SYSTEMS AND METHODS FOR VEHICLE REVERSING DETECTION USING EDGE MACHINE LEARNING

Methods for reversing determination for a vehicle asset are provided. The methods include capturing by a telematics device coupled to the vehicle acceleration data from a three-axis accelerometer, determining by an edge reversing-determination machine learning mode, a machine-learning-determined reversing indication for the vehicle asset. The edge reversing-determination machine-learning model being updated based on a centralized reversing-determination machine-learning model trained using a vehicle-provided reversing indication.