This project examined the concept of introducing coded, differential RF pulses into the rail at grade crossings and, using Time Domain Reflectometry (TDR) techniques, determine the distance to reflections of these pulses caused by approaching trains. Measurements of this train distance information and how it changes over time could conceivably be used to predict train arrival time for the activation of grade crossing warning systems. To test the proposed method, the electrical transmission line properties of railroad track were first determined, including variations for tie type, track ballast quality and moisture content. An electrical analog was then constructed to allow bench development and testing. The research plan then called for preliminary field testing at the Transportation Technology Center, Inc. (TTCI), followed by compatibility testing with the two leading U.S. signal manufacturers, and then a final field trial and demonstration on one of the loop tracks at TTCI.

Two key problems emerged during bench testing, both related to the conductance measured between the two rails, referred to in the industry as "ballast leakage": (1) the relatively poor real-world insulation between the rails leads to a high attenuation rate for differential signals propagating along the rails, and (2) this attenuation is highly dependent on the signal frequency with lower frequencies less affected. Research revealed that the high attenuation could be overcome by applying signal correlation processing methods that can detect extremely weak signals. However, this required the use of increasingly lower frequencies in order to achieve the target range of 8,000 to 12,000 feet, but lowering the frequency also increased the pulse time so that it became impossible to distinguish between the initial transmitted signals and the reflected signals. At present, for existing tracks with less than ideal insulation between the rails, pulse frequencies that are low enough to reduce attenuation to acceptable levels have pulse durations that cause transmit/receive timing overlap, and frequencies high enough to avoid transmit/receive overlap provide only limited range. As a result of these problems, the field testing and compatibility testing were not conducted.