Green vehicle assignment using Geospatial Analysis and Machine Learning algorithms

The project proposes a new methodology for quantifying by type of road the impact of the last mile delivery operation in terms of Co2 emissions and allows the assignment of the vehicle that outperforms in each cluster.

This methodology considers the impact of road conditions, topography and load on CO2 emissions. Our methodology clusters the delivery areas using six parameters: (1) Gradient, (2) Mean velocity, (3) Mean elevation, (4) Average segment length, (5) Percent of the route that is flat, and (6) Percent of the route that is steep.

The output is then used to define the optimal CO2 truck-area assignment.