System and method for inspecting a rail using machine learning

Abstract

An aspect includes a vehicle that includes rail inspection sensors configured for capturing transducer data describing the rail, and a processor configured for receiving and processing the transducer data in near-real time to determine whether the captured transducer data identifies a suspected rail flaw. The processing includes inputting the captured transducer data to a machine learning system that has been trained to identify patterns in transducer data that indicate rail flaws. The processing also includes receiving an output from the machine learning system, the output indicating whether the captured transducer data identifies a suspected rail flaw. An alert is transmitted to an operator of the vehicle based at least in part on the output indicating that the captured transducer data identifies a suspected rail flaw. The alert includes a location of the suspected rail flaw and instructs the operator to stop the vehicle and to perform a repair action.

1 - 19 . (canceled) 
     
     
         20 . A vehicle for inspecting a rail, the vehicle comprising:
 rail inspection sensors configured for capturing data describing the rail; and   a processor configured for:
 receiving the data; and 
 processing the data to determine whether the data identifies a suspected rail flaw or a suspected pattern, the processing comprising:
 inputting the data into a machine learning model that has been trained to classify rail flaws and patterns; and 
 receiving an output from the machine learning model, the output indicating whether the data identifies the suspected rail flaw or a suspected pattern; 
 auditing the machine learning model to identify at least one of a false positive or a false negative, the auditing comparing the output from the machine learning model to an output from a human analyst; 
 adjusting the machine learning model based on the at least one of the false positive or the false negative; and 
 causing an alert to be transmitted based at least in part on the output indicating that the captured transducer data identifies a suspected rail flaw or a suspected pattern, the alert including a location of the suspected rail flaw or a location of the suspected pattern. 
 
   
     
     
         21 . The vehicle of  claim 20 , further comprising a camera for capturing a camera image of the location of the suspected rail flaw or the location of the suspected pattern, wherein the alert includes the camera image, wherein the machine learning model is a first machine learning model, and wherein the processing further comprises:
 inputting the camera image into second machine learning model trained to detect whether a weld is present on the rail.   
     
     
         22 . The vehicle of  claim 20 , wherein the output is transmitted, via a network, to a remote storage device or a remote processor. 
     
     
         23 . The vehicle of  claim 20 , wherein the rail inspection sensors include one or more of an ultrasonic transducer, an induction transducer, and an eddy current transducer. 
     
     
         24 . The vehicle of  claim 23 , wherein the data comprises a stream of overlapping sequences comprising a sequence of overlapping 2-dimensional images, wherein the overlapping 2-dimensional images overlap by a predetermined number of pixels. 
     
     
         25 . The vehicle of  claim 24 , wherein the predetermined number of pixels is greater than half the width of the overlapping 2-dimensional images. 
     
     
         26 . The vehicle of  claim 23 , wherein the data comprises a stream of overlapping sequences comprising a sequence of 1-dimensional signals. 
     
     
         27 . The vehicle of  claim 26 , wherein the data are collected by the eddy current transducer, and wherein a number of the sequence of 1-dimensional signals is 20. 
     
     
         28 . The vehicle of  claim 26 , wherein the data are collected by the induction transducer, and wherein a number of the sequence of 1-dimensional signals is 4. 
     
     
         29 . A method for inspecting a rail, the method comprising:
 receiving data from at least one rail inspection sensor mounted on a vehicle that is located on the rail; and   processing the data to identify locations of suspected rail flaws in the rail, the processing comprising:
 inputting the data into a machine learning model that has been trained to classify rail flaws and patterns; 
 receiving an output from the machine learning model, the output including a list of suspected rail flaws and their corresponding locations on the rail; 
 auditing the machine learning model to identify at least one of a false positive or a false negative, the auditing comparing the output from the machine learning model to output from a human analyst; 
 adjusting the machine learning model based on the at least one of the false positive or the false negative; and 
 initiating a repair action based on the list of suspected rail flaws. 
   
     
     
         30 . The method of  claim 29 , wherein the transducer data includes images. 
     
     
         31 . The method of  claim 30 , wherein the inputting the transducer data includes streaming overlapped images. 
     
     
         32 . The method of  claim 29 , further comprising:
 determining a unique position on the rail based at least in part on a test segment identifier, a rail indicator, and a rail pulse count.   
     
     
         33 . The method of  claim 32 , wherein the test segment identifier indicates a data acquisition recording run on the rail of the suspected rail flaw or the suspected pattern, wherein the rail indicator indicates whether the suspected rail flaw or the suspected pattern is in a left rail or a right rail, and wherein the rail pulse count indicates a distance along the rail from a start of a recording. 
     
     
         34 . The method of  claim 32 , wherein the unique position on the rail is expressed as a latitude and a longitude 
     
     
         35 . The method of  claim 32 , wherein the unique position on the rail is expressed as a milepost and a yardage. 
     
     
         36 . The method of  claim 29 , wherein the at least one rail inspection sensor is selected from the group consisting of ultrasonic transducers, induction transducers, and eddy current transducers. 
     
     
         37 . The method of  claim 29 , wherein the receiving is from a requestor via a network and the method further comprises sending the list of suspected rail flaws to the requestor. 
     
     
         38 . A system for inspecting a rail, the system comprising:
 a non-transitory memory having computer readable instructions; and   one or more processors for executing the computer readable instructions, the computer readable instructions controlling the one or more processors to perform operations comprising:
 receiving data from at least one rail inspection sensor mounted on a vehicle that is located on the rail; 
 processing the data to identify locations of suspected rail flaws in the rail, the processing comprising:
 inputting the data into a machine learning model that has been trained to classify rail flaws; and 
 receiving an output from the machine learning model, the output including a list of suspected rail flaws and their corresponding locations on the rail; 
 auditing the machine learning model to identify at least one of a false positive or a false negative, the auditing comparing the output from the machine learning model to output from a human analyst; 
 adjusting the machine learning model based on the at least one of the false positive or the false negative; and 
 initiating a repair action based on the list of suspected rail flaws. 
 
   
     
     
         39 . A computer program product for inspecting a rail, the computer program product comprising a non-transitory computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to perform operations comprising:
 receiving data from at least one rail inspection sensor mounted on a vehicle that is located on the rail; and   processing the data to identify locations of suspected rail flaws in the rail, the processing comprising:
 inputting the data into a machine learning model that has been trained to classify rail flaws; and 
 receiving an output from the machine learning model, the output including a list of suspected rail flaws and their corresponding locations on the rail; 
 auditing the machine learning model to identify at least one of a false positive or a false negative, the auditing comparing the output from the machine learning model to output from a human analyst; 
 adjusting the machine learning model based on the at least one of the false positive or the false negative; and 
 initiating a repair action based on the list of suspected rail flaws.