Abstract
Bicycling simulation allows for the low-risk experimental study of human factors within transportation environments. A cyclist pedals on a stationary bike trainer, which is instrumented to detect the speed of the wheel and the steering angle of the bicycle. This paper proposes a speed calibration procedure to increase the validity of the simulator results, by using an independent bicycle computer for comparing the simulator speed. The speed ratio, defined as the simulator speed divided by the bike computer speed, approaches one when the simulator is properly calibrated. The effect of tire pressure was analyzed by examining the speed ratio for various tire pressures. The optimal tire pressure was selected as the one that provided a speed ratio closest to one when all other factors were held constant. In the final calibration, a gain factor was used to modify the simulator speed calculation that was embedded in the simulator’s bicycle dynamics model. Following calibration, the final simulation speed was within 99.5% of the bicycle computer speed, indicating that the physical speed of the wheel was accurately modeled in the simulation environment. The calibration procedure uses general equations and techniques that can be applied to other bicycling simulators to calibrate speed measurements and improve the consistency of experimental data worldwide.
In the field of driving and bicycling simulation, simulator sickness has been shown to have a negative impact on driver performance. High latency, where the amount of time between operator input and the response of the visual field are mismatched, is correlated with higher rates of simulator sickness. The Oregon State University Bicycling Simulator was used to develop a framework for evaluating visual latency experienced in a bicycling simulator. A cyclist biked along a bike lane, sharply steered away from an obstacle, then countersteered to return to the bike lane. A relative measure of the steering angle was estimated with video data and was compared to the absolute measurement provided by the simulator. A cross-correlation technique identified a consistent 6 time step lag between the two measurements that represents the sampling rate of the steering cradle. During steady state steering, the delta steering angle approaches zero. Finally, the mean steering latency was found to be 115 milliseconds, with a median around 69 ms. The study provides framework for transportation researchers to measure steering latency which could be used to minimize the mismatch between the user’s control of the system and the response of the visual simulation.