IFC-based generation of semantic obstacle maps for autonomous robotic systems

Abstract: Autonomous Robotic Systems (ARSs) in the construction industry usually have to perform preliminary mapping of construction environments before deployment. For large and complex sites, this can be time-consuming. With Building Information Modeling (BIM), a lot of information is already available about sites. This study proposes a method to make that information available to ARSs to streamline autonomous tasks and remove the need for mapping. This is achieved by automatically generating semantic and color-coded obstacle maps from IFC files. The results are obstacle maps that can be used for autonomous navigation that remove the need for mapping.