METHOD AND SYSTEM FOR DETECTING SYMPTOMS AND DETERMINING AN OPTIMAL REMEDY PATTERN FOR A FAULTY DEVICE

• US 20130305081A1
• Filed: 05/09/2013
• Published: 11/14/2013
• Est. Priority Date: 05/09/2012
• Status: Active Grant

• Current Assignee: Infosys Limited
• Original Assignee: Infosys Limited
• Inventors: Pisipati, Radha Krishna, Agnihotram, Gopichand
• Granted Patent: US 9,063,856 B2
• Time in Patent Office: Days
• US Class Current: 714/2
• CPC Class Codes: G05B 23/02 Electric testing or monitoring
• CPC Class Codes: G05B 23/0262 Confirmation of fault detection, e.g. extra checks to confirm that a failure has indeed occurred
• CPC Class Codes: G05B 23/0286 Modifications to the monitored process, e.g. stopping operation or adapting control
• CPC Class Codes: G06F 11/0739 in a data processing system embedded in automotive or aircraft systems
• CPC Class Codes: G06F 11/0751 Error or fault detection not based on redundancy power supply failures G06F1/30; network fault management H04L41/06
• CPC Class Codes: G06F 11/079 Root cause analysis, i.e. error or fault diagnosis in a hardware test environment G06F11/22; in a software test environment G06F11/36
• CPC Class Codes: G06F 11/0793 Remedial or corrective actions recovery from an exception in an instruction pipeline G06F9/3861; by retry G06F11/1402; for recovering from a failure of a protocol instance or entity H04L69/40
• CPC Class Codes: G07C 5/0808 Diagnosing performance data testing of vehicles G01M17/00; testing of electrical installation on vehicles G01R31/005

Abstract

Computer-implemented systems, methods, and computer-readable media electronic for detecting symptoms and determining an optimal remedy pattern for one or more faulty components of a device is disclosed. First the symptoms of the faulty device are detected and associated faulty components of the device are identified. Different tests are performed to confirm the status of the faulty components. Based on the historical data, cost information and remedy cost function an optimal remedy pattern is determined.

Claims

1. A computer implemented method executed by one or more computing devices for detecting one or more symptoms and determining an optimal remedy pattern for one or more faulty components of a faulty device, the method comprising:

detecting the one or more symptoms in the faulty device that are occurring due to the one or more faulty components in the faulty device;

identifying the one or more faulty components from a plurality of components of the system based on the detected one or more symptoms;

performing one or more tests for each identified faulty component to confirm the one or more faulty components and to check status of the faulty device;

extracting historical data stored in a database; and

determining the optimal remedy pattern for the identified one or more faulty components based on;

the extracted historical data;

cost information;

a remedy cost without testing; and

a remedy cost with testing.

2. The method of claim 1, further comprising

remedying the one or more faulty components following the optimal remedy pattern.

3. The method of claim 2, further comprising:

checking the status of the faulty device after remedying every faulty component among the one or more faulty components following the optimal remedy pattern.

4. The method of claim 1, wherein the historical data comprises:

the one or more symptoms;

the one or more faulty components associated with each of the symptoms;

one or more remedy patterns corresponding to the one or more symptoms; and

the optimal remedy pattern used on the one or more faulty components.

5. The method of claim 1, wherein the cost information comprises one or more of:

a renew cost;

an observation cost for the plurality of components;

an observation cost of the faulty device; and

an investigation cost.

6. The method of claim 1, wherein the one or more symptoms are detected based on sensor data associated with the components of the faulty device using one or more measurement error models.

7. A computer implemented method executed by one or more computing devices for identifying one or more faulty components of a faulty device, the method comprising:

obtaining one or more control samples of the faulty device data when the faulty device operates in a no-fault mode;

obtaining sensor data under condition monitoring;

performing exploratory data analysis on the one or more control samples and the sensor data, wherein the sensor data is obtained corresponding to each component of the faulty device when the faulty device operates in a fault mode;

using measurement error model on the sensor data;

describing the sensor data using a multivariate probability distribution wherein one or more parameters of the multivariate probability distribution are estimated by a maximum-likelihood estimation technique;

comparing the sensor data with the control samples using likelihood ratio test; and

determining the sensor data that indicates the one or more faulty components.

8. A computer implemented method executed by one or more computing devices for determining an optimal remedy pattern for one or more faulty components of a faulty device, the method comprising:

receiving one or more symptoms identified in the faulty device;

identifying one or more faulty components associated with each of the one or more symptoms;

determining status of the faulty device by a probabilistic engine;

computing a remedy cost with testing and a remedy cost without testing based on cost information of the system;

determining the optimal remedy pattern using the remedy cost with testing, provided the remedy cost with testing is less than the remedy cost without testing;

determining the optimal remedy pattern using the remedy cost without testing, provided the remedy cost with testing is more than the remedy cost without testing; and

updating historical data of the probabilistic engine with the determined optimal remedy pattern.

9. The method of claim 8, wherein the cost information comprises one or more of:

a renew cost;

an observation cost for the plurality of components;

an observation cost of the faulty device; and

an investigation cost.

10. The method of claim 8, wherein the probabilistic engine is configured to store information comprising:

one or more faulty device details;

domain knowledge; and

the historical data of one or more components of the faulty device.

11. The method of claim 8, further comprising:

remedying the one or more faulty components following the optimal remedy pattern.

12. A remedy system for detecting symptoms and determining optimal remedy pattern for one or more faulty components of a faulty device, the remedy system comprising:

a symptoms detection and analysis module configured to;

detect one or more symptoms in the faulty device that are occurring due to the one or more faulty components in the faulty device;

determine the one or more faulty components from a plurality of components of the faulty device based on the detected one or more symptoms;

provide the detected one or more symptoms and the determined one or more faulty components to an optimal remedy pattern module;

an activity discovery module configured to;

extract historical data of the faulty device stored in a database and send the historical data to the optimal remedy pattern module;

a testing module configured to;

perform one or more tests for each determined faulty component to confirm the one or more faulty components and to check status of the faulty device; and

the optimal remedy pattern module configured to;

determine the optimal remedy pattern for the detected one or more symptoms based on;

the extracted historical data;

a cost information;

a remedy cost without testing; and

a remedy cost with testing.

13. The remedy system of claim 12, wherein the remedy system further comprises:

an action module configured to;

remedy the one or more faulty components following the optimal remedy pattern.

14. The remedy system of claim 13, wherein the action module is further configured to:

check the status of the faulty device after remedying every faulty component among the one or more faulty components following the optimal remedy pattern.

15. A symptoms detection and analysis system comprising:

a control sample database configured to;

store one or more control samples of the faulty device data when the system operates in a no-fault mode;

an exploratory data analysis module configured to;

perform exploratory data analysis on the one or more control samples and sensor data wherein the sensor data is obtained corresponding to each component of the faulty device when the faulty device operates in a fault mode;

a condition monitoring module configured to;

obtain the sensor data under condition monitoring;

a measurement error model module configured to;

use measurement error model on the sensor data;

a maximum likelihood estimator configured to;

describe the sensor data using a multivariate probability distribution wherein one or more parameters of the multivariate probability distribution are estimated by a maximum-likelihood estimation technique;

a likelihood ratio testing module configured to;

compare the sensor data with the control samples using a likelihood ratio test; and

a determining module configured to;

determine the sensor data that indicates the one or more faulty components.

16. A optimal remedy pattern system comprising:

a probabilistic engine configured to;

receive one or more symptoms identified in a faulty device;

identify one or more faulty components associated with the each symptom;

determine the faulty device status;

a calculating module configured to;

compute a remedy cost with testing and a remedy cost without testing based on cost information of the faulty device;

a processing module configured to;

determine the optimal remedy pattern using the remedy cost with testing provided the remedy cost with testing is less than the remedy cost without testing; and

determine the optimal remedy pattern using the remedy cost without testing provided the remedy cost with testing is more than the remedy cost without testing; and

a remedying module configured to;

update historical data of the probabilistic engine with the determined optimal remedy pattern.

17. The optimal remedy pattern system of claim 16, wherein the cost information comprises one or more of:

a renew cost;

an observation cost for the plurality of components;

an observation cost of the faulty device; and

an investigation cost.

18. The optimal remedy pattern system of claim 16, wherein the probabilistic engine is configured to store information comprising:

one or more faulty device details;

domain knowledge; and

the historical data of one or more components of the faulty device.

19. The optimal remedy pattern system of claim 16, wherein the remedying module is further configured to:

remedy the one or more faulty components following the optimal remedy pattern.

20. A computer program product comprising a computer usable medium having a computer readable program code embodied therein for detecting one or more symptoms and determining an optimal remedy pattern for one or more faulty components of a faulty device, the product comprising:

program code adapted for detecting the one or more symptoms in the faulty device that are occurring due to the one or more faulty components in the faulty device;

program code adapted for identifying the one or more faulty components from a plurality of components of the system based on the detected one or more symptoms;

program code adapted for performing one or more tests for each identified faulty component to confirm the one or more faulty components and to check status of the faulty device;

program code adapted for extracting historical data stored in a database; and

program code adapted for determining the optimal remedy pattern for the identified one or more faulty components based on;

the extracted historical data;

cost information;

a remedy cost without testing; and

a remedy cost with testing.

21. The product of claim 20, wherein the historical data comprises:

the one or more symptoms;

the one or more faulty components associated with each of the symptoms;

one or more remedy patterns corresponding to the one or more symptoms; and

the optimal remedy pattern used on the one or more faulty components.

22. The product of claim 20, further comprising:

program code adapted for remedying the one or more faulty components following the optimal remedy pattern.

23. The product of claim 20, further comprising:

program code adapted for checking the status of the faulty device after remedying every faulty component among the one or more faulty components following the optimal remedy pattern.

24. The product of claim 20, wherein the cost information comprises one or more of:

a renew cost;

an observation cost for the plurality of components;

an observation cost of the faulty device; and

an investigation cost.

25. The product of claim 20, wherein the one or more symptoms are detected based on sensor data associated with the components of the faulty device using one or more measurement error models.

26. A computer program product comprising a computer usable medium having a computer readable program code embodied therein for identifying one or more faulty components of a faulty device, the product comprising:

program code adapted for obtaining one or more control samples of the faulty device data when the faulty device operates in a no-fault mode;

program code adapted for obtaining sensor data under condition monitoring;

program code adapted for performing exploratory data analysis on the one or more control samples and the sensor data, wherein the sensor data is obtained corresponding to each component of the faulty device when the faulty device operates in a fault mode;

program code adapted for using measurement error model on the sensor data;

program code adopted for describing the sensor data using a multivariate probability distribution wherein one or more parameters of the multivariate probability distribution are estimated by a maximum-likelihood estimation technique;

program code adapted for comparing the sensor data with the control samples using likelihood ratio test; and

program code adapted for determining the sensor data that indicates the one or more faulty components.

27. A computer program product comprising a computer usable medium having a computer readable program code embodied therein for determining an optimal remedy pattern for one or more faulty components of a faulty device, the product comprising:

program code adapted for receiving one or more symptoms identified in the faulty device;

program code adapted for identifying one or more faulty components associated with each of the one or more symptoms;

program code adapted for determining status of the faulty device by a probabilistic engine;

program code adapted for computing a remedy cost with testing and a remedy cost without testing based on cost information of the system;

program code adopted for determining the optimal remedy pattern using the remedy cost with testing, provided the remedy cost with testing is less than the remedy cost without testing;

program code adapted for determining the optimal remedy pattern using the remedy cost without testing, provided the remedy cost with testing is more than the remedy cost without testing; and

program code adapted for updating historical data of the probabilistic engine with the determined optimal remedy pattern.

28. The product of claim 27, further comprising:

remedying the one or more faulty components following the optimal remedy pattern.

29. The product of claim 27, wherein the cost information comprises one or more of:

a renew cost;

an observation cost for the plurality of components;

an observation cost of the faulty device; and

an investigation cost.

30. The product of claim 27, wherein the probabilistic engine is configured to store information comprising:

one or more faulty device details;

domain knowledge; and

the historical data of one or more components of the faulty device.