Agility Digit in the Warehouse: The First Humanoid on a Real Commercial Payroll

Most humanoid robots you’ve seen in headlines are demos: a robot folding a shirt in a lab, a robot walking on a treadmill at a trade show, a robot doing backflips for a viral clip. Digit, built by Agility Robotics, is different. It clocks in at a real distribution center, does a repetitive job a human used to do, and its employer tracks its output the way a warehouse manager tracks any other piece of equipment on the floor. That is the entire point of the GXO Logistics deployment, and it’s why Digit is routinely described as the first humanoid robot doing paid commercial work outside a research lab.

The short answer to “what is Digit actually doing in a warehouse” is this: at a GXO-operated distribution facility in Flowery Branch, Georgia — the site handles SPANX shapewear and apparel — Digit lifts totes off autonomous mobile robots and places them onto conveyors, a task that is physically awkward for wheeled automation but comes naturally to a machine with legs, a torso, and two arms. By late November 2025, Agility Robotics announced that Digit had moved more than 100,000 totes in that single live deployment. That’s not a lab benchmark. It’s a production count.

What exactly did GXO and Agility Robotics agree to?

In June 2024, GXO Logistics and Agility Robotics announced what both companies called an industry-first, multi-year Robots-as-a-Service agreement to deploy Digit across GXO’s warehouse operations. The commercial rollout followed a proof-of-concept pilot that ran in late 2023, meaning GXO tested Digit on real freight before committing to a paid, ongoing contract rather than a one-off trial.

The RaaS structure matters as much as the robot itself. Instead of GXO buying hardware outright, Agility retains ownership of the machines and GXO pays for a service — uptime, throughput, support — much the way a company might lease forklifts or contract a staffing agency rather than buying the equipment and hiring the workers directly. That commercial model is central to how humanoid robotics companies are trying to get physical robots into daily operations without asking logistics customers to absorb the capital risk of an unproven robot fleet, a pattern worth understanding if you’re evaluating what these RaaS and per-hour pricing structures mean for the economics of deploying a fleet.

What does Digit actually do on the floor?

Digit’s job at the SPANX facility is narrow and specific: it takes totes off autonomous mobile robots (AMRs) and places them onto a conveyor line. That sounds simple, but it’s a task that sits in an awkward gap for existing automation. AMRs move totes horizontally around a warehouse floor efficiently, but transferring a tote from a mobile cart onto a fixed conveyor at variable heights and angles is exactly the kind of last-foot manipulation that traditional fixed automation struggles to handle economically, especially in a facility with high SKU variability like an apparel distribution center. A bipedal, legged robot with reach and grasping capability can plug into that specific step without a warehouse redesign — the pitch being that Digit adapts to existing infrastructure rather than the other way around.

Why does the 100,000-tote number actually matter?

Numbers like “100,000 totes moved” get repeated constantly, but the reason it’s meaningful isn’t the round figure — it’s what sustaining that count for over a year in a live facility implies about reliability. A demo robot needs to work once, on camera, under controlled conditions. A warehouse robot needs to work every shift, in a facility with real throughput targets, alongside human co-workers and existing safety processes, without becoming a maintenance burden that eats into whatever labor savings it’s supposed to deliver. Reaching a six-figure tote count in production is Agility’s evidence that Digit has moved past pilot-stage novelty into something closer to a dependable, if narrow, warehouse employee.

It’s also worth being precise about what the number doesn’t tell you: it’s a cumulative count from one deployment at one site, not a productivity rate, and Agility hasn’t published comparative throughput or cost-per-tote figures against a human worker or a conventional conveyor-and-sorter system. The milestone demonstrates uptime and task persistence, not necessarily that Digit is currently cheaper or faster than the alternatives it might eventually replace.

What is Digit, physically?

Digit stands roughly 5 ft 9 in (about 175 cm) tall and weighs approximately 65 kg (143 lb) — close to an average adult’s size, which is deliberate, since the robot is designed to move through spaces, doorways, and around equipment built for people rather than for wheeled machines. Its current commercial payload capacity is around 16 kg (35 lb), enough for a standard warehouse tote but not for heavier palletized freight. Some published figures round this payload slightly differently (15.8 kg or 18 kg / 40 lb depending on the outlet and hardware revision), which likely reflects version differences or unit rounding rather than a real discrepancy.

In March 2025, Agility announced hardware and software upgrades to Digit including expanded battery capability, with Agility’s own press materials stating the robot can now run up to four hours between charges, plus a new autonomous docking feature that lets Digit return to a charging station and resume work without human intervention. It’s worth flagging that a handful of secondary aggregator sites cite an “up to 8 hours” runtime for the same hardware generation — those figures aren’t independently sourced and appear to mirror each other rather than cite Agility directly, so the four-hour figure from Agility’s own announcement is the more reliable one to work from.

Digit at a glance: then vs. now

Spec	Current (deployed) Digit	Next-gen Digit v5 (targeted)
Height	~5 ft 9 in (175 cm)	Not disclosed
Weight	~65 kg (143 lb)	Not disclosed
Payload	~16 kg (35 lb)	~22.6 kg (50 lb) — about 43% more
Battery runtime	Up to 4 hours (per Agility, March 2025)	Not yet disclosed
Charging	Autonomous docking (added March 2025)	Expected to carry over
Known live deployment	GXO / SPANX, Flowery Branch, GA	Contracted orders pending delivery

Who is behind Digit, and why is this suddenly a bigger story?

Agility Robotics was co-founded in 2015 by Jonathan Hurst, Damion Shelton, and Mikhail Jones as a spinout of Oregon State University’s Dynamic Robotics Lab — the company’s roots are in legged-robot research, not a pivot from some adjacent industry. Its current CEO is Peggy Johnson, who previously served as an EVP at Microsoft and as CEO of Magic Leap, giving the company an executive with both big-tech and hardware-startup experience heading into its next phase.

That next phase is a public listing. On June 24, 2026, Agility announced a definitive agreement to go public through a SPAC merger with Churchill Capital Corp XI, in a deal valuing the combined company at roughly $2.5 billion and expected to bring in more than $620 million in gross proceeds — about $420 million from Churchill’s trust account plus a roughly $200 million PIPE investment led by Foxconn. The combined company plans to trade under the ticker AGLT. This follows a $400 million Series C round Agility closed in early April 2025, led by WP Global with participation from SoftBank and Amazon. That round’s valuation is reported differently across outlets: SiliconANGLE and Tech Funding News cite a $1.75 billion post-money valuation, while some other secondary sources put the figure closer to $2.1 billion — no two primary sources agree on an exact number, so treat the precise valuation from that round as unresolved rather than settled fact.

What’s harder to dispute is the scale Agility claims heading into its public debut: as of the June 2026 SPAC announcement, the company says Digit is commercially deployed across nine customer facilities and has logged more than 65,000 cumulative operating hours, with named customers including Schaeffler, GXO, Toyota Motor Manufacturing Canada, and Mercado Libre. Agility also says it holds more than $300 million in multi-year contracted orders for the next-generation Digit v5, subject to contractual milestones, and a sales pipeline of more than 30 additional companies.

Is Digit really the “first” commercial humanoid, or just the best-marketed one?

The Robot Report gave Digit its inaugural RBR50 “Robot of the Year” award in 2024, explicitly framing it as the first humanoid robot performing paid commercial work outside a research setting — a meaningful distinction in a field where most competitors were still in pilot phases or closed-door testing at the time. That doesn’t mean Digit lacks competition. Tesla’s Optimus and Figure’s humanoid line are pursuing overlapping bets on general-purpose bipedal labor, with different design philosophies and target industries — see how Figure’s approach to humanoid design compares against Tesla’s Optimus program for a sense of how crowded this race has become. A separate, fast-moving cluster of Chinese humanoid manufacturers led by Unitree and UBTech is also pushing hard on cost and manufacturing scale rather than U.S. warehouse contracts specifically. Digit’s edge so far isn’t necessarily superior hardware — it’s that Agility got a named enterprise customer to sign a paid, multi-year, non-pilot contract and kept the robot running long enough to rack up a genuine production track record.

What happens next for Digit in warehouses?

The near-term roadmap centers on Digit v5, targeting a payload increase to roughly 22.6 kg (50 lb) — about 43% more than the current 16 kg limit — which would let the robot handle a wider range of tote and case weights without redesigning workflows around it. With over $300 million in contracted v5 orders already on the books (pending milestones) and a pipeline Agility describes as more than 30 companies deep, the GXO deployment looks less like a one-off showcase and more like the opening chapter of a broader rollout across logistics operators who are watching Digit’s uptime numbers as closely as its press releases. Whether that translates into humanoids becoming a standard line item in warehouse staffing budgets — rather than a novel supplement to existing AMR fleets — is the question the next few years of contracted deployments will actually answer. For the wider context on how humanoid platforms fit into the broader physical AI robot race, see our overview of humanoid robots and how they work, and for a primer on the terminology used across this beat, check the Physical AI glossary. Coordinating uptime and maintenance across a growing fleet of warehouse robots is itself a discipline — one covered in our broader look at running and maintaining a distributed robot fleet.

For a wider view of how this single deployment fits into the broader humanoid robotics landscape, visit our Humanoids: The Race for the Robot Body hub.