Thesis in the Field of Humanoid Robots & AI - Different Focus Topics

Stack

Responsibilities

• ▹Research and design a multi-agent orchestration architecture for humanoid robots in buildings; develop strategies for distributed localization and orientation (e.g. sensor fusion, RTLS integration); implement a prototype for coordinating multiple robots; evaluate localization accuracy, robustness and scalability
• ▹Set up a simulation environment for tool-fetching tasks (e.g. in Isaac Sim); implement reinforcement learning policies for grasping and transport manipulation in simulation; apply domain randomization and sim-to-real transfer strategies; validate the trained policies on the real humanoid robot and document the transfer
• ▹Collect and prepare demonstration data for tool-fetching tasks (e.g. via teleoperation or kinematic recording); implement and train an imitation learning model to reproduce human fetch-and-carry manipulations; evaluate generalization to different tools and placement locations; test on the humanoid robot and document the results
• ▹Design and implement a teleoperation control pipeline for a humanoid robot; develop methods for visualizing sensor and environment data for the operator; test low-latency control in human-centered environments (e.g. elevator buttons, doors, corridors); evaluate the suitability of teleoperated monitoring in terms of reaction time and situational awareness
• ▹Design a generic runtime environment (harness) for AI agents in care scenarios; unify the integration of tools and information sources (cameras, sensors, RTLS, HMI systems, robots); manage context and world state (states of people, objects and machines) for care activities; implement authorization and verification mechanisms for safe action execution; build a prototype using the example of moving hospital beds and document it

Requirements

• ▹Enrolled at a university or university of applied sciences in computer science, robotics, mechanical engineering, software engineering or a comparable field
• ▹Basic knowledge of Python (ideally with PyTorch) and/or C++
• ▹Familiarity with ROS2
• ▹You enjoy programming and like to learn new things
• ▹Good German and English skills

Nice to have

• ▹Depending on the focus topic, an advantage: experience with simulation and reinforcement learning (e.g. Isaac Sim), C++ (teleoperation), interest in AI agents and multi-agent systems (orchestration, harness) or in learning-based methods (imitation learning)

Soft skills

What we offer

• ▹Central location: only a few minutes' walk from KIT Campus South with excellent public transport connections
• ▹Flexible working hours: your studies come first and the right arrangement is found together
• ▹Remote work possible by individual arrangement with your supervisor
• ▹Practice-oriented projects in collaboration with industry partners and application of research results in humanoid robotics
• ▹Individual supervision with open, communicative exchange within the team
• ▹Exciting tasks, a high degree of personal responsibility and room for your own ideas

About the company

The Fraunhofer Institute of Optronics, System Technologies and Image Exploitation IOSB turns the latest scientific findings into technical innovations.