Home Skip to main content

UGPTI Research Improves Accuracy of Rail and Road Condition Monitoring System

Posted: Jan 16, 2020

UGPTI researchers have developed a system that allows railroad and roadway monitors to use vehicle-mounted sensors to continuously monitor networks for potential problems. A paper recently published in the journal Sensors describes statistical techniques that allow sensors to more precisely identify the location of those problem areas.

Raj Bridgelall, an assistant professor of transportation and logistics and a researcher with NDSU's Upper Great Plains Transportation Institute, wrote "Accuracy Enhancement of Anomaly Localization with Participatory Sensing Vehicles." Denver Tolliver, UGPTI director, was the co-author.

Most cell phones have accelerometers, speed sensors, geospatial positioning systems (GPS), and other instrumentation necessary for the monitoring system, along with the ability to wirelessly transmit collected data. Because most vehicles are not yet equipped with similar instrumentation, cell phones were used as a stand-in for low-cost onboard sensors. However, the poor accuracy of positioning services associated with low-cost GPS hindered the ability of cell phones and sensors to replace the expensive specially equipped profiler vehicles currently used to monitor railways and roads. The statistical method developed for processing the data reduced localization error from more than 100 feet to less than 3 feet.

"The goal of our research and development of the system is to help railroad and roadway managers enhance the efficiency and effectiveness of visual inspections by identifying and prioritizing possible problem areas for inspector scrutiny, based on the magnitude of inertial events logged during normal vehicle operations," Bridgelall said. "More accurately identifying the location of problem areas will help road inspectors and managers focus their efforts."

NDSU Dept 2880P.O. Box 6050Fargo, ND 58108-6050