AI-Engine for Adaptive Sensor Fusion for Traffic Monitoring System

Details

• Creators:
  Dubey, Abhishek ; Work, Daniel B. ; Baroud, Hiba ; Mukhopadhyay, Ayan
• Corporate Creators:
  Vanderbilt University. Institute for Software Integrated Systems
• Corporate Contributors:
  Tennessee. Department of Transportation ; United States. Department of Transportation. Federal Highway Administration
• Subject/TRT Terms:
  Anomaly Detection Artificial Intelligence Data Fusion Emergency Response Time Incident Management Localization Sensor Fusion State Of The Art Traffic Surveillance
• Publication/ Report Number:
  RES2023.17
• Resource Type:
  Tech Report
• Geographical Coverage:
  United States ; Tennessee
• Edition:
  Final Report
• Corporate Publisher:
  Tennessee. Department of Transportation
• Abstract:
  Effective traffic incident management is critical for road safety and operational efficiency. Yet, many transportation agencies rely on predominantly reactionary methods, where incidents are reported by human agents and managed through rule-based frameworks like traditional Traffic Incident Management (TIM) systems. These methods, however, are vulnerable to human error, oversight, and response delays during high-stress conditions. Although recent initiatives incorporating real-time sensor data for corridor monitoring and enhanced roadway information systems represent strides toward modernization, these systems often still require substantial human intervention. Recent advancements in graph-based deep learning models offer promising potential for addressing the limitations of traditional methods. However, state-of-the-art models encounter challenges such as limited availability of high-quality labeled data, variability in real-time traffic measurements, and the complexities of incident localization within dynamic and interconnected road networks. To address these challenges, we propose the Traffic Response Anomaly Capture Engine (TRACE), a novel approach that combines graph neural networks, transformers, and probabilistic normalizing flows to accurately detect and localize traffic anomalies in real time. TRACE captures spatial-temporal dependencies, manages data uncertainty, and enhances automation, supporting more precise and timely incident localization. Our approach is validated on real-world traffic data and demonstrates improved detection accuracy by 5% and incident localization by 10% compared to current state-of-the-art methods, advancing traffic management strategies for safer, more efficient roadways.
• Format:
  PDF
• Funding:
  RES2023.17
• Collection(s):
  US Transportation Collection
• Main Document Checksum:
  urn:sha-512:838cc7b7a8f349090a9c19638ae651ef3af5b9f0b1d8ab4edf830a0ae86d6442ca75c4d5c9c968c69f1729791a072494c085be18564648769f3f72203adfb612
• Download URL:
  https://rosap.ntl.bts.gov/view/dot/79430/dot_79430_DS1.pdf
• File Type:
  [PDF - 2.20 MB ]