Start with a button

Getting a robot arm to press a button sounds simple. Where is the button? How does the arm reach it? Can it still get it right in a different car? Break the problem down, sketch it, explain it, then try it on a real machine.

We're building robots for automotive testing:

  • Inside the car: robot arms that operate touchscreens, buttons, knobs, steering wheels, pedals, and gear selectors.
  • Outside the car: mobile robots that follow trajectories as test targets, carry instruments for measurement and inspection, or work with an arm and tools.

During this internship, you'll work with an automation R&D mentor, starting with one part of the system.

What you'll work on

  • Try SDKs, connect devices, and test prototypes. Help select equipment and work out the interfaces with suppliers.
  • Calibrate systems and write motion scripts. Trace position and force errors so the arm hits the right spot with the right force.
  • Connect mobile robots, sensors, and test systems. Work out how to synchronize trajectories, triggers, and data.
  • Read logs, form hypotheses, and run experiments. Explore how tasks should stop and recover after a disconnect or timeout.
  • Use AI to help with development, and explore constrained device interfaces and task orchestration driven by AI. Check the results against real feedback. Confident answers deserve a test, too.

Pick a starting point from these directions, or bring an idea of your own.

What we hope you bring

  • You're an undergraduate or graduate student in automation, computer science, robotics, electronics, mechatronics, automotive engineering, or a related field.
  • You can program and have built a project. You can explain its interfaces, how it works, and what went wrong. Coursework, competitions, and personal projects all count.
  • You have some grounding in software, hardware, or control, and enough curiosity about machines to follow an unexpected result.
  • You can turn a problem into a testable hypothesis, use data to judge whether a change helped, and explain your findings to others.
  • You've used AI tools, can understand the code they generate, and keep investigating when experiments reveal a problem.

Useful things to have tried

Linux, Git, C++, ROS, or projects involving calibration, trajectory control, sensor synchronization, fixtures, vision, or agent tool use. Bring code, hardware, or experiment notes you'd like to show us.

What might come out of it

Work with your mentor to understand the principles and build your ideas. Try them on a machine and see where reality differs from your prediction. A new approach might start right there.