System and method for estimating travel time and distance

A method for estimating travel time and distance, comprising:

• obtaining a vehicle trip dataset comprising an origin, a destination, a time-of-day, a trip time, and a trip distance associated with each of a plurality of trips, wherein the origin and the destination are represented as origin geo-coordinates and destination geo-coordinates, respectively;
• preprocessing the vehicle trip dataset, wherein the preprocessing comprises:
• discretizing the origin geo-coordinates and the destination geo-coordinates into respective spatial cells;
• converting the origin geo-coordinates and the destination geo-coordinates into geo-coordinates of the respective spatial cells;
• and
• training a neural network model with the preprocessed vehicle trip dataset to obtain a trained model, wherein:
• the neural network model comprises a first module and a second module;
• the first module comprises a first number of neuron layers;
• the first module is configured to receive the origin and the destination as first inputs and generate an estimated travel distance;
• the second module comprises a second number of neuron layers;
• and
• the second module is configured to:
• obtain information of a last layer of the first module, wherein the information comprise activations of the last layer of the first module and the estimated travel distance;
• and
• receive the obtained information of the last layer of the first module and the time-of-day as second inputs and generate an estimated travel time.

A non-transitory computer-readable medium for estimating travel time and distance, comprising instructions stored therein, wherein the instructions, when executed by one or more processors, cause the one or more processors to perform the steps of:

• obtaining a vehicle trip dataset comprising an origin, a destination, a time-of-day, a trip time, and a trip distance associated with each of a plurality of trips, wherein the origin and the destination are represented as origin geo-coordinates and destination geo-coordinates, respectively;
• preprocessing the vehicle trip dataset, wherein the preprocessing comprises:
• discretizing the origin geo-coordinates and the destination geo-coordinates into respective spatial cells;
• converting the origin geo-coordinates and the destination geo-coordinates into geo-coordinates of the respective spatial cells;
• and
• training a neural network model with the preprocessed vehicle trip dataset to obtain a trained model, wherein:
• the neural network model comprises a first module and a second module;
• the first module comprises a first number of neuron layers;
• the first module is configured to receive the origin and the destination as first inputs and generate an estimated travel distance;
• the second module comprises a second number of neuron layers;
• and
• the second module is configured to:
• obtain information of a last layer of the first module, wherein the information comprise activations of the last layer of the first module and the estimated travel distance;
• and
• receive the obtained information of the last layer of the first module and the time-of-day as second inputs and generate an estimated travel time.