Agility Robotics Expands Operations in Direct Competition with Tesla

Agility Robotics is set to open a 60,000-square-foot training facility in Fremont, California, to develop its humanoid robot models. The site is located near the factory where Tesla plans to begin producing its Optimus robots this year.

Tesla has placed increasing emphasis on the Optimus project. Elon Musk recently stated that he anticipates it will become “the biggest product ever” once it can be used outside of Tesla’s own operations sometime next year.

Although Agility Robotics does not have access to Tesla’s substantial funding, it already possesses a functional robot named Digit that is in active use across various industries. This robot generates revenue by transporting boxes and containers in manufacturing and warehouse environments for clients such as Amazon, GXO, Schaeffler, and Toyota Motor Manufacturing Canada. The company has also secured $300 million worth of contract orders for its robots.

“It’s wonderful to have [Tesla] operating in the same region as us,” said CEO Peggy Johnson to TechCrunch. “For a long time, Agility was working alone in this space, so it’s great to see other companies entering the humanoid robotics field as well. We’ve successfully commercialized our technology, and now we understand exactly what’s required to meet industry safety standards, regulatory requirements, and integration needs with existing IT infrastructure and warehouse management systems.”

Agility has not disclosed the exact number of Digits robots it has built or deployed, but external analysts estimate that dozens have been used in pilot projects or for generating revenue. The company has indicated that its Digits units have already moved 100,000 containers at a GXO logistics facility.

Johnson is currently steering Agility through a reverse merger that is expected to establish the company as the first pure-play humanoid robotics firm listed on public stock markets later this year. Founded in 2015 by a group of researchers who developed advanced techniques enabling robots to walk safely on two legs, Agility aims to leverage its early advantages over newer AI-driven robotic startups such as Figure, 1X, the Bot Company, and Sunday Robotics.

While transformer-based neural networks that powered the development of large language models also hold promise for advancing robotic behavior, Agility is pursuing a more practical approach to achieving autonomy.

“Think about self-driving cars as a non-humanoid example,” explained Damion Shelton, co-founder and chairman of Agility Robotics, to TechCrunch. “You certainly don’t want the anti-lock brake system under AI control. For humanoid robots, all safety-related systems need to follow a path that avoids reliance on generative AI — you definitely don’t want to experiment with creative approaches to your safety mechanisms.”

That said, AI does help fulfill the goal of scaling up robotic capabilities.

“Early on, when [Bruce Leak, inventor of Quicktime and a member of Agility’s board] asked us how we planned to write software for our robots, we didn’t have a clear answer,” Shelton said. “The number of tasks a robot could perform is far greater than the number of engineers available to program them. Generative AI resolves this issue completely.”

The new training facility is designed to speed up the deployment of Agility’s robotic solutions. Johnson stated that more than 30 customers are currently in discussions with the company regarding the use of Digit, and it will be at this facility that the six-foot-tall robots learn new skills in environments similar to those they will face in real-world applications.

Unlike many newer competitors entering the humanoid robotics space, Agility does not plan to offer household versions of its robots in the near future. This perspective aligns with most independent robotics experts, who believe that today’s most advanced robots are not yet safe enough for consumer use. While current Digits models operate in spaces free of humans, version 5, scheduled to be released this fall, will include human detection capabilities and won’t require isolation in dedicated robot zones.

Jonathan Hurst, co-founder and chief robot officer at Agility, said that the company has plenty of work ahead of it even just within the manufacturing and logistics sectors.

“Let’s start with handling boxes and containers,” Hurst told TechCrunch. “After that, we can move on to tasks like picking items and assembling products. Then we can tackle more complex challenges such as working with cardboard, loading and unloading tractor trailers, and so on. Once we reach 100 million robots, imagine what a trillion-dollar company we could become.”