Predictive Maintenance

Abstract

Vehicular maintenance is predicted using real time telematics data and historical maintenance data. Different statistical models are used, and an intersecting set of results is generated. Environmental weather may also be used to further refine predictions.

1 . A method, comprising:
 obtaining, by a vehicle controller comprising a processor, at least two different datasets associated with a vehicle from a group of datasets comprising a sensor dataset, a maintenance dataset, and a weather dataset;   predicting, by the vehicle controller, using respective different machine learning models for the at least two different datasets, respective groups of components of the vehicle for maintenance; and   predicting, by the vehicle controller, using an additional machine learning model on an intersecting set of components of the respective groups of components, a final component of the intersecting set of components of the vehicle for maintenance.   
     
     
         2 . The method of  claim 1 , wherein the additional machine learning model is different from the respective different machine learning models. 
     
     
         3 . The method of  claim 1 , further comprising wirelessly sending, by the vehicle controller, the prediction of the final component of the vehicle for maintenance to a remote server. 
     
     
         4 . The method of  claim 1 , further comprising generating, by the vehicle controller, a webpage by the vehicle controller, the webpage describing the prediction of the final component of the vehicle for maintenance. 
     
     
         5 . The method of  claim 1 , further comprising generating, by the vehicle controller, a graphical user interface for a display of the prediction of the final component of the vehicle for maintenance. 
     
     
         6 . The method of  claim 1 , further comprising recommending, by the vehicle controller, a service appointment associated with the prediction of the final component of the vehicle for maintenance. 
     
     
         7 . The method of  claim 1 , wherein at least one of the respective different machine learning models comprises a random forest process. 
     
     
         8 . A system, comprising:
 a processor; and   a memory, coupled to the processor, that stores executable instructions, that when executed by the processor, facilitate performance of operations, comprising:
 retrieving different datasets associated with a vehicle from a group of datasets comprising a sensor dataset, a maintenance dataset, and a weather dataset; 
 determining, based on respective outputs from applying respective different machine learning models to the at least two different datasets, respective components of the vehicle for maintenance; and 
 generating, based on an output of applying an additional machine learning model to an intersecting set of components of the respective components, a prediction for at least one final component of the vehicle that implicates maintenance. 
   
     
     
         9 . The system of  claim 8 , wherein the respective different machine learning models do not comprise the additional machine learning model. 
     
     
         10 . The system of  claim 8 , wherein the operations further comprise generating network-accessible content describing the prediction for the at least one final component of the vehicle. 
     
     
         11 . The system of  claim 10 , wherein the operations further comprise publishing the network-accessible content describing the prediction for the at least one final component of the vehicle. 
     
     
         12 . The system of  claim 8 , wherein the operations further comprise causing a user interface to be rendered describing the prediction for the at least one final component of the vehicle. 
     
     
         13 . The system of  claim 12 , wherein causing the user interface to be rendered comprises causing at least one of a video, an image or audio describing the prediction to be rendered in the vehicle. 
     
     
         14 . The system of  claim 8 , wherein at least one of the respective different machine learning models is configured to generate an output that at least in part applies a random forest process. 
     
     
         15 . The system of  claim 8 , wherein the operations further comprise recommending a specified service appointment associated with the maintenance of the at least one final component of the vehicle. 
     
     
         16 . A non-transitory computer-readable medium having instructions stored thereon that, when executed by a processor, facilitate performance of operations, comprising:
 retrieving different datasets associated with a vehicle from at least two of a sensor dataset, a maintenance dataset, and a weather dataset;   identifying, using respective first machine learning models on the different datasets, respective intermediate sets of components of the vehicle for which maintenance is potentially to be performed, wherein the respective first machine learning models are different from one another; and   generating, using a second machine learning model on an intersecting set of components comprising an intersection of the respective intermediate sets of components, a prediction for a final set of components of the vehicle for which maintenance is to be performed.   
     
     
         17 . The non-transitory computer-readable medium of  claim 16 , wherein the second machine learning model is different from the respective first machine learning models. 
     
     
         18 . The non-transitory computer-readable medium of  claim 16 , wherein the operations further comprise displaying a user interface describing the prediction via a display device that is part of the vehicle. 
     
     
         19 . The non-transitory computer-readable medium of  claim 16 , wherein using of the respective first machine learning models comprises using at least one of the respective first machine learning models that applies a random forest classifier. 
     
     
         20 . The non-transitory computer-readable medium of  claim 16 , wherein the operations further comprise recommending a time and a place for the final set of components of the vehicle to be maintained according to the prediction.