FieldAI, a California-based robotics company, has raised $405M across multiple previously undisclosed rounds to develop what it calls “foundational embodied AI models” — essentially universal robot brains capable of helping machines adapt to diverse and unfamiliar environments.
The most recent $314M round, announced this week, was co-led by Bezos Expeditions, Prysm, and Temasek. Other backers include Khosla Ventures, Intel Capital, and Canaan Partners.
Building adaptable embodied AI
Unlike traditional AI that interprets text or images, embodied AI powers physical robots as they navigate the real world. FieldAI’s approach, called Field Foundation Models, embeds physics into the AI architecture, giving robots a framework to manage uncertainty, assess risk, and make safer decisions.
“The mission is to build a single robot brain that can generalize across robot types and environments,” said Ali Agha, FieldAI founder and CEO. “By grounding models in physics, we help robots not only act but also evaluate confidence levels in their decisions — crucial for safety when working alongside people.”
Agha, who previously worked at NASA and MIT, said this breakthrough enables one AI brain to operate across humanoids, quadrupeds, and self-driving systems.
Advertisement
Applications across industries
Since its 2023 launch, FieldAI has already secured contracts in construction, energy, and urban delivery, though specific clients were not disclosed.
The new funding will be used to accelerate R&D, scale production, and expand globally, with the goal of becoming core infrastructure for the next generation of robotics.
Agha likens the company’s vision to human evolution: “Humans adapt to many environments and tasks. We believe robotics needs the same versatility, beyond single-use cases.”
