Local Path-Planning Optimisation for an Industrial Autonomous Mobile Robot via Dynamic Obstacle Detection

Over the last decade, there has been a substantial surge in the demand for Autonomous Mobile Robots (AMRs) owing to their proficiency in navigating complex environments without the use of physical or electromechanical guidance devices. In particular, the exploration of obstacle avoidance holds paramount importance in AMR research, particularly as these robots are intended to function in settings with the presence of humans, vehicles, or other robots, necessitating the safe evasion of both stationary and moving obstacles. Within this context, this paper presents a solution for navigating in dynamic environments that can be easily integrated into the current ROS2 Navigation Stack (Nav2). The proposed solution is based on a Laser Imaging Detection and Ranging (LiDAR) sensor-based system, which allows for the detection of dynamic obstacles and estimating their velocities using a linear Kalman filter. The obtained information was integrated into the navigation system and the proposed solution was then validated through a campaign of simulation tests on an AMR prototype developed by HSG Engineering Srl. Simulations showed that this architecture guarantees safer navigation and represents a solid starting point for developing the robot’s navigation for industrial use.