Agility opens a Fremont hub as Digit humanoid robots chase commercial scale

Humanoid robots just got a Bay Area training campus with a clear commercial brief. Agility Robotics opened a 60,000-square-foot Fremont facility this week to train, test, and ship new skills for Digit, its bipedal warehouse and factory platform.

The company announcement and TechCrunch coverage both put the site next to the same industrial corridor where Tesla is preparing Optimus production. That geography makes for a neat headline. The marketplace story is different: Agility is betting that software talent, field-like practice floors, and named customer deployments matter more than who starts building robots first on a converted car line.

What Agility opened

Agility's July 16 announcement, published on the company site and PR Newswire, frames Fremont as a Physical AI and capabilities hub. Engineering teams will use the space to advance the AI that helps Digit learn new tasks and run them in customer environments. The building complements RoboFab, Agility's Salem, Oregon manufacturing plant.

Hiring plans are concrete. Agility says it will house nearly 200 existing and new people across hardware engineering, AI and machine learning software, and field operations. Fremont Mayor Raj Salwan welcomed the site as advanced manufacturing and robotics investment for the city.

CEO Peggy Johnson told the company release that Silicon Valley talent should help Digit gain capabilities faster and put them to work for customers immediately. That is an operator pitch, not a demo-day pitch. Buyers on Buy My Robots should hear it as a signal about support density: more field staff and software capacity usually means faster skill packs, denser remote support, and clearer upgrade paths for units already on a floor.

Why Digit, not Optimus, is the near-term product

TechCrunch notes that Tesla still expects Optimus to become a huge product once it is useful outside Tesla sometime next year. Agility's counter is quieter. Digit already generates revenue moving totes and bins for customers including Amazon, GXO, Schaeffler, Toyota Motor Manufacturing Canada, and Mercado Libre.

Agility's own release lists Schaeffler, GXO, Toyota Motor Manufacturing Canada, and Mercado Libre as active humanoid deployments. It also says the company has secured more than 300 million dollars of multi-year orders for Digit v5, subject to contractual milestones, plus a pipeline of over 30 customers. Treat the order figure as company-stated backlog, not cash collected. Still, backlog tied to a named next hardware generation is rarer than vague "interest" language in this category.

TechCrunch reports that Digit currently works in human-free zones, and that Digit v5, expected this fall, is meant to sense people and leave those robot-only cages. That version detail matters for anyone comparing listings. A Digit sold today as a tote mover in a fenced aisle is not the same product thesis as a Digit v5 sold for shared-floor work. Ask which software stack, safety case, and zone rules transfer with the machine.

Public markets and the Fremont timing

Agility is expanding ahead of a planned public listing through a business combination with Churchill Capital Corp XI. The company says the deal would make it the first publicly listed U.S. pure-play humanoid robotics company, subject to the usual shareholder, regulatory, and closing conditions.

That IPO path changes how secondary buyers should think about documentation. Public filings force more disclosure on orders, deployments, and risk factors. If you sell Digit inventory or Digit-adjacent tooling after a listing, expect buyers to compare your hours and maintenance logs against whatever Agility eventually reports. If you buy, use those filings as comps instead of relying only on press quotes.

Fremont also sits near Tesla's Optimus manufacturing bet. Johnson told TechCrunch it is useful to have Tesla nearby after years when Agility felt alone in the humanoid space. Competition for talent and suppliers will rise. So will buyer questions about who has real plant hours versus who has production theater.

What marketplace users should do with this

If you operate warehouses or plants, Agility's message is that skill training is becoming a product feature. A humanoid that can only run one tote workflow is a fixed asset. A platform with a Bay Area hub pushing new skills into the field is closer to a serviceable fleet. When you evaluate Digit against Figure, Tesla Optimus, Unitree, or Boston Dynamics Atlas, ask:

Which tasks are production proven today, and which are still lab demos?
Does the vendor have a named training or software hub tied to customer deployments?
Are order numbers milestone-gated, and what happens if a milestone slips?
Will Digit v5 human-sensing change your facility layout, insurance, and safety case?

Sellers should lean into provenance. A used Digit with GXO-style tote counts, maintenance records, spare actuators, and transferable software sits in a different price band from a lightly used show unit. Pricing guides on this site already treat hours and license transfer as first-order inputs. Agility's Fremont move raises the bar again: buyers will ask whether the unit can receive the next skill pack, not only whether it still walks.

Operators shopping listings should also separate form factor from proof. Walden's wheeled humanoids and Agility's bipedal Digits both chase plant work, but they answer different floor constraints. Digit's Fremont hub does not settle that debate. It does make bipedal commercial hours harder to dismiss as vapor.

What to watch next

Watch four checkpoints. First, whether Agility publishes more Digit v5 milestone detail behind the 300 million dollar order claim. Second, whether the Churchill XI combination closes in 2026 and what the S-4 says about deployments and backlog quality. Third, whether Digit v5 ships human-sensing on the timeline TechCrunch reported. Fourth, whether Tesla's Fremont Optimus line starts slow production and how buyers compare those early units with Digit fleets that already have customer names attached.

Agility did not invent humanoid robots this week. It opened a place to make Digit better at jobs that already pay. For anyone buying, selling, or running humanoid robots, that is the practical news: commercial scale now looks like training campuses and contract backlogs, not only factory photos next door.