Master's thesis: on Distributed and Federated AI for Autonomous Vehicles

Master's thesis: on Distributed and Federated AI for Autonomous Vehicles Background Autonomous vehicles generate large volumes of multimodal sensor data, making centralized processing impractical due to communication, storage, and privacy constraints. The DREAM project (Distributed, Robust and Efficient AI for Autonomous Vehicles) develops efficient real-time federated learning methods for autonomous vehicles, with particular focus on self-supervised learning and knowledge transfer between heterogeneous AI models. Description We offer two Master's thesis topics within DREAM. The first investigates federated self-supervised learning (SSL) to exploit largely unlabeled driving data and reduce dependence on costly annotations. The second investigates knowledge transfer/knowledge distillation between models with different architectures, sensors or hardware, enabling learning to continue as vehicle platforms evolve. The final scope will be defined together with the supervisors and aligned with DREAM's research objectives. Key Responsibilities

• Topic 1 – Federated Self-Supervised Learning: develop and evaluate novel SSL methods for driving data; investigate hybrid supervised/self-supervised federated learning; compare with relevant supervised and federated baselines.
• Topic 2 – Federated Knowledge Transfer: investigate knowledge distillation/transfer across heterogeneous models; study adaptation to changes in model architecture, sensors or hardware; evaluate knowledge transfer within federated learning.
• Implement and experimentally evaluate the proposed methods using autonomous-driving data, including the Zenseact Open Dataset (ZOD), and present the results to DREAM project partners.

Qualifications We are looking for two highly motivated Master’s students with a strong background in machine learning and computer vision. Essential skills include deep learning, Python programming, the ability to read and understand scientific literature, and experience working with complex systems. Experience with federated learning, self-supervised learning, knowledge distillation, or autonomous-driving data is considered a merit. The thesis work will be carried out in close collaboration with the research team at RISE and Zenseact. You are expected to work on-site at the RISE Kista office at least three days per week. Terms

• Recruiting manager: Karin Kraft, PhD
• Industry supervisor: Sima Sinaei, PhD; Henrik Abrahamsson, PhD; Mina Alibeigi, PhD
• Location: RISE, Kista, Stockholm
• Applications are reviewed on a rolling basis, apply as soon as possible, but no later than Dec 15th, 2026
• Starting date: January 11th 2027
• Credits: 30 points
• Compensation: 39 000 SEK upon a successful completion of a high-quality thesis.

Welcome with your application! Send in your application (CV, motivation letter, transcript of records) no later than December 15th. For any questions, please contact:

• Henrik Abrahamsson, henrik.abrahamsson@ri.se
• Sima Sinaei, sima.sinaei@ri.se