Agility plans a public debut as a humanoid robotics company

From university research to factories

Nearly a year ago, Figure AI CEO Colby Adcock predicted that Agility would go bankrupt within 12 months, responding to the company’s mockery on social media. Instead, Agility is preparing to go public.

The company was founded before the current investment boom in physical AI. In 2015, a group of researchers working on safe bipedal robot locomotion spun it out of Oregon State University. Agility focused on using its Digit robots in industry. Its nine customers include Schaeffler, GXO, Toyota Motor Manufacturing Canada and Mercado Libre.

Digit robots move heavy containers. At one GXO logistics facility, they have moved more than 100,000 loads of this kind. CEO Peggy Johnson said the company chose repetitive manual tasks that are hard to staff, rather than attention-grabbing stunts such as dancing and backflips.

Compared with robots from Figure, Tesla, 1X and Boston Dynamics, Digit may seem slow. But it is already working at customer facilities and generating revenue — though that revenue was just $174,000 in the first six months of 2026, most of it from related parties. Agility lost $362 million over the same period, making it important for the company to put its new investment to work successfully.

What Agility is preparing for Digit 5

Agility has planned contracts worth $300 million from 30 customers for its next-generation robot, Digit 5. Unlike the previous model and most humanoid robots, it will be able to work safely alongside people, rather than only in an isolated work area. Johnson called these contracts one of the main reasons the company decided to go public, but did not name the customers or say whether investors are among them.

To close the deals, Digit 5 must meet performance, operating-time and object-placement accuracy targets that have not yet been disclosed. Agility has not shown prototypes of the model. They are planned for CES 2027, with broader availability promised by the end of that year.

In May 2026, four customers signed up for an on-site robot evaluation. This step before a full-scale deployment will cost the customer $500,000. Digit can currently lift heavy loads and place them with 98% accuracy, but, like most modern robots, it cannot yet handle complex manipulation tasks.

The economics and limits of automation

Agility estimates that its robot could save a customer $100,000 a year under a “robot as a service” model; buying the robot could yield greater savings. But those calculations depend on Digit 5’s capabilities: how accurately it performs tasks and how long it can operate without stopping to recharge or receive maintenance.

Johnson gave the example of a person lifting a 50-pound object. At some companies, this is an unofficial threshold beyond which a task is considered a risk for injury. Digit can repeat the operation many times, so, Johnson said, it is better suited to the task.

Two key questions remain for Agility: how quickly it can find new customers for Digit 5, and how quickly it can improve the robot’s capabilities. It is also unclear when generative AI methods will let robots handle a wide range of objects with dexterity. Training on video and remote-control data has not yet produced reliable general autonomy. At the same time, researchers in physical AI see promise in adapting transformer architectures to different types of data.

Johnson noted that no public dataset of robot movements exists, so many companies are trying to generate data through simulations. Agility uses these methods too, but it also has 65,000 hours of real work data gathered from customers at operating facilities. Johnson said these data help speed up Digit’s training.

Raising capital and scaling production

Johnson said the money from the SPAC deal would cover Agility’s current needs, but did not claim it would be enough for the company to operate independently. She did not disclose Digit 5 production plans. Agility says its factory can produce 10,000 robots a year.

The company intends to keep focusing on customer requests and tasks at their facilities. Co-founder and CTO Jonathan Hurst described the development path: first, working with containers and totes; then, order picking; and later, handling cardboard and loading and unloading semi-trailers. In his view, this path could eventually lead to a market for 100 million robots and a trillion-dollar company.

Dataist's take: Going public will give Agility capital to grow, but it does not resolve questions about profitability. Current revenue is small relative to losses, and Digit 5 still has to prove it can meet its stated specifications. The company has customers, contracts and accumulated real-world work data, but the pace of new deployments and the robot’s ability to perform more complex tasks remain uncertain.