Visual servoing control of Baxter robot arms with obstacle avoidance using kinematic redundancy

In this paper, a visual servoing control enhanced by an obstacle avoidance strategy using kinematics redundancy has been developed and tested on Baxter robot. A Point Grey Bumblebee2 stereo camera is used to obtain the 3D point cloud of a target object, which is then utilized to manipulate the closely coupled dual arms of Baxter robot. The object tracking task allocation between two arms has been developed by identifying workspaces of the dual arms and tracing the object location in a convex hull of the workspace. By employment of a simulated artificial robot as a parallel system as well as a task switching weight factor, the actual robot is able to restore back to the natural pose smoothly in the absence of the obstacle. Two sets of experiments carried out demonstrate the effectiveness of the developed servoing control method.