A non-transitory computer-readable medium storing instructions, the instructions comprising:
- one or more instructions that, when executed by one or more processors, cause the one or more processors to:
- receive data identifying network constraints associated with networks connected to vehicle devices of vehicles, and data identifying data collection constraints;
- receive, from the vehicle devices, data identifying vehicle device constraints associated with the vehicle devices;
- determine first sampling rates, first time periods, second sampling rates, and second time periods for collecting vehicle data from the vehicle devices based on the data identifying the network constraints, the data identifying the data collection constraints, and the data identifying the vehicle device constraints,
- wherein the first sampling rates are different than the second sampling rates;
- receive, from the vehicle devices, first vehicle data, provided at the first sampling rates and for the first time periods, and second vehicle data, provided at the second sampling rates and for the second time periods;
- train a machine learning model with the first vehicle data and the second vehicle data to generate a trained machine learning model;
- receive data identifying particular network constraints associated with a particular network connected to a particular vehicle device of a particular vehicle;
- receive, from the particular vehicle device, data identifying particular vehicle device constraints associated with the particular vehicle device;
- determine a third sampling rate, a third time period, a fourth sampling rate, and a fourth time period for collecting vehicle data from the particular vehicle device based on the data identifying the particular network constraints, the data identifying the data collection constraints, and the data identifying the particular vehicle device constraints,
- wherein the third sampling rate is different than the fourth sampling rate;
- receive, from the particular vehicle device, third vehicle data, provided at the third sampling rate and for the third time period, and fourth vehicle data, provided at the fourth sampling rate and for the fourth time period;
- process the third vehicle data and the fourth vehicle data, with the trained machine learning model, to identify a driving behavior or an issue associated with the particular vehicle;
- and
- perform one or more actions based on the driving behavior or the issue associated with the particular vehicle,
- wherein the one or more actions include one or more of:
- providing, to the particular vehicle device, an alert identifying the driving behavior or the issue associated with the particular vehicle,
- providing, to the particular vehicle device, an instruction to address the driving behavior, or
- providing an indication of the driving behavior to a device associated with an insurer of the particular vehicle.