Improvement of Intention Estimation Method for Leader-Follower Robot

A leader-follower robot has the feature of remote operation so that it can be applied to a work environment that is dangerous to its operator. However, there still is a risk that the operator of this robot may make an erroneous operation due to many reasons, such as fatigue. Based on the feature of remote operation, the follower robot will faithfully perform the erroneous operation, which may be dangerous. In this research, we are aiming to develop a method that can suppress the effect of erroneous operation by estimating the operator’s correct intention and using the correct intention to operate the leader-follower robot. In this report, we list and evaluate the effect of the previous intention estimation method, then carry out a method to improve the previous intention estimation method by applying a Kalman filter.