Reachability-Based Path Planning Using RRT for Construction Processes with Ensured Handover Capabilities

The use of automated processes in the construction sector has the potential to increase productivity enormously and to meet the growing demand for new housing. To efficiently utilize robots on construction sites, it is necessary to optimize their placement to minimize time-consuming repositioning. Most current path planning algorithms for component placement do not consider the reachable workspace and the kinematics of a robot. This paper presents a Rapidly-exploring Random Tree (RRT)-based algorithm that plans paths considering both reachable spaces and obstacles. To analyze the reachability a time-efficient classifier is applied, or for simple kinematics the inverse kinematics (IKs) is solved analytically. To allow the placement of components beyond the immediate reach of one robot, a handover condition is incorporated into the algorithm. The results show that the algorithm can efficiently determine feasible paths for the component's pose and placements, which is demonstrated using two example scenarios and different machines.