Predictive Teleoperation using Laser Rangefinder

Presence of delay in communication channels degrades performance and stability in bilateral teleoperation controllers. One method to remedy this problem is the use of predictive controllers in which the slave and environment dynamics are modelled and rendered locally at master to reproduce the contact information and bypass the delayed feedback loop. A major challenge in these controllers is the agility of local model adaptation to the changes in environment dynamics and its synchronization with slave contact events. In this paper, a novel collision prediction method using laser rangefinder is introduced. The proposed method is employed to synchronize the master local model and slave control events in a force position teleoperation architecture. Moreover, this method speeds up environment identification algorithm. The proposed predictive controller is evaluated experimentally on a one degree of freedom (DOF) master-slave teleoperation setup and its performance is compared to that of a conventional force-position controller