FieldAI raises $405M to scale robot intelligence for real-world autonomy

FieldAI announced $405M across two consecutive rounds to expand its hardware-agnostic autonomy platform; the company’s roots sit in field robotics, DARPA programmes, NASA JPL, DeepMind and large-scale robot deployment.

FieldAI announced on August 20, 2025 that it had raised $405 million across two consecutive funding rounds to expand its autonomy software for robots operating in complex physical environments. The Irvine, California company said the funding would support global expansion, continued development across locomotion and manipulation, and hiring as it worked to double headcount by the end of the year. Investors included Bezos Expeditions, BHP Ventures, Canaan Partners, Emerson Collective, Intel Capital, Khosla Ventures, NVentures, Prysm, Temasek and others.

FieldAI builds autonomy software for robots working in sites where GPS, fixed maps and pre-planned routes are unreliable. Its software is designed to run across different machines, including quadrupeds, humanoids, wheeled robots and passenger-scale vehicles, with customer work spanning construction, energy, manufacturing, urban delivery and inspection.

The company was founded in 2023 by Ali Agha, a former NASA Jet Propulsion Laboratory robotics technologist. Agha’s work before FieldAI included DARPA Subterranean Challenge, DARPA RACER, NASA Autonomous Mars Cave Exploration and coordinated Mars helicopter-rover autonomy. That background places FieldAI’s starting point in field robotics, where robots have to work through poor signals, changing terrain and incomplete information.

The Team

FieldAI’s senior team includes people with work across NASA JPL, DARPA programmes, Google, DeepMind, Carnegie Mellon and Toyota Research Institute. Shayegan Omidshafiei leads AI and science after work at Google and DeepMind on large-scale AI models, multi-agent reinforcement learning and sequential decision-making. David Fan leads innovation and technology after chief engineer and chief technologist roles on DARPA SubT and DARPA RACER.

The Technology

FieldAI calls its core system Field Foundation Models. The company describes the models as risk-aware and physics-first, built around a Belief World Model that reasons under uncertainty. The system combines localization, planning and mapping capabilities for complex job sites, with edge operation and support for different robot bodies.

FieldAI develops autonomy for robots operating without prior maps, GPS, pre-planned routes or external communication.

The Round

FieldAI said the funding followed customer adoption, expansion contracts and deployments across hundreds of complex industrial environments. The company said its systems were already in daily operations at customer sites in Japan, Europe and the United States.

Have a robotics update Korthos should review? Send news, deployments, product releases, funding rounds, research, or media to [email protected] or reach out on X at @agkorthos.