NDSU researchers have developed a method to identify and more precisely locate railroad track irregularities using the vast amount of data generated by low-cost sensors or sensors already in use aboard various railroad equipment.
“Signal Feature Extraction and Combination to Enhance the Detection and Localization of Railroad Track Irregularities” was written by Bhavana Bhardwaj, a doctoral student in computer science and a research assistant with the Upper Great Plains Transportation Institute at NDSU. Co-authors were Raj Bridgelall, assistant professor in the NDSU Department of Transportation, Logistics and Finance, and Pan Lu, an associate professor in the department. Bridgelall and Lu are both researchers with the institute.
The work was recently published in the journal IEEE Sensors.
“Traditional track inspection methods are laborious, relatively slow, expensive, and require track closure to search for possible track defects, Bridgelall sid. “A reliable, low-cost method for detecting and identifying the location of irregular rail track geometry can enhance the efficiency of traditional track inspections by focusing inspection resources on high-risk locations.”
Low-cost sensors similar to those found in most smart phones can be used aboard regular service trains to enhance coverage and monitoring frequency, the researchers say. However, GPS and sensor errors limit the accuracy of the method. The researchers developed an analytical framework and statistical techniques that use multiple transversals to reduce errors and improve location accuracy.
As a student-focused, land-grant, research university, we serve our citizens.