Evaluating the effects of cross-sectional roadway design elements and the impact on driver performance using a driving simulator

The combination between roadway design and driver performance has long been at the forefront of creating a safe driving environment. This research initiative explored the relationship between the cross-sectional design elements and the impact on selected driver attributes such as speed profiles and lateral positioning. In this experiment a traditional collector-type roadway was modeled using a fixed-base driving simulator. The existing design, otherwise known as the base scenario, was subsequently reconfigured with four different cross-sectional designs within the same physical right-of-way. Specific alternative design elements included: narrow travel lanes, addition of bicycle lanes, raised center median, and curvilinear roadway profile. Twenty participants completed five simulation scenarios and their respective speed and lateral positioning were recorded throughout each drive. These performance measures were analyzed by comparing data at five controlled checkpoints within the drive. Experimental results were evaluated using descriptive and inferential statistical tests. Results show that the mean participants’ speed was higher than the posted speed limit in all scenarios; however, the speeds at each checkpoint were consistent with the speed data collected from the field. There was no statistically significant difference in speeds between all scenarios except for the curvilinear profile scenario. In the curvilinear profile scenario, the difference in collected mean speeds was statistically significant compared to the base scenario. The difference in lateral positioning between each scenario across the five controlled checkpoints was statistically significant. Overall, there is evidence to suggest that the adaptation of narrow lanes, inclusion of bicycle lanes, or addition of raised medians, will have only a minimal influence on speed reduction in a controlled driving simulator environment.