Virtual Reality Bicycle with Data-Driven Vibrotactile Responses from Road Surface Textures
Ruslan Rakhmatov, Arsen Abdulali, Waseem Hassan, Minji Kim, Seokhee Jeon · 2018
In this paper, we proposed a data-driven vibrotactile rendering system for indoor exercise bicycles. The input of the data-driven model is described by a two-dimensional vector of cycling velocity and tire pressure. We designed a data-collection bike that captures vibrations induced at the handlebar and the cycling velocity for different tire pressures. The data were collected from four real cycling pavements. The level of the tire pressure was varied according to subjects' weight. Data-driven input-output based haptic modeling was adapted for generating road surface texture models. A rendering system was also developed and integrated into the stationary bicycle, and four virtual textures were evaluated by experienced cyclists. The results show that the participants were able to successfully discriminate and identify simulated virtual road surfaces.