Are you passionate about robotics and eager to contribute to the advancement of humanoid robots in real-world household tasks? At H2T, we develop assistive humanoid robots (the ARMAR series) envisioned to operate in human-centered environments, providing help with domestic chores.

We welcome KIT undergraduate and graduate students to join us part-time in developing novel skills for household scenarios. This is a unique opportunity to work with state-of-the-art humanoid robots, algorithms and software, collaborate with researchers in the lab, and translate our cutting-edge research results into real-world applications evaluated on multiple humanoid robots.
Below we provide some reference open topics.

Open topics

Teach a household robot new tasks: Learning from demonstration is crucial for teaching robots new manipulation skills. We have developed fundamental building blocks and controllers for teaching and executing demonstrated motions in a kitchen scenario. Your tasks focus on extending and improving the robot's skills, such as loading and unloading the dishwasher, fetching objects from drawers and fridges, and setting or tidying up a dining table.

Human-robot interaction and collaboration: Humanoid robots must behave naturally when interacting with humans. Understanding human intentions is essential for collaborative tasks. This topic involves developing interactive and collaborative skills, such as adaptive handover, and collaborative bimanual transport tasks, while incorporating human-like behaviors.

Object grasping: Robotic grasping involves perceiving object status, generating and validating grasps, and executing them. Tasks in this area include integrating state-of-the-art computer vision algorithms for better online estimation of object status and improving motion execution in cluttered scenes.
 

Interactive learning and correction: Assistive robots in human-centered environments need to be equipped with basic functionalities, such as learning new objects or concepts from humans and adjusting compliance via verbal interaction. This task focuses on developing tools to interactively demonstrate the robot's functionalities.

More topics will be discussed when you are on board!

Qualifications

We are looking for candidates with:

- good knowledge of the programming languages C++ and Python
- desire to work as part of a team and be self-reliant on the tasks
- commitment to follow through with safety guidelines and responsibilities
- openness to learning new skills

Note that prior knowledge in imitation learning, robot control, computer vision, and deep learning is advantageous.

Responsibilities

In addition to the responsibilities of each particular topic, we expect the student assistant to:

- attend the regular software development meetings
- demonstrate the robot skills and our research outcomes to visitors and guests
- support the maintenance of the robot platforms

How to apply?

Please send an e-mail to Fabian Reister and Jianfeng Gao including your Curriculum Vitae (CV), transcript of records, and other relevant documents. We are looking forward to receiving your applications.

Join us at H2T and be a part of the future of humanoid robotics!