FOV-based Optimal Positioning and Path Planning of Quadruped Robot for Construction Site Monitoring

This study proposes a methodology for developing optimal positioning and path planning of a quadruped robot, addressing the operational inefficiency caused by unplanned exploration and redundant movement during 3D spatial data acquisition at construction sites.A closed-form expression for the minimum shooting distance that guarantees full capture of a target within the camera frame was derived from the width and height of the target, the camera field of view, and the minimum safety clearance of the robot.These positions were then treated as waypoints in a multi-waypoint path optimization scheme combining the A* algorithm with the dynamic programming-based Held-Karp algorithm.The methodology was pre-validated in a BIM-based Unity simulation and demonstrated indoors at a university building using a quadruped robot (Unitree Go2).The experimental results demonstrated that, in this case study, the proposed approach achieved improvements in travel distance, travel time, and re-planning frequency compared to the built-in autonomous navigation during the 3D mapping task, and the route generated by the proposed algorithm matched the most efficient path among candidate visiting orders in the multi-waypoint shooting task.