Goal Predictive Infused Robot Teleoperation with Kinect Depth Camera
thesisposted on 17.02.2017 by Christopher Schultz
In order to distinguish essays and pre-prints from academic theses, we have a separate category. These are often much longer text based documents than a paper.
Using depth camera technology developed in recent years, a human's pose demonstrations can be used as inputs into unilateral robotic teleoperation. This teleoperation system provides an intuitive and effective means of control for the human operator. However, the imprecision of low-cost depth cameras and difficulties with the frame of reference for the human operator introduce inefficiencies in the teleoperation process when performing tasks that require precise robotic interactions with the robot's work space. We developed a goal-predictive teleoperation system that addresses these difficulties for the human operator by adding goal-directed aid to the teleoperation control process. Our approach used inverse optimal control to predict the intended final state of the robotic system from the current motion trajectory in real time and then adapted the degree of autonomy between the operator's demonstrations and autonomous completion of the predicted task. We evaluated our approach by using a Microsoft Kinect depth camera as an input sensor to control a Rethink Robotics Baxter robot. The results verify the effectiveness of our developed goal-predictive teleoperation system.