AI-Enabled Crash-to-Care Integration for Park City
Pilots an AI-enabled crash-to-care system on three Park City corridors, covering multimodal sensing, conflict prediction and warnings, automated crash reporting, and emergency response data sharing.
Partners: Park City Municipal Corporation (lead applicant); AIWaysion
Proactive V2X System for Beyond-Line-of-Sight (BLOS) Hazard Prevention
Develops a roadside system that jointly predicts the motion of all road users in a scene and warns approaching drivers of hazards beyond their line of sight before a conflict develops.
Partners: Utah Department of Transportation
Evaluation of AI-Driven Pavement Monitoring for Optimized Maintenance Resource Allocation
Evaluates AI-driven pavement monitoring as a basis for allocating maintenance resources.
Evaluation of AI-Driven Pavement Monitoring for Optimized Maintenance Resource Allocation
Evaluates AI-driven pavement monitoring as a basis for allocating maintenance resources.
INTERACT: Intent-Driven Human–Vehicle Interaction with Constraint-Aware Control
Uses advanced GPU computing to develop safer, more responsive AI-enabled vehicle-control systems that act on human intent within explicit safety constraints.
Reasoning-Enabled Autonomy System for Open Negotiation
Builds an autonomy architecture in which vision-language models translate human instructions into constraints grounded in sensor data, maps, and traffic law, with every policy update validated against safety invariants.
An AI-Based Reasoning Framework for Proactive Infrastructure Monitoring and Preservation Using CAVs
Turns the sensors already carried by connected and automated vehicles into a continuous inspection function, with a retrieval-augmented language model that grades each detected anomaly and recommends an intervention.
Advancing Vision Zero on Tribal Lands through Real-time Hazard Detection & Behavior Monitoring
Instruments the unstable coastal bluff with geotechnical sensors and US 101 with AI computer vision, fusing both into one early-warning platform for the Trinidad Rancheria’s lifeline route.
Partners: Cher-Ae Heights Indian Community of the Trinidad Rancheria (lead applicant)
Integrated AI Computing and V2X Infrastructure Systems for Safer Streets
Deploys edge-AI sensing and V2X communication at four sites spanning an urban core, a rural county, and a tribal community, with a language-model reasoning layer that feeds findings back into safety action plans.
Partners: Salt Lake City; Garfield County; Trinidad Rancheria; AIWaysion; Johns Hopkins University; Rensselaer Polytechnic Institute
Transportation Certification Program
Builds a 15-credit stackable online graduate certificate in transportation infrastructure electrification; Dr. Liu develops and teaches CVEEN 6592.
Partners: Utah State University
AI-Driven Predictive Safety Framework for V2X-Based Road User Protection
Forecasts short-term road-user trajectories on edge hardware at UDOT’s LiDAR- and V2X-equipped intersections and broadcasts predictive warnings with hazard location, time to collision, and model confidence.
Partners: Utah Department of Transportation
Drone Network Design for Comprehensive Emergency Response in Rural Utah
Builds a siting and fleet-design optimization model for drone bases that support emergency response across rural Utah, paired with an analysis of federal beyond-visual-line-of-sight rules.
Quantifying the Impacts of UTA’s Frequent Transit Network on Congestion, Safety, and Mode Shift
Treats Salt Lake City’s staggered upgrade of three corridors to 15-minute service as a natural experiment to measure effects on ridership, congestion, and safety.
Partners: Utah Transit Authority
Distilling GPT-4o for the Edge: Real-Time Transportation Decision Making via Multi-Modal Knowledge Extraction
Distills knowledge from a large multimodal model into compact models that run on edge devices for real-time transportation decision making.
Integration of Edge-AI in Machine Learning for Civil Computing
Brings edge-AI hardware into machine learning instruction for civil engineering computing.