Autonomous Robotic System for On-site Construction Waste Assessment and Separation in Wuhan - 3Pager

This demonstration was conducted at the Longhu Binjiang Tianjie construction site in Wuhan, China to validate an autonomous robotic system for on-site construction waste assessment and sorting. The system was deployed in the construction environment to perform autonomous navigation, environmental perception, and waste identification. The main data types used included 3D point cloud data from LiDAR sensors and RGB-D image data from visual cameras, which supported environmental mapping and construction waste recognition. The system was developed using the ROS robotic operating system and integrated LiDAR sensors, RGB-D cameras, YOLO-based object detection for waste recognition, and RTAB-Map SLAM for mapping and localization. During the implementation, several challenges were addressed, including the complexity of construction site environments, mixed waste materials, and the need for reliable perception and navigation in dynamic conditions. The results show that the robotic system can support automated waste inspection and reduce dependence onmanual operations. The collected data can contribute to the Circular Potential Information Model (CP-IM) within the REINCARNATE project by providing digital information on waste categories and spatial data for material tracking and circularity analysis. The demonstration also supports cross-work-package collaboration by linking robotic sensing technologies with broader project activities on digital construction, circular material management, and knowledge dissemination.