomes paramount, and this is attainable through the incorporation of the obstacle avoidance and detection systems on these vehicles. This study evaluates three distinct methods for obstacle recognition: utilizing LIDAR sensors, 3D cameras, and a combination of both LIDAR and 3D cameras. The aim is to enhance both the cost-effectiveness and performance metrics of the safety system employed in AMRs. The experiments were conducted using the AMR system developed by Phenikaa-X Company, affirming the effectiveness of the newly integrated safety system in obstacle detection and collision prevention. The experimental results demonstrate that the combination of LIDAR and a 3D camera is the most promising option, compared to using only LIDAR or 3D camera. This approach achieves 94% precision rate in object detection even while the AMR is in motion. The AMR requipped with both LIDAR and a 3D camera exhibits the ability to promptly detect obstacles and come to a stop at a considerable distance from them. This substantial improvement significantly enhances the overall safety of the AMR system.