The same people who shipped a billion iPhones just put their money where AI can't fake it—the factory floor.
The Summary
- Omni Ventures raised $33 million to back pre-seed manufacturing tech, led by ex-Apple engineers Simon Lancaster and Sabrina Paseman who brought products from prototype to production at scale
- VC investment in robotics and physical AI grew from $4B in 2019 to $26B in 2025, with over $23B raised in the first half of 2026 alone
- The fund targets the "digitization stack" under strategic industries: semiconductors, defense, pharma, aerospace, robotics, and energy
- Writing $700K-$1M checks into startups building software, sensors, AI, and robots that make factories smarter
The Signal
Manufacturing tech has been the venture capital equivalent of flyover country for a decade. Software ate the world, crypto tried to rebuild it, and AI promised to replace everyone. Meanwhile, the people who actually make things got ignored. That's changing fast.
Omni Ventures' $33 million fund is a bet that the next AI wave happens in physical space, not pixels. Lancaster and Paseman aren't SaaS tourists slumming in hardware. They spent years at Apple turning prototypes into products at billion-unit scale. They know the gap between a demo and a shipping container full of finished goods. That gap is where money gets lost and companies die.
"Manufacturing is the layer underneath every strategically important industry."
The timing matters. Three forces converged to make this fundable:
- Hardware got cheap enough to instrument. Sensors, edge compute, and robotics that cost $50K five years ago now cost $5K.
- Labor got scarce enough to automate. Reshoring without workers means you need agents on the floor, not just in the cloud.
- Geopolitics got real enough to reshore. Supply chains aren't just cost optimization anymore. They're national security.
Physical AI, the term Nvidia's Jensen Huang uses for AI that acts in the real world, is where the robotics boom intersects manufacturing tech. Investment numbers tell the story: $4 billion in 2019, $26 billion in 2025, over $23 billion in the first six months of 2026. That's not hype inflation. That's capital chasing margin compression and labor replacement in industries that can't be offshored anymore.
Omni's thesis is infrastructure, not point solutions. They're targeting the digitization stack that sits under semiconductors, defense, pharma, aerospace, robotics, and energy. Software that optimizes production lines. Sensors that catch defects before they ship. AI that schedules machines better than humans with clipboards. Robots that don't call in sick.
The Implication
If you're building AI agents, this is your reminder that the highest-value use cases aren't writing emails or summarizing meetings. They're coordinating machines, optimizing yields, and cutting waste in industries with billion-dollar capex budgets. The real agent economy won't be subscription SaaS. It'll be saving 2% on a $500 million factory line.
For workers, the question isn't whether manufacturing comes back. It's what kind of manufacturing. Reshored factories won't look like the ones that left. They'll run leaner, faster, and with fewer people per unit of output. The jobs that come back will require different skills: managing machines, interpreting data, fixing robots. Physical AI doesn't mean no humans. It means different humans.