Hybrid Thesis - Driving Stability through Reinforcement Learning (Gifhorn, DE, 38518) bei IAV

This challenge awaits you:

Are you a student majoring in automotive engineering, physics, computer science, or a related field, and are you interested in reinforcement learning (RL) or similar methods? We offer a fantastic opportunity for a talented and motivated student to join our team and help us apply and further develop modern machine learning methods in various areas such as vehicle dynamics, autonomous driving, virtual testing, and many more.

Your Tasks:

• You will develop a framework that integrates LLMs (Large Language Models) with reinforcement learning to automate the design of reward functions.
• You will integrate a suitable LLM using the Model Context Protocol.
• You adapt the existing RL algorithm to include any additional interfaces that are required.
• You demonstrate the effectiveness of LLM-generated reward functions in accelerating the progress of RL training.

Necessary Skills:

• Current master's degree with a focus on automotive engineering, electrical engineering, computer science, mechatronics, or similar courses of study
• Familiarity with RL frameworks (e.g., Stable Baselines3) and LLMs
• Very good programming skills in Python
• German language skills at least at B2 level

Desired Skills:

• Experience with simulation tools (CarMaker, etc.) and vehicle dynamics modeling is an advantage
• Analytical thinking and problem-solving skills
• Good written and oral communication skills
• Ability to work independently and as part of a team