Master Thesis - Artificial Intelligence (m/w/d) - End-to-End Learning for Hybrid Machine Learning Mo

Do you want to put all your theoretical knowledge into practice, gain operational experience and broaden your skills? We offer a option for you to work with our employees, experts and specialists in a fascinating high-tech environment. We are offering a unique opportunity to prove and recommend yourself for a possible future employment.

 

The AI/ML Team in TGRE is a highly focused and one of the few teams in the world, which focus entirely on the application of AI and ML in the race environment. Our tools are state-of-art and we have close cooperation’s with the leading institutes in data science. We are passionate and diverse ML engineers who drive to contribute to Toyota victories on the race track.  

 

 

Your thesis:

 

Some vehicle signals cannot be measured during the race and therefore need to be estimated based on measurements. Among different methods, the combination of Kalman Filter (KF) and Physics Informed Neural Networks (PINN) shows promising estimation results. The steps of the learning process consisted of training the PINN and tuning the KF after combining them into a hybrid model. Focus of this study is to achieve the best possible performance of the hybrid model by applying end-to-end learning approaches.

 

 

The successful candidate will have:

 

  • Bachelor’s degree in engineering, physics, or related field
  • Courses and student projects with ML in Python
  • Courses in control theory incl. Kalman Filters
  • Ability to communicate with engineering and technicians from various disciplines
  • Well-structured and systematic research approach
  • Very good command of the English language (in speech as well as in writing)
  • Passion for motorsport