Fieldai, a robotics startup based in Irvine, California, previously raised $455 million in multiple private rounds to develop what is called the “basic embodied AI model.” Essentially, it is designed to adapt to all environments, from humanoids to four times more self-driving cars.
The company announced its funds on Wednesday. The latest round raised $314 million in August, co-led by Bezos Expedition, Prysm and Temasek. Other Fieldai supporters include Khosla Ventures, Intel Capital, Canaan Partners, and more.
Unlike traditional AI that processes text and images, embodied AI refers to AI that controls physical robots moving through the real environment. Fieldai will build the Field Foundation Model, a general-purpose, embodied AI model rooted in physics. This approach allows robots to quickly learn and adapt to new environments while being aware of risk, Fieldai founder and CEO Ali Agha told TechCrunch in an interview.
“The mission is to build a single robotic brain that can generalize to a variety of robot types and diverse environments,” Aga said. “To get there, you need to manage risks and safety when you go to these new environments. And that’s a fundamental gap in robotics, and traditional models and traditional approaches weren’t designed to manage that risk and safety.”
Agha said the key to making robots safe to learn in new environments is to add layers of physics to these AI models. This addition provides a second set of information for making decisions, particularly in a new environment, as opposed to the fact that the model responds to everything that it does next, just like a traditional LLMS.
He added that small amounts of AI hallucinations are not harmful in certain situations, but that is possible for dangerous environments and robots working with people.
“All of a sudden you started to have that feeling, how much I know, if I don’t know anything, or if I’m making a decision, how confident I am with it,” Aga said. “As the network becomes accessible to it, it starts to make more secure decisions. This not only vomits, ‘Hey, there’s next action’, but also tells you how confident it is.
TechCrunch Events
San Francisco
|
October 27th-29th, 2025
AGHA has worked on the idea for decades across roles ranging from NASA to the Massachusetts Institute of Technology (MIT). He decided to launch Fieldai when he achieved a technical breakthrough that allowed one robot brain to work with different kinds of robots that perform both the same thing and individual actions.
Since launching the company in 2023, Fieldai has secured contracts across the industry, including construction, energy and urban delivery. The company refused to disclose the customer by name.
The funding will support research and development, helping the company increase production, deploy models to customers, and further expand its reach overseas.
Aga compares field approaches to human evolution. “We can perform different tasks in different environments, and have the ability to learn quickly (and) the need for robotics. Yes, we can optimize for one particular use case, but that’s not the market we’re doing.”
We are constantly aiming to evolve and you can help us by providing insights into TechCrunch and your perspective and feedback on our coverage and events! Fill in this research to let us know how we are doing and get the opportunity to win an award in return!
