Event detection based on vehicle data

A method comprising:

• receiving data associated with a vehicle;
• determining a feature represented by at least a portion of the data;
• determining, based at least in part on the feature and one or more criteria, a potentially unsafe operation of the vehicle;
• receiving ground truth data representing a classification indicating whether the portion of data is associated with an unsafe operation of the vehicle;
• training, based at least in part on the portion of date and the ground truth data, a machine learned model to generate a trained machined learned model;
• and
• transmitting the trained machine learned model to the vehicle, the vehicle configured to input additional data associated with the vehicle into the trained machine learned model and to perform a maneuver based at least in part on an output of the trained machine learned model.

One or more non-transitory computer-readable media storing instructions that, when executed by one or more processors, cause one or more computing devices to perform operations comprising:

• receiving data associated with a vehicle;
• determining a feature represented by at least a portion of the data;
• determining, based at least in part on the feature and one or more criteria, a potentially unsafe operating condition of the vehicle;
• inputting the portion of the input data into a machine learned model;
• receiving, from the machine learned model, a classification for the portion of the data, the classification indicating an unsafe operating condition of the vehicle;
• and
• transmitting the machine learned model to the vehicle, the vehicle configured to input additional data associated with the vehicle into the machine learned model and to perform a maneuver based at least in part on an output of the machine learned model.