Effect of local-adaptive haptic guidance on a path-following task

Abstract

Haptic guidance is a promising robot-mediated training method that assists humans in completing tasks. Although several research studies showed positive effects of haptic guidance, a long-standing assumption indicated that haptic guidance impairs motor learning owing to conflicts between human and robots. In this paper, we propose local-adaptive haptic guidance, which provides varied guiding forces in different localities depending on a human's local performance. With this adaptive training method, we hypothesized that trainees learn unfamiliar tasks with less conflict and become less passive. Three types of guidance training method were compared: i) no guidance, ii) (fixed) haptic guidance, and iii) local-adaptive haptic guidance. We observed a trainee's learning performance while the trainee repeatedly performs a path-following task from three types of guidance. Local-adaptive haptic guidance shows the similar positive effect to haptic guidance in terms of reducing completion time. Moreover, the trainee with local-adaptive haptic guidance perform the task more smoothly than the trainee with haptic guidance. This research provides suggestions for the way to reduce the hampering effect of haptic guidance.