Top Innovation Ideas Aiming to Win $1 Million and Transform U.S. Transportation

The 15 Semifinalist Concepts

The semifinalists include Agentic AI for Adaptive and Resilient Middle-Mile Freight Operations, a system developed by Lacy Greening in Arizona that uses agentic artificial intelligence to optimize freight logistics, anticipate disruptions, and dynamically reroute shipments under real-world challenges. From California, UCLA Mobility Lab proposes CivicTwin, a physics-grounded digital twin of entire cities built from real-world data, capable of simulating mobility systems, infrastructure performance, and urban-scale scenarios before any physical implementation.

The University of Utah’s Energy in Motion: Atoms on Wheels concept introduces safe, monitored transport of mobile microreactors, offering emergency energy deployment for remote communities, military bases, or disaster-response operations. Meanwhile, the University of Michigan Transportation Research Institute and SaferDrive AI deliver a Generative AI City-Scale Digital Twin designed to virtually test autonomous vehicles through millions of simulated scenarios—long before they reach public roads.

Security in navigation emerges as another major theme. NavSentinel, from TruNav LLC of Illinois, presents a resilient GNSS receiver engineered to withstand interference and spoofing attempts. Complementing this approach, Tip & Cue Inc. of New York proposes Providing a Signal Source of Truth, a system that ensures a verified, trustworthy navigation signal for vehicles, trains, drones, and aviation systems.

Robotics and automation also play a critical role. New York–based MetaRE and Arup propose Next-Gen Smart Sensors Turn Every Train into an Inspector, a concept that transforms trains into continuous safety-inspection platforms capable of detecting track and tunnel anomalies without service interruptions. Colorado State University’s Pavement & Airfield Tactical Robotic On-site Locator (PATROL) introduces a robotic inspection system capable of detecting runway and pavement damage in real time. Idaho’s Yang Lu contributes Twin-Guided Robotic Pavement Overlays (T-GRO), which uses digital twin guidance to deliver faster, cleaner, and safer road resurfacing.

The Tennessee team—University of Tennessee Chattanooga, University of Tennessee, and the Oak Ridge Innovation Institute—proposes PRISM, an AI-powered real-time dynamic mapping system designed to improve autonomous vehicle operation through live, adaptive mapping.

Climate resilience is another critical frontier. Ohio’s ARCTOS Technology Solutions introduces Rapid Flood Damage Mitigation and Resilience Through Modular Biocement Infrastructure, a technique using fast-setting biocement to reinforce eroded terrain, stabilize roads, and mitigate flood impacts. Virginia’s CNA Corporation proposes Quantum Cooperative Air Traffic Simulation (Q-CATS), a quantum-inspired simulation system designed to manage complex air traffic interactions with unprecedented precision.

Wisconsin innovators Sikai Chen and Xiaopeng Li contribute SemanticVector, an AI-powered semantic compression system capable of processing up to 1,000 times more transportation data with radically reduced storage needs. Also from Virginia, Leidos proposes The Gravity of Infrastructure, an advanced modeling approach to better predict how bridges, highways, and heavy structures respond to stress, vibration, and loads, extending their operational lifespan.

Finally, the Georgia-based collaboration between Numobility and the Texas Transportation Institute brings TransportationBench, a national framework designed to accelerate responsible AI adoption through benchmarking, standardized evaluations, and transparency.