The first publicly traded humanoid robot company might not be the one with the biggest balance sheet

Agility Robotics' Digit has already moved 100,000 totes inside a GXO logistics facility, and the company says it has $300 million in contract orders from Amazon, Schaeffler, and Toyota Motor Manufacturing Canada. Tesla's Optimus has not shipped a paying customer. Later this year, when Agility's reverse merger closes, the first publicly traded pure-play humanoid robot company will be the smaller one with the working customers, not the one with the larger balance sheet.

Digit is a two-legged, human-shaped robot that Agility says it has been operating in real warehouses and factories for years. The company confirmed a 60,000-square-foot facility in Fremont, California on Thursday, framed as a training and operations base for the robot. The site sits in the same highway corridor as Tesla's planned Optimus manufacturing site, which is why TechCrunch called it "Tesla's backyard." Agility is not hiding the geography; CEO Peggy Johnson told the outlet it is good for the category to have Tesla nearby, even as a competitor.

Outside observers put Agility's active Digit fleet at "dozens" of units. GXO confirmed in November that the robots had passed the 100,000-tote milestone in a live deployment under a multi-year agreement. Amazon, GXO, Schaeffler, and Toyota Motor Manufacturing Canada are named customers on Agility's site. The $300 million figure is a company claim, not an audited backlog, and the deployment count is not disclosed.

Johnson's pitch for why the lead matters is integration, not AI. Humanoid robots only earn their place in a warehouse if they can talk to the warehouse management system, pass the same safety bar as a human picker, and recover when a tote is missing or a conveyor is jammed. Those bars take years to clear with one customer at a time. "It's great to have [Tesla] in the same area as us... it's good to have others in the humanoid space," Johnson told TechCrunch, framing the proximity as category validation rather than a threat.

Co-founder and chairman Damion Shelton pushes the same point from the engineering side. He argues against handing the robot's balance and reflexes to a single end-to-end AI controller, and uses the car industry's anti-lock brake as shorthand. "You really don't want the anti-lock brake controller under AI control," Shelton told TechCrunch. The argument is not that AI is unnecessary; it is that the safety-critical layer of a humanoid that walks next to people has to be a separate, deterministic system. The category definition war, between hybrid autonomy and full model control, will decide which "humanoid" actually ships at scale.

Johnson is leading Agility through a reverse merger the company expects to close later this year, which would make it the first publicly traded pure-play humanoid robot company. Tesla trades as an automaker and an AI platform; it is not investable as a humanoid bet. The first humanoid-only index that follows the merger will not include Optimus. So the moment allocator dollars first have to pick a humanoid ticker, the choice is between a small, revenue-generating company already collecting 100,000-tote reference stories from named customers, and a much larger rival whose humanoid is still a stated expectation.

Two things could break this thesis. Tesla could ship Optimus in commercial volume inside the "sometime next year" window Musk named, and reset the category before Agility's first earnings call. Or Agility's $300 million in contract orders could turn out to be early pilot commitments that do not convert to recurring revenue, the way many robotics pilots have failed to for the last decade. The reverse-merger S-4 will be the first place to test which it is.

The Fremont facility opens now. The S-4 lands this fall. The category's first public investor is about to price operational integration, not balance-sheet size.