Master Thesis: Generative AI for Future Emission Management in Heavy-Duty Vehicles

Background 

Future heavy-duty vehicles must meet increasingly stringent emission legislation while minimizing fuel and AdBlue consumption. Achieving this requires intelligent coordination between the engine and the exhaust aftertreatment system (EATS), balancing multiple objectives such as tailpipe NOx emissions, exhaust temperatures, fuel efficiency, and ammonia storage. 

At the same time, recent advances in Generative AI have created new opportunities to learn complex system behavior from historical data and generate future operating strategies. While traditional control methods typically optimize a single predicted future, generative models can propose multiple alternative future scenarios that may enable improved decision-making over longer time horizons. 

This thesis aims to investigate whether Generative AI can be used as a supervisory planning layer for future emission management of heavy-duty powertrains. 

Where You’ll Belong 

Volvo Group Powertrain Technology provides Volvo Group Trucks and Business Area’s with state-of-the-art research, cutting-edge engineering, product planning and purchasing serives, as well as aftermarket product support. The Powertrain Engineering Sweden organization has the full worldwide product platform responsibility for heavy duty engines and transmissions with a large organization of 750 colleagues in Gothenburg and Malmö. 

Why Choose This Thesis? 

This thesis offers a unique opportunity to work at the intersection of: 

• Artificial Intelligence and Machine Learning 

• Advanced Control 

• Combustion and Emission Control 

• Exhaust Aftertreatment Systems (EATS) 

• Sustainable Transportation 

You will work with real industrial data and contribute to Volvo Group's exploration of future technologies for robust emission compliance and energy-efficient vehicle operation. The project combines state-of-the-art AI methods with real engineering challenges and offers significant opportunities for both research and industrial impact. 

Scope and Content 

The objective is to investigate whether Generative AI can generate future emission-management strategies that improve the trade-off between Fuel consumption, AdBlue consumption and Robust tailpipe NOx compliance. 

The work will include: 

• Literature Review of: 

• • Generative AI 

• • Time-series and sequence modeling 

• • Transformer architectures 

• • AI for control systems 

• • Model Predictive Control (MPC) 

• Analyze real engine and aftertreatment data, including:

• • Engine speed and torque 
  • Engine-out NOx 

• • Exhaust temperatures 

• • Urea dosing 

• • NH₃ storage estimates 

• • Preview information describing future operating conditions 

• Model Development: Develop and train a generative model capable of learning the interaction between engine operation, emissions, thermal behavior, and aftertreatment performance. 

• Strategy Generation and Evaluation: Investigate whether the model can generate alternative future operating strategies and evaluate their potential impact. 

• Demonstrator: Develop a proof-of-concept demonstrator showcasing the potential of Generative AI for supervisory emission management. 

Who are you? 
You are curious, analytical, and motivated by solving complex interdisciplinary problems. You enjoy learning new technologies and working collaboratively in a team environment.  

We are looking for two students who will complete this thesis together, we see that the combination of Control Engineering and Machine Learning / Artificial Intelligence would be prefarble. 

Experience in several of the following areas is beneficial: 

• Automatic Control 

• Model Predictive Control (MPC) 

• Machine Learning and Deep Learning 

• Python 

• Matlab / Simulink 

• Data Analytics 

• Signal Processing 

• Combustion engines, emissions, and aftertreatment 

Ready for the next step? 

Want to kick-start your career? Please apply now, last application date is 14th of October. We value your data privacy and therefore do not accept applications via mail. 

Thesis level: Master Thesis 30p  

Language: English 

Starting date: January 2027 

For any further details please contact: Johan Dahl, Expert Systems Engineer at johan.dahl@volvo.com 

Supervision and examination: 
- Powertrain Technology, TTI, Volvo Group 
- Chalmers