Hierarchical Bayesian modeling to evaluate the impacts of intelligent speed adaptation considering individuals’ usual speeding tendencies: A correlated random parameters approach

Abstract

Although there have been a large number of short- and long-term field trials and driving simulator-based studies on intelligent speed adaptation (ISA), there are a limited number of methods for evaluating the impact of ISA. In particular, appropriately evaluating the impacts of ISA through field trials requires the consideration of a number of factors related to detailed road characteristics. In addition, because of the “target speed” features of ISA itself, it is necessary to consider the usual speeding tendencies of individual drivers to avoid underestimating the impact of the ISA. In this study, a hierarchical Bayesian model with correlated random effects was developed to examine the impact of ISA on driver speed based on the driver’s typical speeding tendency and road characteristic conditions as confounding factors. The model was applied to clarify and compare the impacts of informative and incentive ISAs on community streets with a 30 km/h speed limit (“Zone 30”) based on data collected in a field trial. The modeled effects of many road characteristics were found to match intuitive expectations, suggesting that they were well-controlled in our assessment of the ISAs’ impacts. It was also confirmed that the incentive ISA had a large speed reduction effect on the behavior of drivers who tended to speed, while the informative ISA did not. In particular, although the impact of the incentive ISA on all drivers was only 2 km/h greater on average than that of the informative ISA, the impact of the incentive ISA on drivers who tended to speed was 7 km/h greater than that of the informative ISA when correlated random effects were considered.