| CPC G08G 1/20 (2013.01) [G06Q 10/06 (2013.01); G07C 5/0816 (2013.01); G07C 5/085 (2013.01)] | 18 Claims |
|
1. A Machine Learning (ML)-based computer-implemented system for remote monitoring and maintaining a fleet of electric vehicles comprising:
one or more sensors configured to collect vehicle data associated with each electric vehicle in the fleet of electric vehicles, wherein each of the electric vehicle within the fleet of electric vehicles is located at distinct and instantaneously changing locations; and
a server comprising:
one or more processors; and
a memory coupled to the one or more processors, wherein the memory comprises a set of program instructions in form of a plurality of modules, configured to be executed by the one or more processors, wherein the plurality of modules comprises:
an electric vehicle data tracking module configured to track the vehicle data associated with each of the electric vehicle in the fleet of electric vehicles collected via the one or more sensors, wherein the one or more sensors are deployed in each of the fleet of the electric vehicles and associated charging infrastructure;
a fleet vehicle profile management module operatively coupled to the electric vehicle data tracking module, wherein the fleet vehicle profile management module is configured to:
generate a unique vehicle profile corresponding to each of the electric vehicle in the fleet of the electric vehicles upon tracking of the vehicle data; and
maintain the vehicle data linked with the unique vehicle profile generated for each of the electric vehicle in the fleet of electric vehicles in a vehicle data repository;
an electric vehicle objective computation module operatively coupled to the fleet vehicle profile management module, wherein the electric vehicle objective computation module is configured to:
utilize one or more predefined fleet-level objectives corresponding to each of the electric vehicle in the fleet of the electric vehicles upon accessing the unique vehicle profile generated, wherein the one or more predefined fleet-level objectives are maintained in the vehicle data repository, wherein the one or more predefined fleet-level objectives is associated with one or more goals for the fleet of electric vehicles to determine performance of each of the electric vehicle in the fleet of the electric vehicles, and wherein the one or more predefined fleet-level objectives comprises at least one of: maximizing utilization, achieving target equipment life, maximizing equipment reliability, maximizing battery life, maximizing capacity over effective operating life of the electric vehicle, minimizing probability of equipment failures, and minimizing environmental impact; and
rank the one or more predefined fleet-level objectives in a predefined order for operation of each of the electric vehicle in the fleet of the electric vehicles by interfacing with a fleet manager;
a fleet score computation module operatively coupled to the electric vehicle objective computation module, wherein the fleet score computation module is configured to:
compute a vehicle level score corresponding to each of the one or more predefined fleet-level objectives ranked in the predefined order for each of the electric vehicle in the fleet of the electric vehicles, in real time using a machine learning based predefined score model, wherein the machine learning based predefined score model is configured to obtain at least one of: a vector and a graph to compute the vehicle level score for each of the vehicle in the fleet of electric vehicles;
compute a fleet-level score corresponding to each of the one or more predefined fleet-level objectives ranked in the predefined order for the fleet of the electric vehicles by combining the computed vehicle level score;
compute a fleet-level aggregate score for the fleet of the electric vehicles based on combination of each of the computed fleet-level score using a score combination technique; and
compute a change in the fleet level scores and the fleet-level aggregate scores respectively of the fleet of the electric vehicles for a predefined time interval,
wherein for each of the one or more predefined fleet-level objectives, the fleet score computation module is configured to:
determine a trendline for historical objective-specific scores for each of the one or more predefined fleet-level objectives up to a current time:
implement a low-pass filter to reject noise in the trendline; and
determine a derivative of the trendline at the current time,
wherein the derivative of the trendline is the total change over the predefined time interval; and
a generic fleet action implementation module operatively coupled to the fleet-level score computation module, wherein the generic fleet action implementation module is configured to utilize an implemented fleet action model based on the fleet-level aggregate score for predicting maintenance and performance of the fleet of the electric vehicles through a plurality of vehicle specific actions.
|