Agility Robotics Opens Fremont Training Hub in Tesla’s Orbit as Humanoid Race Heats Up

Agility Robotics is opening a 60,000-square-foot training facility in Fremont, California, placing its humanoid robot operations just miles from the Tesla factory where Optimus production is expected to begin. The move underscores how quickly the commercial humanoid race is intensifying, even as Agility leans on real-world deployments and Tesla continues to build out its long-term bet on robotics.

The Oregon-based company is using the new site to speed up training and rollout of Digit, its bipedal warehouse robot that is already working for customers in logistics and manufacturing. Unlike many newer humanoid startups still focused on prototypes, Agility says Digit is generating revenue today and has accumulated a substantial pipeline of contract demand.

The Fremont expansion also carries symbolic weight. It puts one of the best-known independent humanoid robotics companies in the same region as Tesla’s humanoid ambitions, at a moment when Elon Musk is increasingly positioning Optimus as a future cornerstone of the company’s business.

Why Agility’s Fremont move matters

The new facility matters because it shows Agility is not waiting for the humanoid market to mature before building out scale. Instead, the company is expanding near Silicon Valley’s robotics and manufacturing ecosystem to prepare Digit for more deployments while Tesla and other rivals chase the same broad opportunity.

Agility CEO Peggy Johnson framed the location as both practical and encouraging. She said the company welcomes more players in the humanoid space, especially after spending years operating with relatively little competition from companies at comparable scale.

Johnson said Agility sees value in having Tesla nearby, noting that the company is already commercialized and knows what it takes to clear safety reviews, compliance checks, IT integration requirements and warehouse system onboarding.

That commercial experience is a key part of Agility’s pitch. While many robotics startups are still proving that humanoids can work outside a lab, Agility says it has already learned how to bring robots into live facilities where they must fit into existing industrial workflows.

How Digit is being used today

Digit is already operating in settings such as warehouses and factories, where it carries totes and bins for customers including Amazon, GXO, Schaeffler and Toyota Motor Manufacturing Canada. The robot is not being sold as a general-purpose household assistant or a science-fiction-style helper; it is aimed at repetitive, physically demanding tasks in controlled commercial spaces.

Agility says it has secured $300 million in contract orders for its robots. The company has not publicly disclosed how many Digits have been built or deployed, but outside estimates suggest the fleet is still relatively modest, likely numbering in the dozens across pilot and paid installations.

One example Agility has highlighted is a GXO logistics site where Digit robots have moved 100,000 totes. That figure is important because it gives a sense of the operational maturity Agility is trying to demonstrate: not just a machine that can walk, but one that can repeat work reliably enough to be useful in a business environment.

What Tesla’s Optimus push changes

Tesla’s growing focus on Optimus raises the stakes for every other company in humanoid robotics. Musk has recently described Optimus as potentially the company’s biggest product ever, while also saying he expects it to become useful outside Tesla sometime next year.

That sort of timeline matters because it turns humanoid robotics from a research category into a competitive product race. If Tesla can bring its manufacturing scale, software expertise and brand power to bear, it could help accelerate public interest, investor attention and procurement demand across the sector.

At the same time, Tesla’s approach remains forward-looking rather than proven. Agility, by contrast, is arguing that it already has a product deployed in the field and a playbook for integrating with customer operations. The Fremont move is as much a statement about readiness as it is about geography.

What makes Agility’s strategy different?

Agility’s strategy is built around practical automation rather than a broad consumer vision. The company is not planning to sell in-home humanoid robots in the near term, a stance that aligns with the caution of many robotics specialists who argue today’s systems are not yet safe enough for households.

Instead, Agility is focusing on industrial and logistics environments where the conditions are more controlled and the business case is clearer. That approach reflects the reality that humanoid robots must still overcome major hurdles around safety, reliability, maintenance and cost before they can become common in homes.

Why industrial environments come first

Industrial settings come first because they allow robots to be deployed where human exposure can be limited and tasks can be tightly defined. In warehouses or plants, robots can often be restricted to specific routes, payloads and workflows, which makes compliance and integration more manageable.

Agility says the latest version of Digit, expected to be unveiled this fall, will be able to detect humans and will no longer need to stay in a robot-only zone. That suggests the company is gradually expanding the robot’s operational envelope while still keeping the emphasis on safety and controlled deployment.

Co-founder and chairman Damion Shelton said the company is deliberately avoiding the idea that generative AI should directly control the robot’s most critical safety functions, arguing that the safety stack needs a non-generative path.

That distinction is central to Agility’s philosophy. The company is enthusiastic about AI’s role in accelerating development, but it draws a line between creative task generation and the deterministic control needed for safe movement and hardware protection.

How AI fits into humanoid robotics

AI fits into humanoid robotics by making it possible to scale programming and task creation far beyond what engineers could manually code. Agility’s leadership argues that large language models and related generative systems help solve a major bottleneck: the sheer number of tasks a robot might need to learn in the real world.

According to Shelton, the challenge is not simply making a robot move. It is making a robot economically adaptable across many customer-specific workflows, each with different tools, workflows and facility systems. In that sense, AI helps bridge the gap between a few prototype behaviors and a commercially useful product family.

Agility co-founder and chief robot officer Jonathan Hurst has long focused on the foundational robotics problem: creating a bipedal machine that can move safely and reliably through human-built environments. The company’s founders originally developed techniques intended to allow robots to walk on two legs without compromising safety.

The newer AI wave does not replace that foundation. Instead, Agility sees it as a force multiplier that could make it far easier to teach robots new work patterns and deploy them in many more facilities.

What the company will and will not automate

Agility is starting with material handling tasks such as moving bins, totes, picking, kitting and eventually cardboard handling and trailer loading. Those jobs are repetitive, labor-intensive and common enough to offer a large addressable market.

• Moving totes and bins
• Picking and kitting
• Cardboard handling
• Loading and unloading trailers

The company is not yet aiming at household chores, elder care or open-ended domestic assistance. Those use cases remain far more difficult because they require robots to cope with clutter, unpredictable people and less forgiving safety conditions.

Who is Agility Robotics competing against?

Agility is competing against a wave of newer humanoid startups, including Figure, 1X, the Bot Company and Sunday Robotics, as well as the looming possibility that Tesla becomes a major player if Optimus reaches production and real-world use at scale.

What distinguishes Agility is timing. The company is trying to defend its early lead by pointing to actual deployments, actual customers and actual revenue, rather than relying on promises of future capability. In a market where demos can attract attention quickly, commercial proof may become the most important advantage of all.

That said, the field is still at an early stage. No company has yet established a dominant humanoid standard, and the market may ultimately reward a blend of hardware reliability, AI flexibility, manufacturing scale and customer integration.

How close is Agility to going public?

Agility is close to becoming a public company through a reverse merger, a transaction structure that would make it the first pure-play humanoid robot company on the public markets if completed as expected later this year.

Going public would be important for several reasons. It could give Agility more capital, a more visible market profile and a way to position itself as the sector’s commercial leader. It would also create more transparency around its fleet size, deployment count, revenue trajectory and customer concentration.

Public investors tend to reward stories that combine near-term revenue with long-term category potential, and Agility is trying to present itself as exactly that kind of company. It already sells into working industrial environments, yet it is also tied to a future in which humanoid robots may be common across logistics, manufacturing and eventually other sectors.

Why the new facility is designed for training

The Fremont site is meant to do more than provide office space or light assembly. It is intended to help Agility train robots in environments that resemble the field, which should allow Digit to learn faster and be deployed with fewer surprises once it reaches customers.

That matters because robots often perform well in demos but struggle in messy, real-world conditions. Training on-site in a facility designed for realistic scenarios can help a company close the gap between laboratory performance and dependable work output.

Agility says the new facility will support customers who are already in discussion with the company, with Johnson estimating that more than 30 potential customers are currently talking about Digit deployments. That pipeline suggests the company believes demand is growing fast enough to justify the investment now.

Timeline of Agility’s current push

The company’s latest move is part of a much longer arc that began nearly a decade ago, when its founders focused on safe bipedal locomotion and how robots could move through environments made for people.

What happens next for Digit?

The next major step for Digit is the expected release of version 5, which Agility says will be able to sense humans and operate without needing a robot-only zone. If that capability proves reliable, it could widen the robot’s usefulness in mixed human-machine workplaces.

Just as important is whether Agility can convert its current commercial momentum into repeatable deployment economics. A humanoid robot business will ultimately be judged not only by whether the robot can do the work, but also by whether customers can deploy it, maintain it and justify the investment at scale.

The company’s leadership is betting that the fastest route to that future is not a leap into consumer robotics, but a steady climb through warehouses and factories. In that respect, opening in Fremont near Tesla is both a competitive move and a reminder that the humanoid race may be won first by the company that proves robots can make money in the real world.

Bottom line

Agility Robotics is expanding in California at a strategically important moment for humanoid robotics, positioning itself close to Tesla while emphasizing that its own robot, Digit, is already working in commercial settings. With a new training facility, a growing customer pipeline and a planned public listing, the company is trying to convert an early technical lead into lasting market advantage.