Multi-arm grasp and manipulation of objects with internal degrees of freedom

The modeling and use of easily reconfigured, recursive dynamics algorithms for multi-arm systems manipulating complex articulated task objects and for which the grasps and environmental constraints may be rigid or loose are discussed. Examples are given to illustrate the use of spatial algebra algorithms in the modeling and dynamical analysis of multiple parallel manipulators grasping an articulated object. The modeling and dynamics algorithms given use a natural partitioning, based on their rate of time variation, of the multi-arm robotic system into subcomponents. With this partitioning, a large part of the dynamics algorithm remains independent of the specific nature of the task object as well as the constraints. This permits the optimization of the time-invariant parts of the algorithm, and also the use of efficient dynamics algorithms for unconstrained single manipulators