New researcher studies the traffic impacts of automated vehicles

backseat inspectionStern conducted automated vehicle experiments in 2016. When Raphael Stern was a doctoral candidate, he didn’t just study automated vehicles—he got behind the wheel to test their impacts. Stern brings this expertise to the U of M this year as a new assistant professor in the Department of Civil, Environmental, and Geo- Engineering. His primary research interests are automated vehicles and how they can be used for traffic control and estimation. Other interests include transportation cyber-physical systems and smart and resilient cities. 

Prior to joining the U, Stern was a postdoctoral scholar at the Technical University of Munich. Stern completed his graduate studies in the Department of Civil and Environmental Engineering and the Coordinated Science Laboratory at the University of Illinois at Urbana-Champaign. Read more about his plans and interests below.

 

Tell us about your plans for research and teaching.

We’re currently at a point where technology has the potential to transform how we think about transportation. This includes the emergence of automated vehicles on our roadways, improved sensing capabilities from cameras using deep neural networks, and the ability to analyze large amounts of mobility data in real time to better understand travel patterns and predict future travel demand.

My research interests are in understanding how these technologies, and specifically vehicle automation, can be leveraged to get the most out of our existing infrastructure. My teaching interests align closely with this since the next generation of transportation engineers will not only need to be well versed in traditional traffic flow theory and civil engineering techniques, but also be computationally literate and capable of applying the new computational techniques that will soon be more prevalent in transportation engineering.

 

What are some highlights from your recent work?

While fully automated vehicles may soon be approved for at least limited operation on our streets, adaptive cruise control (ACC) is commercially available now. ACC is a form of radar-assisted cruise control that represents the first step toward an automated future.

Recently, I’ve been interested in getting a better understanding of how ACC vehicles—which will soon be ubiquitous on our roads—will impact traffic flow. For this, I’ve done some work to test commercially available ACC vehicles to understand how they follow a leading vehicle and whether they amplify or dissipate small disturbances in the traffic flow (e.g., sudden braking). This work has important implications not only for traffic flow, but also for traffic safety, since changing the driving behavior of even just a small number of vehicles in the traffic flow could dramatically alter how shock waves propagate, for example.

Experiments I conducted in 2016 demonstrated how a single AV in a stream of 21 human-piloted vehicles could dampen stop-and-go traffic waves. In the tests, the lone AV increased the total throughput of the road by 14 percent, decreased the total number of braking events by 98 percent, and decreased average fuel consumption of the entire experimental fleet by up to 39 percent.

 

What are the implications of your work for Minnesota?

The emergence of partially automated vehicles on Minnesota roads over the next several years will mean large changes in terms of traffic flow and safety, as well as new best practices for traffic management. My research addresses these questions to understand how we can expect traffic flow to change and how best to prepare for the transition to automation to ensure that Minnesota is ready for the next era in transportation and mobility.


Related Link