Thesis - Edge-Deployable Virtual Sensing for Vehicle Dynamics

We are Geely Technology Europe. A unified European R&D centre within Geely Auto Group, where world-class engineers, developers and innovators push the boundaries of intelligent mobility. Our mission is to shape the next generation of vehicle architectures, digital technologies and intelligent systems for global markets. We integrate European customer and regulatory requirements early in the development process and support multiple brands within the Geely portfolio, including Zeekr, Lynk & Co and Geely. With nearly two decades of engineering experience in Europe, we continue to build smart, sustainable and user-centric mobility solutions.

Edge-Deployable Physics Informed Neural Network-based Virtual Sensors for Vehicle Dynamics

Project Description:

Accurate knowledge of vehicle dynamic states is essential for advanced motion control, driver assistance, energy-efficient driving, and automated driving functions. However, many relevant states are difficult, costly, or impractical to measure directly in production vehicles. Virtual sensing offers a way to estimate such quantities using available vehicle signals, models, and data-driven methods. Traditional physics-based approaches provide interpretability and engineering insight, while machine learning can capture complex patterns from data.

This thesis builds upon a previous thesis conducted within the team, Physics-Informed Neural Networks for Vehicle Lateral Dynamics Modeling . Students are encouraged to review the findings and methodology presented in that work as a starting point for understanding the problem domain and existing approaches. The present thesis aims to advance the state of the art by investigating scientific machine learning methods, including physics-informed and physics-augmented machine learning approaches, to improve robustness, generalization, interpretability, and suitability for deployment in real-world automotive systems.

Scope

The scope of the thesis is to investigate machine learning-based virtual sensing methods for estimating selected vehicle dynamic states from available sensors and vehicle network signals. The work should remain open with respect to the final model architecture and focus on how domain knowledge, physical principles, and data-driven learning can be combined in a scientifically grounded way. The study may cover lateral and/or longitudinal vehicle dynamics, depending on data availability and project focus, and should consider robustness across operating conditions, generalization to unseen scenarios, and feasibility for real-time automotive applications.

Objectives

• Develop and evaluate machine learning-based methods for estimating selected vehicle dynamic states from available vehicle signals.
• Investigate physics-informed or physics-augmented machine learning approaches as part of the broader machine learning discipline, combining physical insight with data-driven modelling.
• Validate the proposed approach using representative vehicle data and compare it with relevant baseline methods.
• Assess how well the approach performs across different driving conditions, maneuvers, and operating regions.
• Analyze computational complexity and implementation considerations for potential use in real-time automotive systems.

Expected Outcomes

• A literature review of virtual sensing and scientific machine learning methods relevant to vehicle dynamics estimation.
• A proposed virtual sensing framework for estimating selected vehicle dynamic states using available vehicle signals.
• An evaluation of physics-informed or physics-augmented machine learning methods, positioned within scientific machine learning, for incorporating physical knowledge into data-driven models.
• A validation and benchmarking study comparing selected methods against relevant baseline models using representative vehicle data.
• A practical assessment of robustness, generalization, computational complexity, and feasibility for real-time automotive use.

Your skills and background

• Ongoing Master's degree with the focus on Deep Learning and Applied Mathematics
• Strong background in Deep Learning (CNN, RNN, LSTM, GAN, etc.)
• Experience with Machine Learning for Time-series or Dynamical systems
• Experience with Machine Learning frameworks (e.g., TensorFlow, PyTorch)
• Knowledge in Control systems, Vehicle dynamics or Applied physics
• Strong programming skills in Python, C/C++, or similar languages
• Familiarity with Embedded or Real-time computing systems
• Strong analytical and problem-solving skills

Why you should join Geely Tech Eu

We are engineers, developers, and innovators from around the world. Joined together by entrepreneurship, our unique blend of global culture, and a belief in a smarter more sustainable future. At Geely Tech Eu we fast-track innovation and transform ideas into pioneering technology solutions, doing your master thesis here is no different. We are convinced that a thesis project is a major contribution to our innovation capabilities and long-term development. You'll have a great opportunity to use your skills and creativity to push the boundaries of what's possible.

What happens when you apply
If this sounds interesting and you match the requirements, please don't hesitate to submit your application with a CV and cover letter. Shortlisted candidates will be contacted for an interview to further discuss the project's details and expectations.

This thesis project is intended for 2 students. Applicants may specify a preferred partner in their cover letter.

Supervisor: Karthik Prasad, Expert Motion Systems, karthik.prasad@zeekrtech.eu , please contact for more information about the project

Starting date: January 2027

Last application date: 2026-10-31

Please note that due to GDPR regulations, we can only accept applications sent through the recruitment system, not via email or other channels.