BIM-based semantic building world modeling for robot task planning and execution in built environments

From its internal representation of a building, or, its building world, a robot can retrieve a building's information for task planning and execution. However, a building is a complex construct composed of thousands of sub-elements, therefore manually programming the complete lifecycle information in a robot can be impractical. Our study presents the creation of a semantic building world from a building information model (BIM) that can be used for robot operations. To describe static and dynamic elements, we create a Universal Robot Description Format (URDF) building world using the Industry Foundation Classes (IFC) that a robot can directly query information for task planning and execution. As the case study demonstrates, our study bridges the gap between BIM's lifecycle information and robot operations. By allowing robots to acquire information about operating environments, the proposed methodologies provide the foundation for studies attempting task robotization in different types of facilities throughout their lifecycles.