Tesla is paying up to $48 per hour for people to wear motion-capture suits to help train its humanoid Optimus robot

Tesla is hiring full-time workers for the role and is willing to pay between $25.25 to $48 per hour, according to the job listing. The role is available with multiple shifts, including from 8 a.m. to 4:30 p.m., 4 p.m. to 12:30 a.m., and 12:00 a.m. to 8:30 a.m.

Animesh Garg, a senior researcher at Nvidia Research and a robotics professor at Georgie Institute of Technology, said the job likely entails collecting data through simply performing tasks in the motion-capture suit that can be recorded and later analyzed by data annotators in order to train the AI system, as well as tele-operating the humanoid robot directly.

Tesla’s Optimus social media account provided a peek into what the job entails last year.

The video clip shows an operator in a motion-capture suit, with haptic gloves and a VR headset picking up objects while a virtual version of the Tesla Optimus bot performs the same movements.

Earlier this year, Tesla appeared to go on a hiring spree for the role. More than 50 different people have worked in the role over different periods over the past year, according to a Business Insider review of LinkedIn data.

Tesla is likely looking to gather vast amounts of data, experts say.

“It’s the exact same thing they’re doing with Autopilot,” Jonathan Aitken, a robotics expert out of the University of Sheffield, said. “But rather than collecting data from their fleet of cars, they’re paying people to come into the factory or come into a workplace and undertake these tasks.”

Garg, the Nvidia researcher and robotics professor, estimates that Tesla should be aiming to log anywhere from hundreds of thousands of hours to millions of hours of data in order to train Optimus how to work in Tesla’s factories — a feat that could take a lot of time and money.

“It extremely difficult to produce robots at scale,” Garg said, adding that Tesla’s Optimus robots are likely custom built and assembled, as well as constantly undergoing new iterations.

“The amount of data collection you’d need would easily be half a billion dollars and the real question is ‘Even if you do that, do you succeed?’ Because there is no guarantee of success,” he added.

The use of motion-capture suits and tele-operation in robot training is far from a new practice, but Tesla is among the first to attempt to use the practice at this scale and to perform such a wide variety of generalized tasks, according to Aitken.

“In the manufacturing world, we typically have a robot that can do one task really really well, but it doesn’t usually extend beyond that,” Aitken said.

A spokesperson for Tesla did not respond to a request for comment.

Do you work for Tesla or have a tip? Reach out to the reporter via a non-work email and device at gkay@businessinsider.com or 248-894-6012