Researchers at North Dakota State University are refining techniques for using sensors to more safely and precisely identify railroad track irregularities that could lead to derailments or accidents. The current method of inspection and maintenance is expensive, slow, increases the safety risk to workers, and requires closing the track. Researchers developed a framework that considers a technique called "ensemble average," which leverages a large number of signals from low-cost sensors collected from multiple traversals of a railway segment to enhance the signal's quality and reduce the detection error. Analysis found that the technique could identify the location of an irregularity to within 5 meters.
Researchers at North Dakota State University are refining techniques for using sensors to more safely and precisely identify railroad track irregularities that could lead to derailments or accidents. The current method of inspection and maintenance is expensive, slow, increases the safety risk to workers, and requires closing the track. Researchers developed a framework that considers a technique called "ensemble average," which leverages a large number of signals from low-cost sensors collected from multiple traversals of a railway segment to enhance the signal's quality and reduce the detection error. Analysis found that the technique could identify the location of an irregularity to within 5 meters.