Measuring driver responses at railway level crossings

► We evaluate driver responses towards three conventional warning devices for level crossings. ► We use two methods to collect driver responses data: field video recording and driving simulator experiment. ► Driver responses to passive crossings are less encouraging compared to active ones. ► Field results are consistent with simulator results. ► Hence, field results may be used to calibrate the simulator for further evaluation of alternative warning devices. Railway level crossings are amongst the most complex of road safety control systems, due to the conflicts between road vehicles and rail infrastructure, trains and train operations. Driver behaviour at railway crossings is the major collision factor. The main objective of the present paper was to evaluate the existing conventional warning devices in relation to driver behaviour. The common conventional warning devices in Australia are a stop sign (passive), flashing lights and a half boom-barrier with flashing lights (active). The data were collected using two approaches, namely: field video recordings at selected sites and a driving simulator in a laboratory. This paper describes and compares the driver response results from both the field survey and the driving simulator. The conclusion drawn is that different types of warning systems resulted in varying driver responses at crossings. The results showed that on average driver responses to passive crossings were poor when compared to active ones. The field results were consistent with the simulator results for the existing conventional warning devices and hence they may be used to calibrate the simulator for further evaluation of alternative warning systems.