Help a humanoid robot turn a screwdriver. In this master's thesis at 120ft, you will compare approaches to robotic screw driving, then design, build and validate a working prototype.
Host: 120ft Factory AB, Stockholm
Company supervisor: Hsiu-Fang Chien
Field: Mechanical and electrical design, mechatronics
120ft builds Factory OS for humanoid robots, currently the Unitree G1-D, doing high-mix, low-volume assembly. The current Dex1-1 parallel-jaw gripper cannot reliably align and drive a screw, and there is no dedicated solution for it today.
Several directions are plausible: a handheld screwdriver the gripper picks up, a dedicated end-effector with swappable bit heads, or a screwdriver mounted on the arm or wrist. Each brings different trade-offs in mechanical complexity and how much it limits the gripper's other uses.
What you will buildCompare solutions. Evaluate at least these three approaches against criteria including fastener and thread size range, minimum torque and axial force, wrist or arm adaptor needs, and position precision and accuracy. Choose a direction with reasoning the supervisors can follow.
Design and build. Develop a working prototype of the chosen solution with its own power source and a microcontroller-based control interface.
Validate under teleoperation. Drive the robot through the screwing sequence and demonstrate repeated, reliable use of the tool.
A technical supervisor.
Access to a real Unitree G1-D.
Workshop and prototyping support as available.
You are a master's student in mechatronics, machine design or robotics, comfortable with both mechanical and electrical design.
ApplySend your CV and a GitHub, LinkedIn or website link. Tell us why this thesis interests you and share an example of work you are proud of.
Skills Required
- Master's student in mechatronics, machine design, or robotics
- Comfortable with mechanical and electrical design
- Submit a CV and a GitHub, LinkedIn, or website link
What We Do
120ft Factory is a Swedish manufacturing startup developing autonomous local micro-factories powered by Physical AI. Its Factory-as-a-Service approach uses containerised production cells and humanoid robots for flexible, high-mix, low-volume assembly. The company is building Factory OS and robotic systems to help manufacturing SMEs automate complex assembly, bring production closer to customers, and reduce dependence on rigid global supply chains through locally operated production facilities.
