04 Research Internship / Master's Thesis - Robot Learning Winterthur On site Internship / Thesis

Binabik; Winterthur; Vor Ort; CHF 20.000 - 27.000; Vollzeit

Benefits dieser Stelle

Paid internship

Mentoring

Hardware access

Thesis collaboration possible

Office at Technopark Winterthur

Zusammenfassung

Binabik AI is seeking a Master’s or PhD student to join our robot learning team and improve how efficiently our robots learn from demonstrations. You’ll work on learning-based manipulation for machine tending and logistics, including imitation learning, RL in simulation and sim-to-real transfer, and you’ll evaluate results on our AgiBot G2 and Galaxea R1 Pro.

You’ll define your project with your mentor in the first week, based on your interests and our roadmap.

Qualifikationen

• Enrollment in a Master’s or PhD program in robotics, CS, ML, mechanical or electrical engineering, or a related field.
• Strong Python and PyTorch skills.
• Project or thesis experience in imitation learning, RL, generative models, or robot manipulation.
• Ability to read recent papers and implement methods.
• Patience debugging on real hardware.

Aufgaben

• Own a research project from literature review to new method and real-robot experiments.
• Train and evaluate on GPU compute, in simulation and on hardware.
• Collect data with demonstration and teleoperation setups.
• Share learnings with the team and potentially publish or open-source release.

Kenntnisse

Python

PyTorch

Imitation learning

Reinforcement learning

Robot manipulation

Ausbildung

Master's or PhD student in robotics/CS/ML/mechanical or electrical engineering

Tools

MuJoCo

LeRobot

ROS 2

Jobbeschreibung

Help our humanoids learn new tasks from a few demonstrations, and test your results on real machines.

About Binabik AI

Humanoid robots that run your machines.

Binabik AI builds the software that lets humanoid robots do productive work on the factory floor. Our wheeled humanoids, AgiBot G2 and Galaxea R1 Pro, load and unload CNC lathes and mills, fetch material, carry parts and stack pallets. We focus on machine tending and intralogistics, the repetitive tasks that keep production running and are hard to staff.

Our approach combines proven methods from classical robotics, such as motion planning, perception and control, with recent advances in physical and agentic AI. The classical parts make the system reliable and the AI parts make it adaptable. Teaching a new task takes only a few recorded demonstrations. Binabik Teach captures them, and our Skill Engine breaks the task into reusable skills. Binabik Sim, our custom simulator, then replicates each skill and refines it across variations in parts, positions and environment conditions before it goes to the real machine. A hardware abstraction layer lets skills carry over between robot platforms, and data collected in operation goes back into improving them. Customer data stays on-premise.

We are a ZHAW spin-off with roots at the University of Zurich, the ETH AI Center and MIT. Our founders are experienced robotics and AI researchers who have published over 200 research papers. We are a pre-seed company finishing our first proofs of concept and moving towards industrial pilots. We plan to grow quickly over the next year, so the people who join now will help shape the product, the culture and the company.

The role

We're looking for a Master's or PhD student to join our robot learning team and improve how efficiently our robots learn from demonstrations. You'll work on learning-based manipulation for machine tending and logistics, including imitation learning, VLA fine-tuning, reinforcement learning in simulation and sim-to-real transfer, and you'll evaluate your results on our AgiBot G2 and Galaxea R1 Pro.

You'll define your project with your mentor in the first week, based on your interests and our roadmap. The project can also be carried out as a Master's thesis together with your university. Example topics:

• Few-shot imitation learning for picking and placing parts with tight tolerances

• Fine-tuning an open VLA (e.g. π0, SmolVLA, GR00T) on machine-tending tasks

• RL fine-tuning in simulation with domain randomization, and transfer to the real robot

• Automatic segmentation of demonstrations into reusable skills

• Learning from success and failure signals reported by the production machine

What you'll do

• Own a research project, from literature review and baselines to a new method and real-robot experiments.

• Train and evaluate on our GPU compute, in simulation and on hardware, using our benchmarks.

• Collect the data you need with our demonstration and teleoperation setups.

• Share what you learn with the team and, where it fits, in a publication or open-source release.

What you bring

• Current enrollment in a Master's or PhD program in robotics, CS, ML, mechanical or electrical engineering, or a related field

• Strong Python and PyTorch

• Project or thesis experience in imitation learning, RL, generative models (diffusion or flow matching) or robot manipulation

• The ability to read recent papers and implement their methods

• Patience and care when debugging on real hardware

Nice to have

• Experience with real robot arms or grippers

• MuJoCo, LeRobot, ROS 2

• Publications or open-source work in robot learning

Interning at Binabik

You’re part of the team from day one, with a dedicated mentor, a desk in the lab and the same tools as everyone else.

• Duration: 6 to 12 months

• Start: flexible

• Your project: We scope it with you in your first week, with a clear deliverable. At the end, you present your work to the team or customers.

• Room to grow: We adapt the scope to your skills and drive. If you’re ready for more, we’re happy to give you full ownership of a component or feature.

• Mentoring: weekly 1:1s with your mentor, and the rest of the team is around every day.

• Impact: Projects are chosen so the results can be used. Depending on the project, your work may run on our robots, go into customer deployments or become part of a publication.

What we offer

• Paid internship: 6 to 12 months

• Mentoring: a dedicated mentor and a project with a clear goal and practical relevance

• Hardware access: You’ll work hands-on with our humanoid robots, lab equipment and GPU compute.

• Next steps: possible continuation as a full-time employee or through a thesis collaboration

• Location: Our office at Technopark Winterthur, with the robot lab on site, is a short train ride from Zürich.

Hiring process

1. Introductory call: a short conversation about your background, your goals and what we’re building.

2. Technical interview: We go into your experience in depth and work through a problem drawn from our day‑to‑day work. For commercial roles, this step is a case discussion.

3. A day with us: You’ll spend a day at our office in Winterthur, see the robots in operation and meet the team.