Intelligent “Wingman”: Enhancing Visual Awareness in Robotic Construction Processes

Le résumé fourni par la source

Visual awareness plays a critical role in robotic construction. However, it is challenging to achieve comprehensive visual awareness throughout the construction processes with robotic onboard cameras or environmental cameras because of the large-scale workspaces and objects, as well as frequent occlusions caused by site objects. To address this issue, this research introduces a mobile supervising robot equipped with a camera mounted on a manipulator as the intelligent “wingman.” Specifically, the supervising robot adaptively adjusts its poses and manipulates its camera to obtain visual information of the Region of Interest (ROI) during robotic construction processes. A two-step camera view selection algorithm is proposed to optimize the camera view for the supervising robot. To validate the feasibility of the proposed approach, a rough carpentry (framing) task is carried out as the case study. The supervising robot demonstrates the capability to automatically adjust its position and the camera’s pose to gain an optimal view for the ongoing stage of wooden frame construction. This research has the potential to improve the quality and safety of robotic construction work while reducing the cost and effort required to set up camera systems on-site manually.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.