Semi-static Object Detection using Polygonal Maps for Safe Navigation of Industrial Robots

The collision and safety control of industrial UGVs equipped with laser range finders is often based on conservative area-oriented policies that lack in flexibility and does not deal well with non ephemeral environment changes due to semi-static objects (e.g. passive misplaced objects). In this paper, we propose a method to detect and represent semi-static objects using polygonal local maps in order to improve robot navigation. Each local map consists of polylines representing the boundary of an object detected inside a safety area. Polylines are extracted from laser scans and associated with the polylines of a reference map using a similarity measure criterion. Finally, the map is updated by merging the new polylines. The proposed polygonal representation allows the recognition of new semi-static obstacles in the environment and supports more flexible policies for safe navigation. An EKF localizer using artificial landmarks and a fixed path navigation system have been implemented to replicate the navigation system of industrial UGVs. The precision of environment reconstruction has been assessed with experiments in simulated and real environments.