The main aim of the project was to provide typical values of reaction behavior in safety-critical situations in traffic. Additionally, test conditions were to be defined in order to enable a comparable examination of reaction behavior by other researchers. The results can be used to design and evaluate advanced driver assistance systems in standardized conditions. To this aim three driving simulator studies were conducted using a static simulator. Each of the studies used a different context (town: 50 km/h, rural road: 100 km/h, highway 130 km/h). For validation purposes a comparable study with a real car was done for the town context. In each of the studies the influence of the surrounding, expectations of the driver and of cognitive distraction was examined with regard to the type of reaction (steering, braking, combined reaction) and reaction times. N = 131 participants were between 20 to 40 years of age. About half of them were female. In the critical situations, either a pedestrian or a standing car suddenly appeared in some distance of the own, moving car. In an additional situation a lead car suddenly braked. Comparing simulator and real driving showed a high validity of the simulator results. The type of reaction depended on the time left to react and the available space. Brake reaction times were between 1.0 s (town) and 1.3 s (highway). Steering reaction times were between 0.7s (town) and 1.1s (highway). Reaction time during car following were about 0.2-0.3 s slower. There was no effect of cognitive distraction. However, there were strong learning effects leading to reaction times which were between 0.2-0.4s shorter. From the results, a set of 9 situations is defined which can be used to examine different reaction types and reaction times depending on relevant situational characteristics. (Author/publisher)
Abstract