Better understanding of commercial vehicle energy consumption

Challenge

Empirically predict the energy consumption of vehicles based on environmental conditions and driving characteristics such as slope, speed and acceleration, in order to better control the consumption.

Resolution

Application of advanced oversampling techniques to overcome the problem of imbalance in different kinds of available data. Application of various classification, regression and clustering algorithms on the oversampled data in order to precisely determine driving behaviors influencing the average consumption per kilometer.

Techno

Python/scikit-learn/SMOTE

After this project, I adapted my proper driving techniques based upon results we discovered, and I got to significantly lower my fuel consumption!

Better understanding of commercial vehicle energy consumption

Challenge

Resolution

Techno

Data Scientist

Louvain-la-Neuve or Brussels

Project Manager

Louvain-la-Neuve or Brussels