For a decade, Agility Robotics has been the most prominent argument that legs beat wheels. Its Digit humanoid walks with a gait borrowed from ostriches, and CEO Jonathan Hurst has spent his career researching legged locomotion — first at Oregon State University’s Dynamic Robotics Laboratory, then in commercial form. So when Hurst confirms by email that the company is now “exploring a range of robot designs and form factors, including wheels,” it lands as more than a product rumour.
The confirmation came to The Robot Report, published yesterday (25 September, NZT). The prompt was a rendering that flashed on screen at the end of Agility’s recent Digit 5 launch video: a wheeled robot wearing more than a passing family resemblance to Digit, if Digit had traded its legs for a set of wheels. Asked whether this is a wheeled variant of Digit or a new machine entirely, Hurst’s answer was refreshingly unspun: “Too early to say! I have guesses and inclinations, but really we’re just starting.” No product name, no timeline.
And there is more than imagery to go on. Agility’s own vision page names a “Wheeled Base Concept” alongside Digit 5 and a light bipedal concept, describing its purpose as “quicker initial access to public facing work” — with no specifications, pricing or timeline, and the caveat that some imagery is illustrative animation. Independent watchers spotted the same thing: animations of a wheeled form working in a grocery store and a data centre in the launch material. So this is not one reporter over-reading a video frame; the wheels are on Agility’s written roadmap, at concept stage.
Why would the legs company hedge on wheels? Hurst’s own framing is the honest version of the answer. “As the safety, reliability and capabilities of legged robots continue to advance, I expect they’ll become the right solution for an increasingly broad range of applications,” he said. “In the meantime, we’re interested in exploring other approaches that could help us bring robotics to more environments sooner.” Read that closely and it is a business statement, not a technical retreat: legs are the long game, wheels are the bridge that pays the bills while the long game matures.
If that sounds like an outlier move, it is actually the industry’s most reliable pattern. Apptronik ships Apollo 2 in both bipedal and wheeled configurations, explicitly designing its wheeled base to conform to existing industrial mobile-robot safety standards so it can drop into customer operations today. UK startup Humanoid took the same path and got receipts: its wheeled HMND 01 Alpha completed an eight-hour autonomous tote-handling shift at Siemens’ Erlangen electronics factory — 60 moves per hour, above 90 per cent pick-and-place success — and the company has a binding agreement to put a four-digit number of wheeled units on Schaeffler’s factory floors by 2032. Even Nucleus, a European humanoid startup, dropped its legs entirely within a month of leaving stealth because plant managers asked bluntly for more payload, multi-hour uptime and “no legs.” When the German factory says no legs, the German factory gets wheels. When the customer list includes Schaeffler and Siemens, that is not a niche — it is the near-term industrial market voting with purchase orders.
Legs and wheels are dividing the industry — we covered the split on stage at the World Robot Conference in Beijing, where wheeled humanoids from Chinese makers stood alongside their legged stablemates as near-equals. Wheels cannot climb stairs, but they cost less, draw less power, and fall over far less often in the warehouse environments where most of these machines will actually work. Hurst is not abandoning the thesis; he is acknowledging that the customer base for “gets the job done today” is much larger than the customer base for “climbs stairs impressively.”
There is a financial context that makes the timing less coincidental. Digit 5, launched earlier this month, dropped the backwards-bending knees that made Agility’s robots instantly recognisable — a design choice aimed squarely at industrial adoption rather than spectacle. The same pragmatism is visible in Agility’s filings: the company is going public via a SPAC merger with Churchill Capital Corp. XI at a $2.5 billion valuation, expected to gross more than $620 million. Its 2025 numbers, disclosed in the S-4, show $1.8 million of revenue against a $140 million operating loss, cushioned by more than $300 million in multi-year Digit v5 orders from a single customer.
A company carrying that kind of burn into the public markets has every reason to widen its addressable market before investors can vote on it. A wheeled product aimed at flat-floored logistics environments would be cheaper to build, easier to certify, and faster to ship than another legged platform — and it would let Agility sell into facilities that have no stairs but plenty of budget.
None of this is confirmed as a product. Hurst’s own words bracket the whole story in “exploring” and “too early to say,” and the vision page offers a concept, not a datasheet. But the direction of travel is the news — and the context around it is no longer Agility-shaped alone. Apptronik sells both formats today, Humanoid’s wheeled robots have passed an eight-hour shift at Siemens, and customers keep saying the same three words: payload, uptime, stability. The company whose robots defined what a legged humanoid looks like is now willing to imagine one on wheels. If even the legs company is hedging, the form-factor debate in robotics is no longer ideological — it is commercial.
