From Zero to 150 Robots in Just 20 Months | Mike LeBlanc, Foundation Future Industries

What It Covers

Foundation Future Industries cofounder Mike LeBlanc describes building 150 humanoid robots in 20 months, deploying them in an automotive factory in Georgia on 24-hour shifts, pursuing Department of Defense contracts across Army, Navy, Air Force, and Marine Corps, and planning battlefield testing in Ukraine — while host Craig Cannon maintains skepticism about production claims and technical readiness.

Key Questions Answered

• •Deployment economics: Foundation prices robots at $100,000 per robot per year on a full-service lease model that includes repairs, over-the-air software updates, and a human worker replacement guarantee if the robot goes down. Customers must commit to a minimum of 100 robots per contract, with the largest current contract covering 2,000 units across six total customers in automotive manufacturing.
• •Training methodology: Foundation uses approximately 100 hours of first-person and third-person video data per task to train its vision-language-action models, avoiding heavy reliance on teleoperation or simulation. Engineers wear GoPro chest and helmet cameras alongside tripod-mounted external cameras to capture human task execution, which then feeds a latent-space variable analysis model that learns object relationships and physics rather than mimicking specific movements.
• •Task scoping strategy: Rather than pursuing full generalization immediately, Foundation targets narrow, repeatable industrial tasks — injection-molded part inspection, labeling, and packaging in 15-by-15-foot work cells — then aggregates that real-world data across deployments to build toward a generalizable model. This mirrors Tesla's autonomous driving approach: collect fleet data from constrained real-world use before expanding capability scope.
• •Actuator differentiation: Foundation builds its own actuators in-house, treating them as the primary cost and performance differentiator in humanoid robots. Since actuators represent the majority of a humanoid's bill of materials and determine speed, reliability, and economic viability, vertical integration on this component is positioned as a competitive moat against companies sourcing from third-party suppliers with limited scalability.
• •Military use case prioritization: Defense deployments focus on high-risk, low-duration tasks rather than continuous operation — perimeter sweeps around supply trucks, landmine detection, casualty extraction, and reconnaissance outside fortified positions. Battery life currently runs four hours, which aligns with mission-specific use rather than factory-style 24-hour shifts, with the expectation that robots operate from trucks serving as mobile charging stations.
• •Market segmentation by robot weight class: LeBlanc frames the humanoid market as analogous to the auto industry, with lightweight consumer robots targeting homes at sub-$30,000 price points and heavier industrial robots — Foundation's new model stands six feet tall at 185 pounds with arms rated at 80 kilograms of lift each — commanding premium pricing for ports, defense, and manufacturing jobs currently paying $180,000 annually that employers struggle to fill.