Early warning system for potential train collisions: Enhancing safety levels of high-speed trains

Collision avoidance support for high-speed trains is a major safety target for railway operators. This paper presents a new concept for a train collision early warning system (TCEWS) that addresses the problem of train-to-train collisions. Similar to existing collision avoidance systems used by railway networks, the TCEWS approach allows drivers to “see” its neighborhood leading target train and have up-to-date and accurate knowledge of the traffic situation in the vicinity. However, the system architecture and operational principles of TCEWS have certain differences. The core idea is to build a system that is independent of the current railway signaling-based systems, e.g. the Chinese Train Control System level-2 or level-3. With additional sensors and prior information from the electronic track map database, the Global Navigation Satellite System is able to provide the ability for train state detection and collision risk assessment without relying on Automatic Train Protection or other existing train-borne equipments. Wireless communication enables the state of the train to be monitored and an early warning to be issued due to cooperation between the train and trackside facilities. The system components and working principles of the TCEWS are detailed in this paper, and the characteristics and advantages are summarized by comparisons with collision avoidance systems in railway and other transport modes. Implementation and field test of TCEWS on the Hankou–Yichang Railway are introduced and analyzed. It is demonstrated that the TCEWS has sufficient capability to be a “safety overlay” system on top of existing railway signaling systems, and has huge potential in ensuring safety levels of high-speed railways using the Global Positioning System, the BeiDou Navigation Satellite System or other global navigation satellite systems.


  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01599941
  • Record Type: Publication
  • Files: TRIS
  • Created Date: May 4 2016 9:12AM