Assessing the travel impacts of subnetworks for automated driving: An exploratory study. Issue 1 (March 2019)

• Record Type: Journal Article
• Authors: Madadi, Bahman; van Nes, Rob; Snelder, Maaike; van Arem, Bart
• Is Part Of: Case studies on transport policy. Volume 7:Issue 1(2019)
• Page Start: 48
• Page End: 56
• Publication Date: 2019-03
• Subjects: Automated driving impacts -- Mixed traffic -- Network design problem -- Stochastic user equilibrium; Transportation and state -- Case studies -- Periodicals; Transportation -- Planning -- Case studies -- Periodicals; Transportation -- Research -- Case studies -- Periodicals; 388.05
• DOI: 10.1016/j.cstp.2018.11.006
• Languages: English
• ISSNs: 2213-624X

Abstract

Highlights: A facilitated road network for automated driving reduces total travel cost. It also increases safety and aids in a smooth transition to full automation. Higher market penetration of automated cars leads to more time and cost gains. This facilitated network also leads to a traffic shift among different road types. Abstract: This study explores a network configuration concept for vehicle automation levels 3–4 (according to SAE classifications) in an urban road network having mixed traffic and demonstrates its potential impacts. We assume automated driving will be allowed on a selection of roads. For the remaining roads, manual driving (although supported by assisting driving automation systems) will be compulsory. Accordingly, we introduce an approach for road selection and present relevant operational concepts. To evaluate the impacts of this configuration and model different vehicles' route choice behavior in mixed traffic, a static multi-class stochastic user equilibrium traffic assignment with a path-size logit route choice model and a Monte Carlo labeling route-set generation is adapted. Two user-classes are distinguished: vehicles with automation levels 0–2 and vehicles with automation levels 3–4 having a different passenger car unit to account for lower driving headways, lower value of time, and higher fuel efficiency. The results indicate a decrease in total travel cost with the increase in market penetration rate of higher automation levels, a decrease in total travel time, and a minor increase in total travel distance. Although in most cases vehicles with higher automation levels benefit more from the improvements, no deterioration in travel conditions is observed for the rest of the vehicles in any scenario. Furthermore, a noticeable shift of traffic from roads with access function to roads with flow function and distributors is observed. Sensitivity analysis shows that the extent of changes in the impacts is not strongly dependent on the input parameters.