Model development framework for remote monitoring condition-based maintenance
Embodiments include techniques for developing a model framework for remote unit monitoring condition-based maintenance. The techniques include collecting data associated with unplanned service requests, and generating one or more models from the collected data. The techniques also include predicting unplanned service requests based at least in part on the one or more models, and transmitting an output of the prediction of the unplanned service request.
1. A system for developing a model framework for remote unit monitoring condition-based maintenance, the system comprising:
a processor and a storage medium coupled to the processor;
the processor configured to:
execute one or more models operable to perform operations comprising predicting an unplanned service request;
wherein the one or more models are trained to perform the operations by identifying an actual unplanned service request that has occurred, collecting unplanned service request data limited to a time interval leading up to the actual unplanned service request, wherein the unplanned service request data limited to the time interval leading up to the actual unplanned service request comprises a set of data indicating a pattern of operational symptoms predictive of the unplanned service request, and using the unplanned service request data to train the one or more models to perform the operations;
wherein performing the operations comprise generating a predicted unplanned service request based at least in part on the one or more models; and
transmit an output of the predicted unplanned service request.
2. The system of claim 1 , wherein the unplanned service request data comprises at least one of unit data, performance data, or service records.
3. The system of claim 2 , wherein the unplanned service request data comprises at least two of the unit data, performance data, or service records.
4. The system of claim 2 , wherein the unplanned service request data comprises the unit data, performance data, and service records.
5. The system of claim 1 , wherein the actual unplanned service request is for a unit, wherein the unit is at least one of an elevator unit or an escalator unit.
6. The system of claim 2 , wherein the one or more models are generated using a supervised machine-learning process.
7. The system of claim 4 , wherein generating the one or more models comprises generating a health score for the unit.
8. The system of claim 1 , wherein the processor is configured to select a model of the one or more models, and test the selected model based on an independent dataset from the unplanned service request data.
9. The system of claim 1 , the output comprises at least one of a unit ID, an estimated score related to the probability of a request for service request, and one or more features that contributed to the estimated score.
10. The system of claim 9 , wherein the output is used to update a maintenance schedule.
11. A method for developing a model framework for remote unit monitoring condition-based maintenance, the method comprising:
executing one or more models operable to perform operations comprising predicting an unplanned service request;
wherein the one or more models are trained to perform the operations by identifying an actual unplanned service request that has occurred, collecting unplanned service request data limited to a time interval leading up to the actual unplanned service request, wherein the unplanned service request data limited to the time interval leading up to the actual unplanned service request comprises a set of data indicating a pattern of operational symptoms predictive of the unplanned service request, and using the unplanned service request data to train the one or more models to perform the operations;
wherein performing the operations comprise generating a predicted unplanned service requests based at least in part on the one or more models; and
transmitting an output of the predicted unplanned service request.
12. The method of claim 11 , wherein the unplanned service request data comprises at least one of unit data, performance data, or service records.
13. The method of claim 12 , wherein the unplanned service request data comprises at least two of the unit data, performance data, or service records.
14. The method of claim 12 , wherein the unplanned service request data comprises unit data, performance data, and service records.
15. The method of claim 11 , wherein the actual unplanned service request is for a unit, wherein the unit is at least one of an elevator unit or an escalator unit.
16. The method of claim 12 , wherein the one or more models are generated using a supervised machine-learning process.
17. The method of claim 14 , wherein generating the one or more models comprises generating a health score for the unit.
18. The method of claim 11 , further comprising selecting a model of the one or more models, and testing the selected model based on an independent dataset from the unplanned service request data.
19. The method of claim 11 , wherein the output comprises at least one of a unit ID, an estimated score related to the probability of a request for service request, and one or more features that contributed to the estimated score.
20. The method of claim 19 , further comprises updating a maintenance schedule based at least in part on the output.