HASS COE and Virginia Tech Predict the Impact of Automated Vehicles on Roadway Congestion
In partnership with the Virginia Polytechnic Institute and State University (Virginia Tech), the Highly Automated Systems Safety Center of Excellence (HASS COE) recently sponsored a student-led capstone project on the potential impact of automated vehicles (AVs) on urban road congestion.
Working with HASS COE Senior Strategist Ashley Nylen, PMP and Senior Scientist Dr. Mubassira Khan, three Virginia Tech Computational Modeling and Data Analytics (CMDA) students—seniors Cam Taylor, Joe Harrison, and John Lutz—created a computational model to evaluate the impact roadway AVs could potentially introduce into the broader transportation system. The students worked with HASS COE on the project design, execution, and interpretation of findings.
Taylor, Harrison, and Lutz sought to primarily model the impact of additional vehicle miles traveled by AVs on local and state roadways in order to provide a metric of potential congestion and associated economic and environmental risks. The goal was to create a predictive model that could evolve over time with refinement and iterative testing, as better open-source data becomes available (see Figure 1).
Using real-world data, the students analyzed and processed open-source traffic data to demonstrate how their automated model could assist local and state transportation organizations in predicting the impact of AVs on roadway congestion (see Figure 2).
The base model could hypothetically inform cities of the magnitude and degree of impact that AVs will have on the future of urban roadways. Starting with a base model, the students argued, would better equip local and regional DOTs to prepare for imminent changes to their roadway infrastructure.
The students presented their automated model to HASS COE staff. Based on the students’ demonstration of their research, HASS COE staff provided input on how the automated model could be leveraged by cities around the United States, with additional testing and iterations to the base model. The students incorporated HASS COE feedback, completed their findings, and presented their final capstone project for the semester on December 12, 2023 (see Figure 3).
Through the HASS COE partnership, Harrison said that he “gained many perspectives regarding traffic” and learned many new “technical project management skills and GUI skills along the way.”
The result demonstrated a real-world collaboration with transportation experts, providing practical guidance to students learning how computer science—and in particular, artificial intelligence—will continue to change how local and state DOTs plan for the widespread adoption of AVs in the future.
HASS COE is exploring and inviting future collaborations with students to empower tomorrow’s leaders with the expertise of current transportation subject matter experts. The results of the Virginia Tech pilot program have demonstrated the value of collaboration in developing new highly automated safety systems and the HASS COE commitment to workforce development for the future of this domain.
To explore a potential partnership with HASS COE, please contact hass-coe@dot.gov.
Contact Us
Denise Bakar
Communications Manager
denise.bakar@dot.gov
General Inquiries
hass-coe@dot.gov
