Dextrous haptic interaction in virtual environments: human performance evaluations

Abstract

In this paper, we report two human factors studies performed to investigate human performance in tasks involving dextrous manipulations of virtual objects with haptic feedback. These studies were performed at the Man-Machine Interface Laboratory (Rutgers University, NJ, USA) and at the Laboratoire de Robotique de Paris (CRIIF-LRP). Different haptic devices were used : the Rutgers Master, the LRP Force Feedback Glove (LRP), and the ITR device (Interface Technology Research, UK). These experiments involved (1) accurate placements of virtual objects and (2) grasping-force regulation. In the former case, we compared the LRP FFG (that provides complex haptic cueing) and the ITR device(that provides binary haptic cueing : grasped/not grasped). In the latter case, we investigated the effect of sensory substitution (synesthetic representation of haptic informations) to convey virtual haptic cues using the visual and auditory sensory channels. Results of the experiments showed that (1) information redundancy (display of visual or auditory force feedback in addition to finger force feedback) allowed to increase performance compared to finger force feedback alone, and (2) simple tactual feedback (provided by the ITR device) allowed similar performance than complex force feedback (provided by the LRP FFG). This result led us to think that complex force feedback cues are not required in simple manipulation task (i.e. when grasping regulation forces is not crucial). In this case binary haptic information could be sufficient.