Driving simulator validation for speed research on curves of two-lane rural roads

Abstract

The numerous advantages of driving simulators make them a useful research tool. However, validation studies need to be conducted before extrapolating driving simulator results from the laboratory to a real-world driving environment. In this study, validation of mean speeds on curves obtained in a driving simulator was performed by monitoring speed on corresponding real-world curves. A mountainous two-lane rural road in Iran was simulated and the driving speeds of 30 participants were collected at the beginning, middle and end of five curves in the simulator. The driving speeds of 40 different participants were also measured at specified points of five curves in the real road using a radar gun. The curves studied included a horizontal curve on a positive grade, three combined crest curves and one combined sag curve. The results of bilateral z-tests indicated that both experiments belonged to the same population and independent t-tests indicated equality of variances for all curves. In most curves, except the combined crest curve with a radius of 300 m, the mean speeds in the driving simulator were higher than the real-road condition. The trends in speed variation from the beginning to the end of the curve were similar in both experiments.