CV Safety Alert and Predictive Crash Location Integration [Final Implementation Report]

Details

• Creators:
  Hegemier, Nick ; Sekela, Treea ; Magold, Darlene ; Wolpert, Andrew ; Kieffe, Adam
• Corporate Creators:
  Jacobs ; ETCH
• Corporate Contributors:
  Ohio. Department of Transportation ; Strengthening Mobility and Revolutionizing Transportation (SMART) Program ; United States. Department of Transportation. Office of the Assistant Secretary for Research and Technology
• Subject/TRT Terms:
  • Safety
  • Predictive models
  • Crash locations
  • Evaluation
  • Benefit cost analysis
  • Performance evaluations
  • Countermeasures
  • Incident management
• Resource Type:
  Tech Report
• Geographical Coverage:
  United States ; Ohio
• Corporate Publisher:
  Strengthening Mobility and Revolutionizing Transportation (SMART) Program
• Abstract:
  This project, funded by a USDOT SMART Grant, evaluates the feasibility and performance of two connected vehicle safety use cases designed to transition highway safety operations from a reactive state to a proactive model. Leveraging an Event Streaming Platform, the project aims to accelerate emergency dispatch and enable preventative roadway management. Methodology: The evaluation assesses two primary use cases:1. Incident Detection: Automatically ingests raw connected vehicle airbag deployments and automatic emergency call (eCall) data directly from manufacturers. Alerts are standardized, enriched with linear referencing, and routed to regional 911 dispatch centers.2. Crash Prediction: Implements a machine learning model trained on historical crash data. It fuses real-time weather, traffic volume, and telematics (e.g., hard braking) to generate segment-specific roadway risk scores between 0 and 1.Conclusions: Stage 1 prototype evaluations validated that the incident detection use case enriches airbag alerts in under one second and maps crash coordinates within 10meters, proving successful end-to-end routing to local dispatch. The predictive use case successfully isolated 20% of all crashes onto high-risk segments representing just 4% of hourly network observations. These results establish a scalable public-private framework for national proactive highway safety.
• Format:
  PDF
• Funding:
  69A3552341032
• Download URL:
  https://rosap.ntl.bts.gov/view/dot/93172/dot_93172_DS1.pdf
• File Type:
  [PDF - 1.72 MB]
• Collection(s):
  US Transportation Collection
• Main Document Checksum:
  urn:sha-512:b76acfa2af566c7c21cd5282f5f82d322498f1429f0932db26bea0eff77f18cc5e5f2b07754a13bfdf44d9abc5d034228e058a341b4b903b702fd71cfabdd1e0