Claims:WE CLAIM:
1. A method of predicting failures in a diverse set of asset types in an enterprise, the method comprising:
receiving, by a failure prediction system (105), asset information (211) related to one or more assets (101) from one or more data sources (103) associated with the one or more assets (101), wherein the one or more assets (101) belong to one or more asset types;
determining, by the failure prediction system (105), an asset type and a failure mode of each of the one or more assets (101) based on analysis of the asset information (211);
selecting, by the failure prediction system (105), one of a plurality of prediction models (107) for predicting failures in each of the one or more assets (101) based on the asset type and the failure mode of each of the one or more assets (101); and
analysing, by the failure prediction system (105), the asset information (211) using selected one of the plurality of prediction models (107) for predicting the failures in each of the one or more assets (101).

2. The method as claimed in claim 1, wherein the asset information (211) comprises at least one of an asset identifier corresponding to each of the one or more assets (101), values of operating parameters of the one or more assets (101), values of operating parameters of an ambient environment of the one or more assets (101), events indicating changes in operational states of the one or more assets (101) and alarms indicating variations in operation of the one or more assets (101).

3. The method as claimed in claim 1, wherein the asset information (211) is received in real-time or at predetermined periodical intervals.

4. The method as claimed in claim 1, wherein the one or more data sources (103) comprises one or more sensors configured with the one or more assets (101) and data logs indicating operation, maintenance and servicing of the one or more assets (101).

5. The method as claimed in claim 1, wherein receiving the asset information (211) further comprises:
performing one or more data cleansing operations on the asset information (211) for eliminating one or more irregularities in the asset information (211); and
determining sufficiency of the asset information (211) based on comparison of the asset information (211) with one or more predetermined asset parameters.

6. The method as claimed in claim 1, wherein determining the asset type and the failure mode of each of the one or more assets (101) comprises:
determining an asset identifier corresponding to the each of the one or more assets (101) using the asset information (211); and
determining the asset type and the failure mode of each of the one or more assets (101) based on the asset identifier and asset metadata (213) stored in a metadata store associated with the failure prediction system (105).

7. The method as claimed in claim 1, wherein selecting the one of the plurality of prediction models (107) for predicting failures in each of the one or more assets (101) comprises:
comparing the asset type and the failure mode of each of the one or more assets (101) with pretrained asset type and pretrained failure mode used for training each of the plurality of prediction models (107); and
selecting one of the plurality of prediction models (107) for each of the one or more assets (101) based on comparison.

8. The method as claimed in claim 7 further comprises dynamically crating new prediction models (107) for predicting failures in the one or more assets (101) when the asset type and the failure mode of the one or more assets (101) do not match with the pretrained asset type and the pretrained failure mode of the plurality of prediction models (107), wherein each of the new prediction models (107) are stored in a model repository associated with the failure prediction system (105) for subsequent prediction of failures in the one or more assets (101).

9. The method as claimed in claim 1 further comprises training the plurality of prediction models (107) for predicting failures in the one or more assets (101) when prediction accuracy level of the plurality of prediction models (107) is less than a predetermined threshold.

10. The method as claimed in claim 1 further comprises generating and transmitting one or more notification events to asset management personnel, associated with the one or more assets (101), upon predicting failures in the one or more assets (101).

11. The method as claimed in claim 1, wherein each of the one or more asset types is associated with a plurality of prediction models (107) and each of the plurality of prediction models (107) is associated with a predetermined failure mode.

12. A failure prediction system (105) for predicting failures in a diverse set of asset types in an enterprise, the failure prediction system (105) comprising:
a processor (203); and
a memory (205), communicatively coupled to the processor (203), wherein the memory (205) stores processor-executable instructions, which on execution, cause the processor (203) to:
receive asset information (211) related to one or more assets (101) from one or more data sources (103) associated with the one or more assets (101), wherein the one or more assets (101) belong to one or more asset types;
determine an asset type and a failure mode of each of the one or more assets (101) based on analysis of the asset information (211);
select one of a plurality of prediction models (107) to predict failures in each of the one or more assets (101) based on the asset type and the failure mode of each of the one or more assets (101); and
analyse the asset information (211) using selected one of the plurality of prediction models (107) to predict the failures in each of the one or more assets (101).

13. The failure prediction system (105) as claimed in claim 12, wherein the asset information (211) comprises at least one of an asset identifier corresponding to each of the one or more assets (101), values of operating parameters of the one or more assets (101), values of operating parameters of an ambient environment of the one or more assets (101), events indicating changes in operational states of the one or more assets (101) and alarms indicating variations in operation of the one or more assets (101).

14. The failure prediction system (105) as claimed in claim 12, wherein the processor (203) receives the asset information (211) in real-time or at predetermined periodical intervals.

15. The failure prediction system (105) as claimed in claim 12, wherein the one or more data sources (103) comprises one or more sensors configured with the one or more assets (101) and data logs indicating operation, maintenance and servicing of the one or more assets (101).

16. The failure prediction system (105) as claimed in claim 12, wherein the processor (203) is further configured to:
perform one or more data cleansing operations on the asset information (211) to eliminate one or more irregularities in the asset information (211); and
determine sufficiency of the asset information (211) based on comparison of the asset information (211) with one or more predetermined asset parameters.

17. The failure prediction system (105) as claimed in claim 12, wherein to determine the asset type of each of the one or more assets (101), the processor (203) is configured to:
determine an asset identifier corresponding to the each of the one or more assets (101) using the asset information (211); and
determine the asset type and the failure mode of each of the one or more assets (101) based on the asset identifier and asset metadata (213) stored in a metadata store associated with the failure prediction system (105).

18. The failure prediction system (105) as claimed in claim 12, wherein to select the one of the plurality of prediction models (107) for predicting failures in each of the one or more assets (101), the processor (203) is configured to:
compare the asset type and the failure mode of each of the one or more assets (101) with pretrained asset type and pretrained failure mode used for training each of the plurality of prediction models (107); and
select one of the plurality of prediction models (107) for each of the one or more assets (101) based on comparison.

19. The failure prediction system (105) as claimed in claim 18, wherein the processor (203) is configured to dynamically crate new prediction models (107) for predicting failures in the one or more assets (101) when the asset type and the failure mode of the one or more assets (101) do not match with the pretrained asset type and the pretrained failure mode of the plurality of prediction models (107), wherein the processor (203) stores each of the new prediction models (107) in a model repository associated with the failure prediction system (105) for subsequent prediction of failures in the one or more assets (101).

20. The failure prediction system (105) as claimed in claim 12, wherein the processor (203) is configured to train the plurality of prediction models (107) for predicting failures in the one or more assets (101) when prediction accuracy level of the plurality of prediction models (107) is less than a predetermined threshold.

21. The failure prediction system (105) as claimed in claim 12, wherein the processor (203) is configured to generate and transmit one or more notification events to asset management personnel, associated with the one or more assets (101), upon predicting failures in the one or more assets (101).

22. The failure prediction system (105) as claimed in claim 12, wherein the processor (203) associates each of the one or more asset types with a plurality of prediction models (107) and each of the plurality of prediction models (107) with a predetermined failure mode.

Dated this 25th day of January 2019

MADHUSUDAN S T
OF K&S PARTNERS
ATTORNEY FOR THE APPLICANT
IN/PA -1297
, Description:TECHNICAL FIELD
The present subject matter is, in general, related to enterprise asset management and more particularly, but not exclusively, to a method and system for predicting failures in a diverse set of asset types in an enterprise.