Experimental Real-Time Setup for Vision Driven Hand-Over with a Collaborative Robot

Collaborative robotics aims to make possible a close collaboration between human operators and robots in the industry scenario. To achieve this goal, the robot must be able to adapt its behavior to the movements of human co-worker. This work presents a control strategy that permits to drive a UR3 collaborative robot in order to perform a hand-over task with a human operator. The movements of the operator are acquired by a 3D vision system made of multiple Microsoft Kinect sensors, so to prevent possible occlusions related to the presence of the robot or dynamic obstacles in the field of view of the sensors. Thus, the human skeletons extracted from depth sensors are handled by an algorithm that generates an optimized skeleton. A path-planning algorithm uses the skeleton information to calculates the robot joints velocities to accomplish the task. Experimental tests have been conducted to verify the effectiveness of the hand-over strategy here proposed. The hardware setup prepared for the experimental phase of the work and the results obtained from the tests are presented and discussed.