Minimizing multi-stage system variation using Kalman filter for a tele-sandblasting system

In teleoperation systems which generally require human-robot collaboration to operate the tasks, task performances suffer from uncontrollable factors, i.e., human skills, time delay, and robot movement errors. To incorporate with this situation, the system must be designed to withstand those factors' effects. Recently, system robustness has become an essential issue for designing manufacturing systems. In this paper, we present a robust design to control the variation for multi-stage tele-sandblasting system. The objective is to locate the sandblasting nozzle at desired positions while minimizing system variation. A multi-stage variation estimation model with Kalman filter is proposed to estimate and control the system variation over time. The results show that our approach significantly decreases tele-sand blasting system variation.