Formic CEO: 95% of US factories have no robots — capex-free robotics-as-a-service can change that

Summary

Formic CEO Salman makes a case that the bottleneck to American re-industrialization is not robot capability — it's deployment infrastructure. With 250,000 factories in the US and 95% of small and mid-size manufacturers running zero robots, the gap is vast. The typical American factory runs fewer than 2,000 production hours a year against a theoretical maximum of 8,700 and a Chinese benchmark of 7,500 hours. The constraint is labor, not demand. Millions of factory jobs sit unfilled because the work is physically brutal and low-paid, particularly in CPG where wages cap around $15 an hour.

The deployment problem, not the robot problem

Salman spent a decade as a VC backing more than 50 robotics companies before founding Formic, and his central argument is that the industry has been solving the wrong problem. Getting a robot to perform a task is roughly 10% of the work; error handling, maintenance, programming, and ongoing management are the other 90%. Of Formic's own deployment costs across hundreds of robots, less than 5% goes to programming the robot itself. The rest is everything else.

A fully automated robot work cell typically costs between a few hundred thousand and a few million dollars, and only about 30% of that is hardware — the robot arm, conveyor, and peripherals. The remaining 70% is custom engineering: integrators, consultants, site evaluation, simulation, and configuration. That cost structure locks out the small and mid-size factories that make up 99% of US manufacturing. Formic's pitch is that it cuts out that 70% through software — using computer vision and LiDAR for automated site scans that compress months of manual evaluation into minutes, AI-driven robot programming, and a full remote management stack covering teleoperation, error recovery, and preventive maintenance.

Why CPG first

Formic started broad — metal fabrication, plastic injection molding, CPG — then narrowed. CPG factories run 20 to 24 hours a day, versus six hours of actual production in a typical metal parts shop. Higher utilization means automation compounds faster. CPG products are also low-value and low-margin enough that ocean shipping is uneconomical, so the manufacturing base is domestic and accessible. And the jobs being displaced — packing heavy cases of drinks, sorting food — are the most physically damaging at the lowest wages.

Formic now has hundreds of robots across more than 100 US factories, producing aircraft and automotive parts, lawnmower and golf cart components, Chipotle nachos, pet food, and laundry detergent.

The business model

The capex-free model — robotics delivered as a service with no upfront hardware purchase — is explicitly central to Formic's go-to-market. Salman doesn't dismiss the "financial innovation" framing but argues the software infrastructure is what makes the model defensible: the cost savings come from automating the engineering work, not just from shifting capital to operating expense.

Robot supply chain

The major industrial robot arm makers are all non-American — Fanuc and Yaskawa from Japan, Mitsubishi, ABB, and Kuka (now Chinese-owned after being acquired from Germany). A robot arm alone is functionally useless without a full work cell: PLCs from Siemens or Rockwell Automation, sensors from companies like Keyence, grippers, safety scanners from companies including Cognex, conveyors, and light curtains. Most deployments today rely on a systems integrator to source and assemble all of it. Formic's software stack is designed to replace much of that integrator layer.

On humanoids, Salman is measured. Six-axis articulated arms cover the vast majority of factory tasks because humans in factories primarily use their arms. Humanoids have a role for novel task types, but the form factor is largely beside the point — at the software abstraction layer, different robot types are variations on the same motion-control problem.

Re-industrializing the US, Salman argues, is a 20-year minimum journey. Formic is trying to compress that timeline, but the scale of the gap — 95% of factories with no automation, utilization rates a quarter of China's — means the opportunity is large and the work is long.