Development of a Shared Controller for Obstacle Avoidance in a Teleoperation System
- PDF / 3,811,966 Bytes
- 9 Pages / 594.77 x 793.026 pts Page_size
- 92 Downloads / 182 Views
ISSN:1598-6446 eISSN:2005-4092 http://www.springer.com/12555
Development of a Shared Controller for Obstacle Avoidance in a Teleoperation System JiWoong Han, Kyunghwan Cho, Inhoon Jang, Chanyoung Ju, Hyoung Il Son, and Gi-Hun Yang* Abstract: Several methods have been investigated to increase the efficiency of the operator in teleoperation, but remote devices still cannot be operated efficiently in the presence of the obstacle. In this study, a virtual link and virtual joints were created within the end-effector of the slave robot, and a shared controller was designed to implement an effective obstacle avoidance algorithm for the remote control system. Teleoperation experiments were conducted to verify the algorithm. Completion time and the NASA Task Load Index (NASA-TLX) were measured to evaluate the improvement of teleoperator work efficiency. When the obstacle avoidance algorithm was used, completion time decreased by 8.64 %, and the average NASA-TLX decreased by 30.33 % as compared without the algorithm. Our method effectively improved completion time and NASA-TLX scores for both skilled and nonskilled human-operators. Keywords: Obstacle avoidance, object detection, shared control, teleoperation.
1.
INTRODUCTION
In master-slave teleoperation, the operator controls the master device in a constrained environment where the teleoperator cannot see the slave robot directly. Depending on the number of cameras installed and the angles involved, operators may not get sufficient on-site information, so as many as 30-50 % of the direct manipulation results are adversely affected [1]. To overcome limited visibility and improve teleoperation task efficiency, haptic interaction is typically introduced into a teleoperation control system [1–3]. However, when there is an obstacle within the slave robot workspace, the constraints of teleoperation make it difficult for the operator to accurately perceive the distance between the obstacle and the slave robot. Therefore, control of the slave robot can impose a heavy workload on the teleoperator. In the worst case, when the obstacle is not visible on the camera screen, the slave robot may collide with the obstacle. Many obstacle avoidance algorithms have been developed. The potential field is widely used as a method for avoiding obstacles [4–6]. Repulsive potential fields near the obstacle and attractive potential fields near the target are generated. Potential fields given by the sum of repul-
sive and attractive fields are used for path planning from the initial position to the goal position. Furthermore, a model predictive control has been used to overcome disadvantages of potential field [7–10]. However, such path planning is not possible in a remote-control system as the desired position of the slave robot changes in real time with the movement of the master device. Virtual fixtures can help the teleoperator guide the slave robot towards a goal position [11–16]. The virtual fixture is applied to the master device with hybrid positionvelocity control to assist the operator
Data Loading...
