Our paper, Real-time lane-level abnormal traffic detection on freeways using sparse telematics data, is published in Accident Analysis & Prevention.
News
Roger Wilmot travel grant for TRB 2027
Awarded the Roger Wilmot Memorial Fund Travel Grant, administered by UW E-RIDE, to support planned participation in TRB 2027, with a focus on multimodal AI, vehicle telematics, and research partnerships for rural transportation safety. Details.
Wisconsin highway delay study completed
The Wisconsin Highway Delay Causation Study (2025–2026), supported by WisDOT, is complete. The study examines recurring and nonrecurring highway delay and contributing factors across Wisconsin. Final report.
UW Data Science Faculty Fellowship
I was selected as a 2026–2027 UW Data Science Faculty Fellow to develop vehicle-telematics methods for real-time roadway monitoring and proactive safety management on Wyoming highways, supported by the UW Data Science Center. Details.
NASA EPSCoR award for roadway weather AI
Awarded a Wyoming NASA EPSCoR Faculty Research Award to develop lightweight vision-language models for interpreting roadway weather from camera imagery, with natural-language explanations and evaluation on edge devices. Details.
SCOUT-WY receives seed funding
SCOUT-WY received support through the UW CEPS Engineering Initiative seed grant program for 2026–2027 to link vehicle observations across sparse roadside cameras and support travel-time estimation, traffic monitoring, and anomaly detection on rural Wyoming corridors. Details.
GeoLaneRep preprint
New preprint: Behavior-Grounded Lane Representation Learning for Multi-Task Traffic Digital Twins, introducing GeoLaneRep for behavior-aware traffic digital twins.
CrashSight at the CVPR DriveX Workshop
We will present our work, CrashSight, at the DriveX Workshop at CVPR 2026.
Support for the proposed WyCAIT center
Our proposed Wyoming Center for Artificial Intelligence in Transportation (WyCAIT) received support from the UW Science Institute for 2026–2027 to advance multimodal sensing, resilient AI, and corridor digital twins for rural transportation safety, operations, and workforce development. Details.
Agentic privacy workflow preprint
New preprint: An Agentic Workflow for Detecting Personally Identifiable Information in Crash Narratives, presenting a locally deployable workflow for privacy-sensitive crash data processing.
V2X-QA preprint
New preprint: V2X-QA, a dataset and benchmark for multimodal large language models in autonomous driving across ego, infrastructure, and cooperative views.