Weight lifting motion planning for a Puma 762 robot

In this paper we develop a point-to-point weight lifting motion planner for open-chained robots. The joint trajectories are defined by B-spline polynomials along with a time-scale factor. Physical limitations of a Puma 762 robot are incorporated into the formulation. The torque limits are formulated as a penalty function (soft constraints) added into the objective function while the position and velocity limits are formulated as linear inequalities (hard constraints). The problem is solved by a gradient-based optimization method in which the gradient is determined analytically. Not only does the optimal path extend the payload capability, but also reduces the joint torques by passing through singular configurations.