Human-centered geometric design of roads using an autonomous vehicle problem

Geometric design of roads is a key component of an undergraduate civil engineering curriculum. At the University of Illinois, geometric design is taught in a senior design class for upperclassmen and new graduate students. In order to encourage critical analysis of design standards, especially human-related factors, a new design project was implemented that incorporates Autonomous Vehicles (AVs). Working in multi-disciplinary teams, students were required to develop new geometric design standards for fully autonomous AVs together with human cyclists on a principal arterial in Puerto Rico. The class was given an introductory lecture on AVs, after which they were required to review the literature, modify existing standards, and implement the findings into their roadway designs. The teams evaluated the human-related aspects of design by re-considering multiple geometric design parameters for AVs, as well as interaction between AVs and cyclists. All teams recommending that lane width, perception reaction time, and stopping sight distance criteria be modified, while no team recommended that speed limit and maximum rate of superelevation be modified. All teams recommended that cyclists be separated from AV lanes, with most teams recommending a shared shoulder. During peer assessment, students also expressed strong satisfaction with their teams while working on this futuristic roadway problem.