An experimental study of perceived instability during haptic texture rendering: effects of collision detection algorithm

This paper presents a quantitative characterization of the instability that a human user often experiences while interacting with a virtual textured surface rendered with a force-reflecting haptic interface. Psychophysical experiments were conducted to measure the maximum stiffness under which virtual textured surfaces were perceived to be stable in a variety of conditions differing in texture model parameters, rendering method, and exploration mode. Unlike our previous study that used a collision detection algorithm with an inherent step change in force magnitude near the textured surface boundary, these experiments used an algorithm proposed by Ho et al. (1999) that produces continuously changing force magnitudes at the cost of increased computational complexity. We found that the stiffness thresholds resulted from the collision detection algorithm of Ho et al. were not always higher than those obtained with the algorithm used in (Choi and Tan, 2002). The stiffness thresholds depended on the texture rendering method and the exploration mode used in the experiments. We also discuss the types of instability experienced by the subjects and the corresponding characteristics of the proximal stimuli that invoked the perception of each type of instability. With this knowledge, our future work will investigate techniques to mitigate the problem of perceived instability.