Island Radar Completes Train Detection Test

In May 2013, Island Radar completed its long-term evaluation of a radar-based, BNSF train detection application. The dual radar system was installed in Lamar Missouri, at a BNSF and MNA track crossover. The radar system detects trains and potenitally dropped cars in the track circuit dead zone surrounding the crossover. 

Track circuits cannot be maintained in and around crossovers, creating a ‘dead zone’ between insulators in which trains or possible dropped cars cannot be detected. This condition is exacerbated to the extent that insulator joints are moved away from the crossover structure  - minimizing insulator fouling, but creating detection dead zones that may be longer than a railroad car length.

RadarAtCrossoverOptPD Loop systems can provide detection, but they are typically clamped to the tops of ties and exposed in a manner that is susceptible to damage. A dual radar system was developed and deployed to test the performance and applicability of a non-loop system in this application.

Following installation, configuration, and a period of initial testing, comprehensive detection performance data was collected from both the radar system and a PD Loop system on a north-south BNSF track corridor in Lamar Missouri. During this formal test period, which ran from mid-February to mid-April 2013, 835 BNSF trains were detected by the radar detection system with a 100% correlation to detection events registered by the PD Loop system. No false detection events were recorded from snow, rain, or other meteorological conditions.  128 trains were detected on the east-west MNA track at the site, but it was not possible to connect to corresponding PD Loop state information from them MNA track due to the PD Loop system’s auxiliary output circuitry or its configuration.

With the proper timing and attenuation settings, visual data illustrated the radar’s return emission signal power levels (signal to noise ratios) to be substantial and continuous when a train occupied the track segment and virtually zero when the track segment was clear. Radar masts, generally 20 feet in height, were installed at convenient, but not optimal, locations at the test site. But the wide detection footprint of the radar sensors provides substantial mast siting flexibility. It is possible that certain applications based on the local landscape, track geometry, and surrounding features, may be fulfilled with a single radar sensor.

It is expected that the radar sensors in this application would, in practice, be interfaced into the existing signal equipment in a manner that establishes a ‘trap circuit’. In this manner, the radar presence detection system would provide a trap release function to the signal system logic as a train proceeds past the crossover structure. This implementation precludes any necessary alteration of the radar sensor background integration processes and provides a means of automatically resetting the trap circuit without requiring site visits by railroad maintenance personnel.