Foundation Robotics — American-built humanoid robots for industry and defense

• April 2024
  Founded
• San Francisco, CA
  Headquarters
• Phantom MK-1
  Latest flagship (2025)
• ~$21M disclosed
  Funding raised
• USA
  Country
• Private
  Status

About Foundation Robotics

Foundation Robotics Labs, Inc. — commonly referred to as Foundation and occasionally as Foundation Future Industries — was founded in April 2024 by Sankaet Pathak (CEO, formerly co-founder and CEO of the banking-as-a-service platform Synapse Financial Technologies), Arjun Sethi (CEO of venture firm Tribe Capital, which led Foundation's pre-seed round), and Mike LeBlanc (co-founder of security-robotics company Cobalt Robotics and a 14-year U.S. Marine Corps veteran). The company is headquartered in San Francisco, California, and operates an in-house manufacturing facility it calls Robo Factory 1.

Foundation moved unusually fast from a standing start. In August 2024 — four months after launching — the company closed an approximately $11 million pre-seed round led by Tribe Capital, and in the same month acquired Boardwalk Robotics, a Florida-based humanoid startup founded in 2017 and a longtime commercial partner of the Institute for Human and Machine Cognition (IHMC). That acquisition brought Foundation the Alex humanoid platform, an experienced hardware team, and — most importantly — the cycloidal, rolling-contact actuator technology that now underpins the Phantom program. It also came with an inherited approximately $1.8 million U.S. Air Force SBIR award, giving the young company a foothold with U.S. defense buyers on day one.

Roughly 13 months after founding, Foundation publicly unveiled its first-generation humanoid: Phantom MK-1. The company has since gathered talent from Tesla, Boston Dynamics, SpaceX, and 1X Technologies, with manufacturing leadership reportedly drawn from engineers involved in the Tesla Model X and Model Y production ramps and researchers recruited from IHMC and the Technical University of Munich. As of mid-2025 the company employed approximately 45 people. Reported subsequent financing has taken total disclosed private funding to approximately $21 million by mid-2026, with press reports of a larger round of up to $500 million at a target valuation of up to $3 billion still in the market rather than closed. Separately, Foundation is understood to have secured approximately $24 million in U.S. Army, Navy, and Air Force research contracts, including SBIR Phase 3 designation.

Foundation's stated positioning is deliberately American and deliberately dual-use: build a general-purpose humanoid engineered in the United States, assembled in San Francisco, and capable of being deployed on both the factory floor and — in properly authorized configurations — in defense-adjacent roles. That places Foundation in direct competitive overlap with the current AI-boom cohort of general-purpose humanoid developers, which includes Figure AI, Apptronik, Agility Robotics, 1X Technologies, Tesla Optimus, and Boston Dynamics. In February 2026 Foundation confirmed sending two Phantom MK-1 units to Ukraine for a frontline reconnaissance test, and has publicly stated a manufacturing ambition of approximately 50,000 humanoid robots produced by the end of 2027 across industrial and defense orders — a scale-up target that, if achieved, would put the company among the highest-volume humanoid programs in the world.

Technology stack

Foundation's platform is a vertically integrated combination of custom mechatronics — most notably its cycloidal actuator design — camera-first perception, and an in-house physics-informed control model. Unlike several peers, Foundation has publicly emphasized not using LiDAR on Phantom, treating high-resolution vision plus proprioception as sufficient for the industrial and defense tasks it is targeting.

Perception

Phantom MK-1 uses a multi-camera vision system — publicly described as six to eight cameras distributed around the head and torso — providing wide-angle situational awareness without a rotating LiDAR. That camera-first approach lowers bill-of-materials cost and eliminates a moving optical assembly, at the price of pushing more work onto the vision model. Foundation supplements the vision stack with proprioceptive feedback from its cycloidal actuators, whose sub-1 Nm back-drivability makes joint-level force sensing usable for contact-rich manipulation without dedicated six-axis wrist force-torque sensors on every arm.

Cortex control model and fleet coherence

Foundation refers to its onboard control system as Cortex, described publicly as a physics-action model that maps perception and language input to actuator commands. On top of Cortex, Foundation has publicly promoted a software layer it calls fleet coherence, which allows multiple Phantom units to share a real-world data flywheel — pooling task, environment, and failure data across the fleet so that improvements accrue faster than they would from single-robot learning. That fleet-scale learning framing places Foundation in the same conceptual camp as Figure's Helix VLA model and Tesla's Optimus data-collection strategy.

Actuators, power, and compute

The mechanical differentiator on Phantom is its proprietary cycloidal, rolling-contact actuator. Foundation has publicly stated a peak torque of 160 Nm, upper-body degrees of freedom of 19, back-drivability below 1 Nm, and a walking speed of up to 1.7 m/s. The hands are five-fingered, anthropomorphic, and tendon-driven for human-scale dexterity. Onboard compute is supplied via a public partnership with AMD, which provides the GPU-class silicon that runs the Cortex model and the vision pipeline — making Foundation one of the more notable non-NVIDIA compute wins in the humanoid category. Battery runtime has not been publicly disclosed as of the latest reporting.

Power management on a platform of this class is punishing. Each joint's servo controller must deliver hundreds of watts of instantaneous torque without thermally throttling, which drives the use of high-efficiency wide-bandgap semiconductors — SiC and GaN — in the motor-driver stages and requires tightly matched precision passives on every gate-drive and current-sense channel. The onboard battery pack integrates cell balancing, thermal management, and multi-stage protection ICs; the reliability and quality of those battery-management ICs and their supporting MLCC capacitors and precision resistors is often the practical ceiling on how long a humanoid platform can run between charges and how gracefully it handles transient loads on the motor rails.

The robotics supply chain — where Hybrid Electronics fits in

A modern humanoid robot is one of the most component-dense products ever built. A single Phantom-class platform depends on thousands of individual electronic components: precision BLDC motor drivers for the high-torque cycloidal actuators, high-resolution encoders and current-sense front-ends, IMUs and 6-DoF force-torque sensors, multi-camera image sensors and camera-serial interface bridges, GPU compute boards, battery-management ICs, wide-bandgap SiC/GaN power semiconductors, high-speed board-to-board connectors, precision resistors, and thousands of MLCC capacitors distributed across every power and signal rail.

Building that kind of hardware puts enormous pressure on the electronic-components supply chain. Prototype pilots and low-volume production runs frequently hit allocation, EOL notices, and long lead times on the exact parts they need, and R&D and field-service teams routinely have to support older platforms that use semiconductors no longer in mainstream distribution. During the current humanoid robotics and general-purpose robot AI boom, this pressure is amplified: dozens of well-funded teams are chasing the same short list of high-torque motor drivers, precision IMUs, GPU compute modules, and automotive-grade power semiconductors, and franchise stock on many key parts is now measured in months of lead time rather than weeks.

Hybrid Electronics has been sourcing electronic components for industrial and embedded systems since 1995. For robotics teams — humanoid, mobile, industrial, and research — we specialize in finding the parts you need when mainstream distribution can't:

Trademark and affiliation notice. Foundation Robotics®, Foundation®, Phantom™, and Phantom MK-1™ are trademarks of Foundation Robotics Labs, Inc. Hybrid Electronics is an independent electronic-components sourcing company and is not affiliated with, endorsed by, or authorized by Foundation Robotics Labs, Inc. All information on this page is compiled from public sources for educational and reference purposes.