S. Darvish1, M. McNulty1, J. Pon1, H. Tallarida1, J. Moody2, S. Conant1, D.B. Thorp1

1Gonzaga University, Spokane, WA; 2XENOFI, LLC, Sheridan, WY

There are limited data on the effects of virtual reality (VR) on speed perception and motor control. Previous studies examining changes in walking and running speed in relation to visual flow speed (VFS) in 2D environments have found 15-30% underestimations in VFS compared to locomotion speed. None have studied the effects of cycling speed (CS) in a 3D, VR environment. PURPOSE: The aim was to determine if 1) subjects can match CS to VFS, and 2) if surreptitious alterations in the VFS affects CS during cycling in a VR environment. METHODS: Subjects (n=18) participated in three trials on an electronically-braked cycle ergometer, while wearing a VR headset displaying an outdoor path. VFS was specifically created to move independent of the subjects CS. During all trials, subjects were told to match their CS to VFS for the first 2 minutes and then continue cycling. During the familiarization trial, VFS was at 20 km/h continuously for 8 minutes, where the final minute served as the control (CFAM). During the two counterbalanced experimental trials (7 minutes), VFS started at 20 km/h for 3 minutes, serving as the in-trial controls, then either increased (fast trial) or decreased (slow trial) by 15% during the third minute. This experimental speed was held for the final 3 minutes. CS and heart rate (HR) were recorded every 10 seconds. A RM ANOVA analyzed the repeatability of CS between the control trials based on the order in which they were completed, CFAM, in-trial control 1 (C1) and 2 (C2). To assess any differences in CS before and after VFS was altered, a paired samples t-test compared the ratios of experimental/control speeds in the fast and slow trials. Significance set at p<0.05. RESULTS: CS during the CFAM (21.0±4.1 km/h, +5.3% compared to VFS) was significantly greater than in C1 (17.9±4.5 km/h, -10.7%) (p<0.05), while C1 and C2 (18.7±3.9 km/h, -6.5%) were not significantly different (p=0.161). There was no significant difference between fast (1.04±0.07) and slow (1.13±0.46) trial speed ratios (p=0.46). Average HR across all trials was 109±5 bpm, indicating low intensity exercise. CONCLUSION: While CS was not perfectly matched to VFS, our findings suggest that VR cycling reduces the underestimation of VFS to CS when compared to 2D running environments. Further, CS was not altered in a predictable way when VFS was altered surreptitiously.

This document is currently not available here.