Event-based dynamic bandwidth management for teleoperation

Abstract

In real time network based teleoperation, different tasks or events requires different feedback resolutions and thus different rates of feedback are needed for reliable and efficient task completion. For example, in navigation if obstacles are distant then rate of video feedback can be reduced versus when the vehicle is approaching an obstacle. In order to efficiently teleoperate mobile robots, dynamic bandwidth management of feedback data is necessary for efficient use of networking resources. In this work, we introduce a bandwidth management algorithm based on the task conditions. In other words, if nothing (something) of relevance to the task is occurring in the environment the rate of feedback can be decreased (increased). The algorithm was validated experimentally teleoperating a mobile robot using a haptic device while providing visual and haptic feedback. Results demonstrate that comparable performance can be accomplished using dynamic bandwidth allocation using lower network recourses.