Collecting robot training data is dirty, unglamorous work. Some AI labs are already paying XDOF to do it.

Two weeks ago, OpenAI said it would relaunch the robotics program it shuttered in 2021 — the latest signal that the biggest AI labs are racing to teach machines to operate in the physical world. But building capable robots requires something the AI industry doesn’t yet have, which is the training data to match that used for language models.

That gap is creating a new kind of infrastructure business. Unlike LLMs that were trained on a vast sea of publicly available text, robots need data that captures physical interaction, and that kind of data barely exists. YouTube videos and footage captured by gig workers are low-fidelity and hard to reconcile with the physical world.

XDOF (pronounced “ecks-doff”), emerging from stealth today, is betting that the next great bottleneck in AI isn’t models or chips, but the data feedback loop needed to teach robots how to interact with the physical world.

The startup aims to build the data pipelines, collection tools, and annotation systems that frontier labs and robotics companies can’t easily build themselves — and has raised $70 million from Thrive Capital, Spark Capital, a16z, Lux, and WndrCo to do it. Co-founder and CEO Philipp Wu says XDOF, which has about 60 employees, is already working with 20 customers, including several frontier AI labs, but cannot name them.

“All of the top labs are trying to pursue robotics,” Wu said. “We’ve already seen some of the downfalls of falling a little bit behind in the language model race … you don’t want to be in this type of situation where you pursue this technology too late, and everyone is in this boat where physical AI is the next frontier.”

Wu ran into this problem himself as a PhD student at UC Berkeley. His focus was on enabling robots to learn skills from large-scale datasets. There was just one problem.

“We didn’t have large-scale data to work with,” he told TechCrunch. “There was this chicken-and-egg problem — we first needed to actually collect data before we could even ask how to train a foundation model for robotics.”

Wu and his future XDOF co-founder and CTO, Fred Shentu, worked on a project called GELLO, a low-cost teleoperation system that lets a human operator control a robotic arm to generate training data. “It ended up becoming a very influential paper in robotics, because a lot of people had similar needs and bottlenecks, and many started leveraging this type of device for data collection,” Wu said.

Spotting the opportunity, Wu, Shentu, and third co-founder and Chief Operating Officer Nemo Jin launched XDOF in October 2024 to provide a data ecosystem for companies pursuing robotics models. Mindful that data provision alone can be a dead-end business, the company is also focused on data cleaning, tooling, and annotation — creating a self-reinforcing feedback loop for robot trainers.

As a starting point, the company is partnering with UC Berkeley’s AI Research lab to release what it believes is the largest collection of high-quality robot training data ever assembled, dubbed ABC. It includes 130,000 trajectories of robot manipulation data, 300 hours of simulation, and 100 hours of evaluations. That kind of scaled-up pre-training data has never been available to academia before.

“We’ve seen in language, image generation, and other fields, that when models and data are released, the community achieves things that you wouldn’t necessarily have expected,” David McAllister, a Berkeley PhD student who helped organize the release, told TechCrunch.

The team has already used the data to train robots on benchmark tasks like folding T-shirts and flattening boxes, or loading AirPods into their cases.

Unlimited degrees of freedom

The company plans to work across three tiers of a data pyramid. The most valuable tier is teleoperation data collected on the actual robot being deployed; next comes teleoperated robots gathering more general data, as with GELLO; and finally “egocentric” data gathered by humans performing everyday tasks, for which XDOF plans to build its own wearable sensors.

“Your camera choice is going to affect the quality of your data — which is going to affect how your hand-tracking algorithm performs,” Wu said. “If you don’t design the hardware well from the start, the data you collect might have very specific problems that you didn’t anticipate.”

The company plans to hire and train armies of teleoperators and egocentric data operators around the world — a labor-intensive model that raises an obvious question: Why aren’t the major labs doing this data production work themselves?

“You need a warehouse of hundreds of thousands of square feet with hundreds of robots,” Wu said. “You need to maintain these robots, calibrate their physical parameters, and properly train operators.”

It’s a build-out that requires focus, capital, and operational scale that most AI labs would rather outsource — which is precisely the market XDOF is betting on.

The name XDOF is a play on the robotics term “degrees of freedom,” which describes the number of independent motions a robot can perform. Your arm, from shoulder to wrist, has seven degrees of freedom. Humanoid robotics company Figure AI’s latest robot has 30. The X in the company’s name captures its ambition: “Arbitrary degrees of freedom, unlimited degrees of freedom,” Wu says.