Comparison of sampling based motion planning algorithms specialized for robot manipulators

This paper presents a comparison on the same framework between seven different sampling based motion planning algorithms that are working in configuration (C-space) and Cartesian space and are focused on applications for robot manipulators. A planner that works in Cartesian space is implemented and compared with the others. The main result of this comparison is to illustrate the benefit as well as the drawback of each algorithm. This work examines the behavior of the algorithms in different environment conditions. Feasibility and connectivity in free and narrow passage configuration spaces, completeness and total trajectory cost(length) are examined. Moreover results with extra constraints in motion like end effector’s orientation are going to be presented. The planners are based on Rapidly exploring Random Trees (RRT). The robotic system consists of a 7 Degrees of Freedom (DoF) robotic arm mounted on the rehabilitation system FRIEND (Functional Robot arm with user fIENDly interface).