Robotic Manipulation by 3D Multifingered Caging

笼形控制是一种通过机器人本体使物体无法从封闭区域内逃脱的方法。以前的研究大多只考虑平面笼，而这项研究涉及多指手的3D笼。我们将三维多指笼形划分为三个典型的类别，并推导出笼形的充分条件。利用这些充分条件，我们开发了一种机器人运动规划算法，以实现锁定。该规划算法已成功应用于实际机器人的笼形规划。

Caging is a method to make an object inescapable from a closed region by robot bodies. Most of previous studies considered only planar caging, while this study dealt with 3D caging by multifingered hands. We classified 3D multifingered caging into three typical categories and derived sufficient conditions for caging. Using the sufficient conditions, we developed a motion planning algorithm for robots to achieve caging. The planning algorithm was successfully applied to caging by an actual robot.