For decades, railroads have relied on track circuits todetect train presence and broken rails. However, track circuitry is expensive to maintain, and does not always reliably detect trains due to such factors as contamination at the wheel-rail interface. Moreover, a substantial percentage of rail breaks occur in which electrical continuity is maintained, and are therefore not detected by the track circuits. This research investigated the feasibility of using fiber-optic filaments buried along the right-of-way to detect trains and to detect the energy released when rail breaks occur. The approach was to use coherent optical time-domain reflectometry (COTDR) in concert with advanced signal-processing techniques and neural networks in buried fiber-optic filaments to detect and locate trains and the ballistic event characteristic of rails breaking under stress.
An optical transmission through a continuous length of low-loss, telecommunications-grade fiber buried along the right-of-way, yet away from track maintenance operations, was investigated to determine whether it held any promise for providing an inexpensive, reliable alternative to conventional track circuitry for train presence and broken rail detection. Another potential advantage is that buried fiber-optic filament is free of the problems associated with the electromagnetic interference encountered with track circuits. The objective was a low-cost, reliable alternative to conventional track circuits for near real-time detection and location of rail break events, as well as detection and location of moving trains that can be commercially developed for application to the railroad. If successful, this technology could also facilitate the railroad industry movement toward communications-based train control systems and away from track-circuit dependent train control.
A state-of-the-art coherent laser is used to pulse a buried communications-grade optical fiber. Information is extracted from polarization shift in the laser pulse backscatter light to establish train presence and rail break events as well as the location and time of events. The laser employs coherent continuous waves with a line width of approximately 10 kilohertz. The laser beam is pulsed at 30 nanoseconds over a 0.1 millisecond period to provide a 2-meter resolution in a 20-kilometer fiber length. The concept was that the system would recognize that a train has stopped by registering the cessation of activity at the last known location.
The system has the potential to continuously monitor train movement, direction, and location while monitoring the track structure for rail breaks. Additionally the system has the potential to detect and discriminate among various in-train defects, e.g., flat wheels, dragging equipment, and stuck brakes.
The final report for this IDEA project can be found at:
