In-depth analysis · About 6 min read
GXO and Agility plan two more U.S. Digit sites for early 2027, with deployment evidence still site-specific
An expanded partnership targets Statesville and Lebanon. The announced schedule, past tote movements, fleet-wide operating hours and factory design capacity describe different things.
- Published here
- Source published
- Source checked
GXO and Agility Robotics announced on October 26 that their humanoid-robot partnership would expand to two more U.S. warehouses. Deployments in Statesville, North Carolina, and Lebanon, Tennessee, are expected to begin in early 2027. The announcement describes a plan and earlier operating experience, rather than confirming that the two new sites are already running Digit or that a repeatable financial return has been independently demonstrated.
Listen to the news
Read this article aloud or keep listening to the news.
Preparing the reader…
Device voice
Speech is handled by your browser or device. AISubHub does not send the text to an audio API. Voice availability and playback with a locked screen vary by device.
Resuming may repeat a short phrase.
At a glance
- The planned e-commerce deployments start with Digit 4 and are to migrate to Digit 5 when it becomes commercially available. No precise start, shipment or Digit 5 availability date is established.
- More than 100,000 totes refers to the existing Flowery Branch facility. More than 65,000 operating hours covers Digit’s commercial operations and customer programs more broadly; the figures cannot be combined into a measured average handling rate.
- RoboFab is described as a 70,000-square-foot facility designed for up to 10,000 robots a year. Design capacity is not actual production or delivery, and Digit 5’s cooperative-safety design is not a risk-free deployment certification.
Two future sites build on an older commercial relationship
The new element is the intended extension to Statesville and Lebanon for repetitive material handling in e-commerce fulfillment. The release expects work to begin in early 2027 with Digit 4, followed by a move to Digit 5 after commercial availability. That sequence has a dependency: an announced next-generation model is not necessarily a product ready to ship. The text does not specify exact commissioning dates, robot counts at the two locations or a dated migration schedule.
The relationship itself is older. A proof-of-concept pilot began at Flowery Branch, Georgia, in late 2023, and a multi-year Robots-as-a-Service agreement followed in June 2024. Management treats a customer expanding its use as evidence of value. AISubHub analysis: renewal or expansion can indicate willingness to continue, but it does not disclose total cost, labor effects, payback time or a comparison against an alternative. The older agreement is context, not a second new event.
The workflow is tote handling within existing automation
At Flowery Branch, the described work is moving totes and placing them onto conveyors alongside existing automation. It is a bounded workflow within a logistics operation, not evidence that the robot autonomously performs every warehouse task. For the two planned sites, the release identifies repetitive material handling and e-commerce fulfillment without publishing detailed task lists, floor layouts, exception policies or acceptance criteria.
Agility Arc is described as a cloud-based platform for managing robot fleets and integrating them into customer operations. That is a role in orchestration and workflow integration, not a published specification of every interface or a guarantee that another site can adopt the setup unchanged. AISubHub analysis: handoffs to conveyors, existing automation and people form operational boundaries. A repeat deployment needs those boundaries checked locally; the announcement supplies no independent integration test or full warehouse-autonomy evaluation.
The two operating figures have different denominators
The release reports more than 100,000 totes moved at Flowery Branch since the earlier deployment. Separately, it says Digit has accumulated more than 65,000 hours across commercial operations and customer programs. One number belongs to a particular facility and handling activity; the other aggregates broader operating experience. They are company-reported quantities, not a matched numerator and denominator from one disclosed study.
Dividing tote movements by those hours would therefore manufacture an average that the source does not measure. The same figures do not establish uptime, successful attempts, safety incident rates or labor productivity. AISubHub analysis: a useful comparison would need aligned locations, fleet sizes, observation periods, completed and failed tasks, and maintenance or waiting time. Those details are not supplied. Experience can inform product development without the release proving a universal throughput or reliability result.
A new robot and a factory ceiling are not delivery evidence
Digit 5 was unveiled on September 15, 2026, according to the release. Agility describes it as designed for cooperatively safe work near people and a broader range of industrial workflows. These are intended design properties, not evidence that the two new warehouses have already accepted the robot or that every human interaction is certified safe. The future migration still depends on the robot becoming commercially available; no precise date is given here.
The manufacturing statement has a separate boundary. RoboFab in Salem, Oregon, is described as occupying 70,000 square feet and being designed for annual capacity of up to 10,000 robots. Neither floor area nor a design ceiling is a count of units produced, ordered or delivered. The release also points to expanded deployment, software and service capabilities. AISubHub analysis: manufacturing, commissioning and ongoing support are different parts of scaling; stated capacity alone cannot verify any customer’s delivery schedule.
Repeat deployment still needs local acceptance and support
A second and third site could test whether work learned in one warehouse can be reused elsewhere. AISubHub analysis: the meaningful question is what transfers and what must be adapted. Site acceptance, unusual tote or conveyor conditions, maintenance access, recovery after an interruption and coordination with people still need clear responsibilities. These are engineering questions raised by the proposed expansion, not results that AISubHub observed or claims that the release publishes complete answers to them.
The source is a joint commercial announcement with explicit forward-looking statements. It identifies uncertainty around expected deployments, migration, skills, manufacturing and commercial expansion, and warns that plans are not guarantees. Its corporate outlook is not evidence of completed operations. No independent site logs, comparison method, quantified return, deployment price or customer acceptance report is provided. AISubHub has not visited the warehouses, run the robots or audited either company’s metrics. The plan should be followed as a plan, with completion assessed from later evidence.
Why it matters
AISubHub analysis: the expansion is useful as a concrete proposed move from one operating location to several, rather than a new general claim that humanoids can do warehouse work. Keeping the schedule, workflow, site-level tote count, broader operating hours and manufacturing ceiling separate makes it possible to ask what later deployment evidence would actually confirm.
Limits and availability
The evidence is one official commercial release. New-site operations, robot quantities, exact dates, pricing, ROI, incident rates, detailed acceptance conditions and verified safety performance remain unestablished. Design intent is not certification; company-reported history is not an independent test. Broader corporate and commercial forecasts remain forward-looking. No third-party article, separate customer announcement, investment claim or image-derived metric is used.
Sources and evidence
Source published: · Source checked:
AI-assisted editorial articles checked against linked primary sources. Vendor claims are attributed; inclusion is not endorsement.