Development of a Benchmarking Environment for Active Learning Algorithms
We are looking for a motivated and technically skilled student research assistant (HiWi) to support the ALeSCo project, which focuses on active learning of dynamical systems.
Task Description
To adequately evaluate data-driven algorithms, simulation-based assessments are essential. Static problems such as image recognition and natural language processing are typically evaluated using predefined, comprehensive datasets. Dynamic tasks, as commonly encountered in robotics, require correspondingly dynamic environments. While environments based on Gymnasium provide a solid foundation for this, their quality criteria and tasks are designed to assess an algorithm's performance at the end of training. Many practical control tasks — such as robust stabilization and compliance with safety requirements, including during interaction with the system — are not captured by this approach.
You will support the development of a benchmark for active learning systems. The benchmark is intended to enable the automated and reproducible evaluation of algorithms in dynamic systems. To this end, you will develop a benchmarking environment with defined interfaces for simulation environments, agents, and evaluation modules. Specifically, you will design and implement interfaces for simulation environments such as MuJoCo, as well as evaluation modules based on suboptimality criteria — known as "regret" — and document their usage through examples.
Your Profile
- Enrolled in engineering, computer science, mathematics, physics, or a comparable program at TUM.
- Enthusiasm, creativity, and the ability to work independently and responsibly.
- Solid knowledge of control theory and machine learning, ideally also statistics and robotics.
- Experience with higher-level programming languages such as Python. Knowledge of MATLAB or Julia is a plus.
We Offer
- Salary according to Tarifvertrag TV-L “HiWi”.
- A central location in the heart of Munich at the Campus Innenstadt.
- Collaboration with a dynamic and innovative team.
- Opportunity for direct involvement in the latest developments in research and technology.
Application
Please send your application including a brief motivation letter, CV, and transcript of records by e-mail to
- M.Sc. Max Beier (max.beier@tum.de)
- M.Sc. Sami Leon Noel Aziz Hannah (ge69yot@mytum.de)
- M.Sc. Konrad Reichel (konrad.reichel@tum.de)
TUM strongly encourages applications from women and underrepresented groups, in line with its commitment to diversity in engineering education, research, and practice.
ecoBay Project
We are looking for a motivated and technically skilled student research assistant (HiWi) to support the ecoBay project.
About the Project
ecoBay is a newly launched research project (January 2026 – February 2031), funded by the Bavarian State Ministry of Education, Science and the Arts as part of the BayKlif2 climate research network. The project combines AI-driven autonomous robotics with genetic biodiversity monitoring to detect early signs of ecological change in Bavarian lakes. An autonomous surface vehicle (USV) collects environmental DNA/RNA (eDNA/eRNA) samples and continuous physico-chemical measurements across lakes such as Ammersee and Großer Ostersee. AI-based adaptive mission planning ensures that the most informative locations are sampled at the right time. The project is carried out in close collaboration with TUM Global Change Limnology, LMU Palaeontology & Geobiology, and MIRMI.
Your Tasks
- Support the development and testing of software modules for autonomous mission planning and control of the uncrewed surface vehicle (USV)
- Assist with the integration and evaluation of sensor data pipelines (e.g., water quality sensors, GPS, cameras)
- Contribute to simulation and real-world experiments at outdoor test sites (lakes in the greater Munich area)
- Help with data collection, processing, and documentation
Your Profile
- Currently enrolled in a Bachelor's or Master's programme in Engineering, Computer Science, Robotics, Mathematics, or a related field at TUM
- Enthusiasm for robotics, autonomous systems, and/or environmental sensing
- Experience with programming languages such as Python, C++, or MATLAB
- Knowledge of ROS (Robot Operating System) is a plus
- Prior experience with autonomous vehicles, embedded systems, or field robotics is welcome but not required
- Independent, reliable, and motivated work style
We Offer
- Remuneration according to TUM student assistant rates
- Active participation in a cutting-edge interdisciplinary research project at the interface of robotics, AI, and environmental science
- Collaboration with a dynamic and international team at MIRMI and ITR
- Hands-on fieldwork at lakes in Bavaria
- Flexible working hours compatible with your study schedule
- Possibility to develop your involvement into a Bachelor's or Master's thesis
Application
Please send your application – including a brief motivation letter, your CV, a current transcript of records, your planned start date, and your expected availability (hours per week) – to:
Dr.-Ing. Stefan Sosnowski sosnowski@tum.de
Please send all documents collected in a single PDF file.