Adrian Cottam, Ph.D.

My research develops digital twins, large-scale models, and AI tools for transportation systems, with safety as the thread that runs through all of it.

Research areas

1. Transportation digital twins and simulation-based safety evaluation

I build digital twins of transportation facilities from field data and use them to evaluate designs and operations before they are built or changed. I developed a stochastic calibration method that matches simulation output to the full distribution of observed behavior rather than to averages, so that models of military entry control facilities reproduce real operational variability. These models are coupled with a full-motion driving simulator and with reinforcement and imitation learning agents that generate adversarial scenarios a hand-written test plan would never produce. We now also build roadway digital twins automatically from LiDAR for virtual curve safety studies.

A participant seated in a full-motion driving simulator with a red motion platform and wraparound monitors, inside a glass enclosure.
Our full-motion, high-speed driving simulator, integrated with an Unreal Engine and CarSim simulation suite.

2. Large-scale transportation system modeling with sparse data

Agencies manage networks of thousands of miles with sensors on only a fraction of them. My VEMLAN framework imputes volumes for failed freeway sensors, my crowdsourced estimation work recovers statewide traffic volumes and passenger car equivalency factors where no sensors exist, and my team's statewide analysis of probe vehicle composition earned a 2025 TRB Best Overall Paper Award.

3. AI for transportation safety and operations

I apply large language models, reinforcement learning, and imitation learning where they change what engineers can do. My team demonstrated zero-shot, open-vocabulary annotation of roadside assets in LiDAR point clouds using large language models, and built an AI Safety Analyst that uses multimodal large language models to make data-driven safety assessment accessible to agencies without specialized staff.

Research projects

Click a project to see details.

Public Transportation Management and Rural Transit Assistance ProgramAlabama Department of Transportation (ALDOT) · Co-PI · Oct 2026 – Sep 2029 · $3.98MCurrent

Provides training, technical assistance, and technology transfer to public and rural transit agencies across Alabama, bringing research tools and data-driven practice to the people who run transit service day to day.

A Data-Driven Comprehensive Action Plan for Multimodal Campus Road SafetyUSDOT Safe Streets and Roads for All (SS4A) · Co-PI · Awarded Oct 2026 · $1.09MCurrent

Develops a data-driven comprehensive safety action plan for a campus transportation network, addressing the safety of everyone who travels on it: pedestrians, bicyclists, micromobility riders, transit users, and drivers.

Research and Technology to Promote Safety of Commercial Motor Vehicles, Drivers, and Carriers (IDIQ)Federal Motor Carrier Safety Administration (FMCSA) · Co-PI · Jul 2026 – Jul 2031Current

Our AUTRI team was one of twelve teams selected nationally for this indefinite delivery, indefinite quantity contract, under which FMCSA can award task orders totaling up to $60 million through 2031 on commercial motor vehicle safety research.

Safety and Operation Analysis for Rectangular Rapid Flashing Beacon WarrantsUSDOT Safe Streets and Roads for All (SS4A) · Co-PI · Mar 2026 – Sep 2028 · $1.05MCurrent

Analyzes the safety and operational effects of rectangular rapid flashing beacons (RRFBs) at pedestrian crossings to inform warrants for when and where agencies should install them.

Rail Center for Research Enhancing Short-line Transportation (RailCREST)Federal Railroad Administration (FRA) · Co-PI, AUTRI team · Aug 2025 – Jul 2028 · $1.83MCurrent

Highway-rail grade crossings are among the biggest safety risks on the rail network: the U.S. has more than 203,000 of them, and 228 fatal crossing incidents occurred in 2023. RailCREST is a Penn State–led center founded through an FRA CRISI grant that our team helped write. Within it, our team is building a low-cost, real-time safety monitor for these crossings. A solar-powered camera and edge computing device at the crossing use machine learning to detect vehicles and pedestrians stopped on the tracks, then send the data over cellular networks to a central server. The server alerts railway operators and enforcement when something is on the tracks, and a web tool reports safety surrogate metrics such as how many vehicles stop on the tracks each day.

Because each unit carries its own solar power and cellular connection, the modular system can be deployed at almost any crossing in the U.S., which makes it especially useful for rural crossings and short-line railroads.

System diagram: a camera at a highway-rail grade crossing feeds an edge computing device and cellular hotspot in the field; data travels over the cellular network to an HRGC events database and web-tool server on the Auburn University intranet, which deliver safety surrogate analysis through a web host and risk alerts through an API.
How the system works: from the camera at the crossing, through the cellular network and Auburn servers, to risk alerts and safety analysis.
Safe and Efficient Entry Control Points at Military InstallationsDepartment of Defense · Research Team Leader · Aug 2023 – present · ~$1.4MCurrent

Led research team to evaluate Entry Control Facilities (ECFs) using traffic simulation techniques. Developed ECF simulation models to analyze traffic flow into military installations, and performed calibration and validation to ensure models matched real-world operations. Built a driving simulation suite on a full-motion platform to study threat-vehicle scenarios with human, reinforcement learning, and imitation learning drivers.

News article

Adrian Cottam with two colleagues in suits standing in front of a brick building at Auburn.
Traffic Data Analytics Platform DevelopmentArizona Department of Transportation · Lead Graduate Research Assistant · 2018 – 2023Completed

Developed the Statewide Mobility Analytics in Real Time (SMART) tool as the lead graduate research assistant of a joint project between the University of Arizona and the Arizona Department of Transportation (ADOT). Developed data collection software, databases for multi-sourced big data, and created the SMART tool dashboard.

ITS AZ 2021 Presentation

Screenshot of the SMART tool dashboard with a freeway map and travel time charts.
Optimizing Traffic Signals Using Multi-Source DataCity of Tucson · Research Assistant · 2017 – 2018Completed

Supported a multi-phase project between the City of Tucson and the University of Arizona to optimize traffic signals in the Tucson area. Assisted with developing a data visualization dashboard (tucson.ua-star.org) to assist the city with evaluating roadway performance.

Screenshot of the Tucson traffic dashboard with a map of intersections and a flow-rate chart.

Awards and fellowships

  • TRB Committee on Highway Traffic Monitoring (ACP70) Best Overall Paper Award, 2025
  • TRB Freeway Operations Committee (ACP20) Young Professional / Student Best Paper Award, 2025 (student-led)
  • NITC Dissertation Fellowship, 2023
  • Dwight David Eisenhower Transportation Fellowship, 2021 – 2022
  • GAANN Fellowship, 2019 – 2022

Professional service

  • Member, TRB Standing Committee on Intelligent Transportation Systems (ACP15), 2025; Paper Review Coordinator, 2024 – 2025
  • Friend of TRB committees on Artificial Intelligence and Advanced Computing Applications (AED50) and Information Systems and Technology (AED30)
  • Reviewer for Transportation Research Record, Transportation Research Part A, Journal of Transportation Engineering, Part A, and the TRB Annual Meeting (2021 – 2023, 2026)

Conference photos

Adrian at the ASLRRA conference beside the large ASLRRA letters and the association's logo.
ASLRRA conference
Adrian and graduate student Corey McDaniels beside their poster on zero-shot annotation of roadside assets in LiDAR data using large language models.
TRB Annual Meeting, January 2026: zero-shot LiDAR annotation with large language models
Adrian beside his poster on low-cost real-time threat notifications and surrogate safety metrics at highway-rail grade crossings.
ASCE ICTD 2025: real-time safety monitoring at highway-rail grade crossings
Adrian beside the poster 'A Statewide Highway System Network Analysis: Impact of Probe Vehicle Composition'.
TRB Annual Meeting, January 2025: statewide probe vehicle analysis (ACP70 Best Overall Paper)
Adrian standing among the large TRB letters at the Annual Meeting registration hall.
TRB Annual Meeting, January 2025
Adrian presenting at a podium at a conference session.
Lectern Session Presentation at TRB Annual Meeting 2024 in Washington DC
A full conference session audience watching a presentation.
Lectern Session Presentation at TRB Annual Meeting 2024 in Washington DC
Adrian standing beside his research poster.
Presenting Poster at TRB Annual Meeting 2024 in Washington DC
Adrian in front of the large TRB letters at the Annual Meeting.
TRB Annual Meeting 2024
Adrian discussing his poster with a conference attendee.
Presenting Accepted Peer-Reviewed Poster at TRB Annual Meeting 2023 in Washington DC
A group of colleagues in front of the TRB letters at the Annual Meeting.
2023 TRB Annual Meeting Washington DC
Adrian presenting the SMART Tool on stage, slide titled 'How does the SMART Tool work?'.
Presenting at 2022 ITS Arizona Conference
University of Arizona team at an exhibit booth with laptops and posters.
Helping with University of Arizona booth at 2022 Arizona Conference on Roads and Streets
Adrian beside his research poster at the TRB Annual Meeting.
2022 TRB Annual Meeting Eisenhower Fellow Poster Session Washington DC
Four colleagues standing in front of the TRB letters.
2020 TRB Annual Meeting Washington DC
A group of colleagues at a conference exhibit hall.
2019 ITS AZ Conference
Adrian standing in front of a Beijing University of Technology building, with large red Chinese characters on the lawn.
Beijing University of Technology Collaboration Meeting 2019
Three colleagues standing in front of a research poster.
2019 TRB Annual Meeting Washington DC
Adrian and a colleague in matching red shirts at a conference exhibit hall.
2018 NRITS Conference Arizona
Adrian and a colleague at a scenic overlook above a coastal city.
2018 COTA-ISETT Conference Hawaii
1 / 20