Exhibit D - Research Project Requirement: PDF Link

Project Summary: This proposal details an Integrated Hierarchical Digital Twin (DT) framework designed to revolutionize Transportation Incident Management (TIM) and hazard mitigation within rural contexts. Rural networks are characterized by sparse sensing and delayed response, leading to disproportionately high fatality rates. Our solution divides the DT into two interactive layers: an Abstract Digital Twin and a Physics-based Digital Twin. The Abstract DT uses Graph Neural Networks (GNN) and Diffusion modeling to maintain network-wide situational awareness, inferring traffic states from spatiotemporally diverse, multiscale data sources. The critical innovation is an event-triggered state estimation mechanism: when the Abstract DT identifies a high-risk spatiotemporal pattern or incident, it triggers the Physics-based DT, a high-fidelity Unity engine simulation, for that specific "hot-spot." This localized layer uses physics-based models to calculate precise danger thresholds for vehicle stability during incidents, such as floods or icing events. The framework concludes with an AI agent-based decision-support system to coordinate responders, aiming to reduce detection and response latency and provide vehicle-specific hazard warnings via connected infrastructure and/or C-V2X.

• TBD

TBD

TBD

TBD