Incident-aware Resilient Traffic Management for Urban Road Networks (InTURN)

• Applicant: Professor Dr.-Ing. Sven Tomforde
• Subject Area: Traffic and Transport Systems, Intelligent and Automated Traffic ; Computer Architecture, Embedded and Massively Parallel Systems
• Term: from 2019 to 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 420542957

• Final Report Year: 2023

Final Report Abstract

A desirable goal in the use of today’s urban traffic networks is traversing it with the utmost speed and dependability, all the while avoiding any bottlenecks and disruptions due to traffic accidents as well as saving fuel and emissions at the same time. Additionally, this goal should be accomplished without repeatedly and excessively sharing the present situation, location and contextual information with large data-driven corporations, thus safeguarding your privacy. This project explored the realisation of such a system as a self-adaptive and self-organising (SASO) mechanism. As the base SASO system, we used our preliminary work known as Organic Traffic Control (OTC), an intersection-centric system for traffic light control, progressive signal systems and route guidance. Concerning the idea of data minimisation, the strategies of this decentralised approach address the combination of substantial traffic demands, intertwined with the underlying dynamics and ever-changing behaviours, along with the presence of disruptions such as incidents. Machine learning (ML) techniques were used to identify abnormal traffic conditions and incidents. While aiming for automatic categorisation, we consider the effect on nearby intersections and road segments. This relationship was used for a cooperative validation of detected incidents and, building upon that, for an incident-aware formation of distributed progressive signal systems. Using simulated urban road networks and relying only on simple vehicle-counting detectors (loop detectors), we explored several approaches for incident detection. That includes decision-tree-based classifiers that perform well for stronglypronounced incidents that close off whole sections. As expected, partial closures, such as blocking off only one lane, are more challenging to recognise. To partially compensate for that, incidents can be validated using information from neighbouring intersections. The results were peer-reviewed and published. Together with the software, these can aid further research.

Publications

• A Self-organising System Combining Self-adaptive Traffic Control and Urban Platooning: A Concept for Autonomous Driving. Proceedings of the 7th International Conference on Vehicle Technology and Intelligent Transport Systems, 429-437. SCITEPRESS -Science and Technology Publications.
  Hamann, Heiko; Schwarzat, Julian; Thomsen, Ingo & Tomforde, Sven
  
• Incident-aware Resilient Traffic Management for Urban Road Networks, volume 18 of Intelligent Embedded Systems, pages 125–138. kassel university press
  I. Thomsen
  
• Urban Traffic Incident Detection for Organic Traffic Control: A Density-based Clustering Approach. Proceedings of the 7th International Conference on Vehicle Technology and Intelligent Transport Systems, 152-160. SCITEPRESS -Science and Technology Publications.
  Thomsen, Ingo; Zapfe, Yannick & Tomforde, Sven
  
• A Concept for Collaborative Incident Validation in a Self-organised Traffic Management System. Proceedings of the 8th International Conference on Vehicle Technology and Intelligent Transport Systems, 316-323. SCITEPRESS -Science and Technology Publications.
  Tomforde, Sven & Thomsen, Ingo
  
• Adaptive Approaches for Tidal-Flow Lanes in Urban-Road Networks. Future Transportation, 2(3), 567-588.
  Striewski, Sören; Thomsen, Ingo & Tomforde, Sven
  
• Detecting Extended Incidents in Urban Road Networks for Organic Traffic Control Using Density-Based Clustering of Traffic Flows. Smart Cities, Green Technologies, and Intelligent Transport Systems, 330–347.
  Thomsen, Ingo & Tomforde, Sven
  
• Intersection-centric Urban Traffic Flow Clustering for Incident Detection in Organic Traffic Control. Proceedings of the 8th International Conference on Vehicle Technology and Intelligent Transport Systems, 410-417. SCITEPRESS -Science and Technology Publications.
  Thomsen, Ingo & Tomforde, Sven
  
• Traffic Flow based Detection of Incidents in Urban Road Networks, volume 20 of Intelligent Embedded Systems, pages 179–194. kassel university press
  I. Thomsen
  
• Abnormal Behaviour Detection of Self-Adaptive Agents in Traffic Environments. 2023 IEEE International Conference on Autonomic Computing and Self-Organizing Systems Companion (ACSOS-C), 41-46. IEEE.
  Goller, Martin; Thomsen, Ingo; Al-Falouji, Ghassan & Tomforde, Sven
  
• Approaches to Automatic Road Traffic Incident Detection and Incident Forecasting. Proceedings of the 9th International Conference on Vehicle Technology and Intelligent Transport Systems, 289-296. SCITEPRESS -Science and Technology Publications.
  Striewski, Sören; Thomsen, Ingo & Tomforde, Sven
  
• Distributed Collaborative Incident Validation in a Self-Organised Traffic Control System. Proceedings of the 9th International Conference on Vehicle Technology and Intelligent Transport Systems, 152-159. SCITEPRESS -Science and Technology Publications.
  Thomsen, Ingo; Brennecke, Torben & Tomforde, Sven
  
• Incident-Aware Distributed Signal Systems in Self-Organised Traffic Control Systems. Proceedings of the 9th International Conference on Vehicle Technology and Intelligent Transport Systems, 15-26. SCITEPRESS -Science and Technology Publications.
  Tomforde, Sven; Ohl, Yanneck & Thomsen, Ingo
  
• Inturn software. Version 1.0.
  S. Tomforde & I. Thomsen