Cooperative manipulation with least number of robots via robust caging

One problem in multi-robot cooperative manipulation is redundancy. Too many robots are waste of hardware and increase control complexity. This paper solves the problem of redundancy by robust caging. Robust caging calculates caging positions from translational immobilization with respect to translational constraints and rotational constraints. On the one hand, robust caging helps to reduce the necessary number of robots in cooperation. On the other hand, the initial positions of necessary robots in robust caging are optimized to offer large robustness to control errors. Our proposal with robust caging is implemented to transport target objects over slopes. The algorithm can choose least number of robots with respect to shape of target objects and requirements of robustness. At the same time, each robot may endure as much as 256 ms time step and 1 cm control error, showing the superiority of robust caging.