Internship | Machine Learning for Heavy-Duty CO2 Emissions

About this position

In this internship, you will develop a second-by-second CO2 emission model for heavy-duty vehicles using XGBoost, building on previous work for light-duty vehicles. The main challenge is that HD vehicle mass can vary strongly with payload and is often unknown, so part of the project is to estimate the current vehicle weight from available signals and use that information in a robust predictive CO2 emission model. The final goal is to create an accurate, data-driven model that can capture real-world driving behaviour and improve CO2 emission estimation for heavy-duty transport. What will be your role?

During the internship, the student will work on building and improving a CO2 emission model for heavy-duty vehicles using XGBoost. In practice, this means exploring real vehicle data, cleaning and preparing datasets, selecting relevant driving and vehicle features, and developing methods to estimate the vehicle’s current weight when it is ...