Robotic Stacking of Irregular Objects with Load Position Identification and Compensation

Autonomous facility construction is a challenging task in the field of robotic manipulation, which is faced with the uncertainty of operating objects and the construction environment. This paper proposes a set of algorithms for the autonomous stacking of irregular objects by robots, including size recognition of objects, position identification and compensation of load using a force sensor, and stacking position detection. The method effectively compensates for the uncertainty in the position of the object relative to the gripper after grasping. We have performed stacking experiments on the robot. For four randomly placed irregular stones in the sand, the robot autonomously and successfully completed a stable positive pyramidal stacking. Our experiments demonstrate the feasibility of the proposed method.