Target prediction to improve human errors in robot teleoperation system

Abstract

In robot teleoperation system, human operates and respond rely on what the video screen displays. However, unavoidable delays, such as network transmission, cause asynchronous communication between human and robot and further result in human operation errors, such as redundant or duplicate operations or even wrong commands. Different from users in large-scale industry, users in daily life may not be experts or even trained to use the robots so there are new challenges for developing home-used robots. In this paper, we design a module-based remote control system to not only illustrate the possible human operation errors caused by network delay but also provide a user-friendly framework for simulating solutions to improve human errors. We then proposed two target prediction methods, Target Area Predictor (TAP) and Stop Position Predictor (SPP), to decrease these human errors. The experiment results show that TAP and SPP effectively improve the performance by reducing 60% task completion time and 36% contour errors.