Asset failure prediction with location uncertainty

Patent metadata

Field	Value
Publication number	US-10414416-B2
Application number	US-201313962287-A
Country	US
Kind code	B2
Filing date	Aug 8, 2013
Priority date	Jan 11, 2013
Publication date	Sep 17, 2019
Grant date	Sep 17, 2019

Abstract

Geo-defect repair modeling with location uncertainty is provided. A method includes logically dividing a railroad network into segments each of a specified length. The method also includes identifying, via a computer processor, geo-defects and approximated locations of the geo-defects occurring at each inspection run for each of the segments. The method also includes calculating, via the computer processor, a rate of increase in amplitude of each of the geo-defects for each of the segments between inspection runs, determining a correlation of the geo-defects between the inspection runs as a function of the approximated locations, and predicting a deterioration rate for each of the geo-defects based on the calculating.

First claim

The invention claimed is:

1. A method, comprising:
logically dividing a railroad network into segments each of a specified length, the railroad network comprising a railroad network infrastructure that includes a plurality of data sources, the plurality of data sources comprising sensors configured to monitor operational aspects of components of the railroad network;
collecting data from the plurality of data sources with respect to monitored operational aspects of the components;
identifying, via a computer processor that is communicatively coupled to the plurality of data sources, geo-defects and approximated locations of the geo-defects occurring at each inspection run for each of the segments;
resolving location uncertainty of the approximated locations of each of the geodefects based on geo-defect type, geo-defect amplitude, and location proximity for corresponding geo-defects assessed at each inspection run, wherein the corresponding geo-defects are determined by matching identified geo-defects from multiple inspection runs within each segment of a specified length;
calculating, via the computer processor, a rate of increase in amplitude of each of the geo-defects for each of the segments between inspection runs;
predicting a deterioration rate for each of the geo-defects based on the calculating;
upon determining multiple geo-defects exist for one of the segments at one of the inspection runs and at least one geo-defect exists for the one of the segments at another of the inspection runs, aggregating the multiple geo-defects to reflect the single geo-defect;
determining an instantaneous rate of failure probability, the determining of the instantons rate of failure probability comprising for each of the segments:
calculating a likelihood of the derailment occurring at a time t;
calculating a likelihood that the segment will survive until time t; and
calculating the instantaneous rate of failure probability at time t given that the derailment has not occurred until at least time t by dividing the likelihood of the derailment occurring at time t by the likelihood that the segment will survive until time t;
generating a repair decision for each of the geo-defects based on one or more of the predicting, the calculated likelihood of derailment, or the determined instantaneous rate of failure probability; and
selectively repairing one or more of the geo-defects based on the repair decision.

2. The method of claim 1 , wherein the segments of the railroad network are of equal length.

3. The method of claim 1 , wherein the segments do not overlap.

4. The method of claim 1 , wherein the predicting includes processing only the geo-defects having a rate of increase in amplitude at a defined threshold value resulting from the calculating.

5. The method of claim 1 , wherein a geo-defect identified at a current inspection run is determined to be the same geo-defect as a geo-defect identified at a previous inspection run, despite the geo-defect of the current inspection run being documented as residing at a location that is different than a location of the geo-defect of the previous inspection run, when the geo-defect of the current inspection run:
is determined to have a common geo-defect type as the geo-defect type of the geo-defect identified for the previous inspection run;
has an equal or greater amplitude than the geo-defect of the previous inspection run; and
has a location proximity within a predetermined distance of the geo-defect of the previous inspection run.

6. The method of claim 1 , wherein the sensors are configured to measure qualitative aspects of the components of the railroad network and ambient conditions, the components including tracks and cars.

7. The method of claim 6 , wherein the sensors include at least one of:
machine vision detectors (MVDs);
wheel impact load detectors (WILDs);
optical geometry detectors (OCDs);
truck performance detectors (TPDs);
acoustic bay detectors (ABDs);
hot box detectors;
warm bearing detectors; and
hot wheel/cold wheel detectors.

8. The method of claim 1 , wherein the geo-defect types include at least two of: align;
cant;
dip;
gauge;
harmonic cross-level;
over-elevation;
reverse cross-level;
super cross-level elevation;
surf;
twist;
warp; and
wear.

9. A computer program product comprising a computer-readable storage medium having program code embodied thereon, which when executed by a computer processor, causes the computer processor to implement a method, the method comprising:
logically dividing a railroad network into segments each of a specified length, the railroad network comprising a railroad network infrastructure that includes a plurality of data sources, the plurality of data sources comprising sensors configured to monitor operational aspects of components of the railroad network;
collecting data from the plurality of data sources with respect to monitored operational aspects of the components;
identifying geo-defects and approximated locations of the geo-defects occurring at each inspection run for each of the segments;
resolving location uncertainty of the approximated locations of each of the geodefects based on geo-defect type, geo-defect amplitude, and location proximity for corresponding geo-defects assessed at each inspection run, wherein the corresponding geo-defects are determined by matching identified geo-defects from multiple inspection runs within each segment of a specified length;
calculating a rate of increase in amplitude of each of the geo-defects for each of the segments between inspection runs; and
predicting a deterioration rate for each of the geo-defects based on the calculating;
upon determining multiple geo-defects exist for one of the segments at one of the inspection runs and at least one geo-defect exists for the one of the segments at another of the inspection runs, aggregating the multiple geo-defects to reflect the single geo-defect;
determining an instantaneous rate of failure probability, the determining of the instantons rate of failure probability comprising for each of the segments:
calculating a likelihood of the derailment occurring at a time t;
calculating a likelihood that the segment will survive until time t; and
calculating the instantaneous rate of failure probability at time t given that the derailment has not occurred until at least time t by dividing the likelihood of the derailment occurring at time t by the likelihood that the segment will survive until time t;
generating a repair decision for each of the geo-defects based on one or more of the predicting, the calculated likelihood of derailment, or the determined instantaneous rate of failure probability; and
selectively repairing one or more of the geo-defects based on the repair decision.

10. The computer program product of claim 9 , wherein the segments of the railroad network are of equal length.

11. The computer program product of claim 9 , wherein the segments do not overlap.

12. The computer program product of claim 9 , wherein the predicting includes processing only the geo-defects having a rate of increase in amplitude at a defined threshold value resulting from the calculating.

13. The computer program product of claim 9 , wherein a geo-defect identified at a current inspection run is determined to be the same geo-defect as a geo-defect identified at a previous inspection run, despite the geo-defect of the current inspection run being documented as residing at a location that is different than a location of the geo-defect of the previous inspection run, when the geo-defect of the current inspection run:
is determined to hav

Assignees

• IBM

Inventors

• Hampapur Arun
• Li Hongfei
• Parikh Dhaivat P

Classifications

• G06F30/20

  Design optimisation, verification or simulation (optimisation, verification or simulation of circuit designs G06F30/30) · CPC title

• B61L23/042

  Track changes detection · CPC title

• B61L23/047

  Track or rail movements · CPC title

• G06Q10/0635

  Risk analysis of enterprise or organisation activities · CPC title

• B61L23/044

  Broken rails · CPC title