Validation of a Fixed-Base Driving Simulator to Assess Behavioural Effects of Road Geometrics

Driving simulators need to be validated to ensure that they can be used to conduct experiments, and that collected data can be used to understand driver behaviour in real conditions. Several researchers have performed this fundamental activity by considering different driving responses (i.e., speed, acceleration, lateral position, actions on commands, etc.). With this investigation the authors sought to validate the fixed-base driving simulator at the Politecnico di Torino (Italy). Comparative and statistical approaches were used to analyse free flow speed data collected in both simulated and naturalistic driving conditions. A comparison was then made with the 15th, 50th and 85th percentiles of speed data collected at a frequency of 4 Hz. For statistical analysis, 127 sections spaced 100 m apart and characterized by different geometric characteristics were considered. Results confirm that simulator speeds are slightly higher than naturalistic driving ones. Different types of validity were obtained for different road elements. Absolute validity was achieved in 56% of cases - mainly on sharp curves, and short tangents, while relative validity is predominant in the case of shallow curves. Long tangents, intersection approaches and departures achieve almost equal levels of validity in both absolute and relative terms.