System and method for maintenance of a fleet of machines

1. A method for maintenance of a machine among a fleet of machines, comprising:

receiving a service request corresponding to a machine, wherein the service request comprises a description of a fault condition in the machine;

retrieving log record data corresponding to the fleet of machines, wherein the log record data comprises a plurality of historical service requests, a plurality of recommendations corresponding to each of the plurality of historical service requests, system logs comprising a plurality message codes, a part list comprising part descriptions, or a combination thereof;

obtaining a service architecture corresponding to the fleet of machines, wherein the service architecture comprises a service dictionary, a plurality of classification schemes organized in a tree data structure, and a part dictionary comprising a plurality of part cluster names and corresponding plurality of part cluster descriptions, wherein the part dictionary is based on the plurality of historical service requests, the plurality of recommendations, and the part list;

processing the service request based on the service dictionary and a text parsing technique to generate a list of descriptive words;

generating a recommendation based on the list of descriptive words and the service architecture, wherein the recommendation comprises at least one of an on-line repair activity, an on-site repair activity and a part replacement activity; and

servicing the fault condition of the machine based on the recommendation.

2. The method of claim 1 , wherein the service dictionary comprises a plurality of descriptive key-words and a corresponding plurality of message codes extracted from system logs.

3. The method of claim 1 , wherein processing the service request comprises:

removing special characters, proper nouns and dates from the service request to generate the list of descriptive words; and

creating a first binary vector based on the service dictionary and list of descriptive words corresponding to the service request, wherein the first binary vector is representative of presence of descriptive words of the list of descriptive words in the service dictionary.

4. The method of claim 3 , wherein generating the recommendation comprises:

selecting a classification scheme among the plurality of classification schemes based on the first binary vector and a deep learning model; and

evaluating a machine learning model corresponding to the classification schemes based on the first binary vector.

5. The method of claim 1 , wherein obtaining the service architecture comprises:

determining the service dictionary based on the plurality of historical service requests, the plurality of recommendations and the system logs;

and

determining the plurality of classification schemes based on the log record data, the service dictionary and the part dictionary.

6. The method of claim 5 , wherein determining the service dictionary comprises:

determining a plurality of descriptive key-words from the plurality of historical service requests using a text processing technique;

extracting a plurality of message codes from the system logs, corresponding to the plurality of historical service requests; and

associating the plurality of message codes with the plurality of descriptive key-words in the service dictionary.

7. The method of claim 5 , wherein determining the parts dictionary comprises:

determining part names corresponding to a plurality of historical service requests;

determining a term frequency-inverse document frequency parameter for each of the part names;

determining a distance between a pair of service requests using cosine similarity metric;

clustering a plurality of part names based on a density-based spatial clustering of applications with noise technique.

8. The method of claim 7 , wherein generating the parts dictionary comprises clustering a plurality of co-occurring parts from association rules based on a confidence value and a lift value.

9. The method of claim 5 , wherein obtaining the service architecture comprises:

determining a first classification scheme and a second classification scheme among the plurality of classification schemes based on the log record data, the service dictionary and the part dictionary, wherein the first classification scheme is a binary classification scheme configured to provide one of a remote servicing option and a field servicing option, and wherein the second classification scheme is a multi-label classification scheme configured to provide a plurality of off-site repair recommendations to be performed.

10. The method of claim 9 , wherein obtaining the service architecture comprises determining a third classification scheme, a fourth classification scheme and a fifth classification scheme among the plurality of classification schemes based on the log record data, the service dictionary and the part dictionary, wherein the third classification scheme is a binary classification scheme configured to provide one of a part replacement option and a repairing option, wherein the fourth classification scheme is a multi-label classification scheme configured to provide a list of parts to be replaced, and wherein the fifth classification scheme is a multi-label classification scheme configured to provide a list of on-site repair recommendations.

11. A system for maintenance of a machine among a fleet of machines, comprising:

a memory unit configured to:

store a request database comprising a service request corresponding to the machine, wherein the service request comprises a description of a fault condition in the machine;

retrieve log record data corresponding to the fleet of machines, wherein the log record data comprises a plurality of historical service requests, a plurality of recommendations corresponding to each of the plurality of historical service requests, system logs comprising a plurality message codes, a part list comprising part descriptions, or a combination thereof; and

store a service architecture corresponding to the fleet of machines, wherein the service architecture comprises a service dictionary, a plurality of classification schemes organized in a tree data structure, and a part dictionary comprising a plurality of part cluster names and corresponding plurality of part cluster descriptions; and

a machine learning unit communicatively coupled to the memory unit and configured to:

determine the service dictionary based on the plurality of historical service requests, the plurality of recommendations, and the system logs; and

determine the part dictionary based on the plurality of historical service requests, the plurality of recommendations, and the part list;

a text processing unit communicatively coupled to the memory unit and configured to process the service request based on the service dictionary and a text parsing technique to generate a list of descriptive words;

a recommendation unit communicatively coupled to the text processing unit and configured to generate a recommendation based on the list of descriptive words and the service architecture, wherein the recommendation comprises at least one of an on-line repair activity, an on-site repair activity and a part replacement activity; and

a controller unit communicatively coupled to the recommendation unit and configured to service the fault condition of the machine based on the recommendation.

12. The system of claim 11 , wherein the service dictionary comprises a plurality of descriptive key-words and a corresponding plurality of message codes extracted from system logs.

13. The system of claim 11 , wherein the text processing unit is further configured to:

remove special characters, proper nouns and dates from the service request to generate a list of descriptive words; and

create a first binary vector based on the service dictionary and list of descriptive words corresponding to service request, wherein the first binary vector is representative of presence of descriptive words in the service dictionary.

14. The system of claim 13 , wherein the recommendation unit is further configured to:

select a classification scheme among the plurality of classification schemes based on the first binary vector and a deep learning model; and

evaluate a machine learning model corresponding to the classification scheme based on the first binary vector.

15. The system of claim 11 ,

wherein the machine learning unit is further configured to:

determine the plurality of classification schemes based on the log record data, the service dictionary and the part dictionary.

16. The system of claim 15 , wherein the machine learning unit is further configured to:

determine part names corresponding to a plurality of service requests;

determine a term frequency-inverse document frequency parameter for each of the part names;

determine a distance between a pair of service requests using cosine similarity metric;

cluster a plurality of part names based on Density-based spatial clustering of applications with noise.

17. The system of claim 16 , wherein the machine learning unit is further configured to cluster a plurality of co-occurring parts from association rules based on a confidence value and a lift value.

18. The system of claim 17 , wherein the machine learning unit is configured to determine a first classification scheme and a second classification scheme among the plurality of classification schemes, wherein the first classification scheme is a binary classification scheme configured to provide one of a remote servicing option and a field servicing option and wherein the second classification scheme is a multi-label classification scheme configured to provide a plurality of off-site repair recommendations to be performed.