Mind Robotics Crosses $1B in Total Funding as Physical AI Comes to Factory Floors

Six months. Three funding rounds. More than a billion dollars raised. And a $3.4 billion valuation for a company that didn’t exist until November 2025. Mind Robotics has become the most striking bet in physical AI this year — and the $400 million Series B announced on May 13 may only be the beginning.

The company was founded by RJ Scaringe, the entrepreneur who previously founded and still leads Rivian, the electric vehicle manufacturer. Scaringe has now raised a combined $12 billion across three startups; the speed and scale of Mind Robotics’ fundraising suggests investors believe physical AI — putting foundation models inside industrial robots — may be the most consequential application of the current AI wave that nobody is talking about loudly enough.

From “Project Synapse” to $1B+ Company

Mind Robotics did not emerge from a garage. It was incubated inside Rivian’s engineering organization as “Project Synapse,” a quiet internal effort to apply the reasoning capabilities of large language models to the practical problem of manufacturing. Rivian’s Normal, Illinois factory — one of the largest EV production facilities in North America — became the proving ground: a real industrial environment where robots needed to handle dexterous, variable, often unpredictable physical tasks that traditional automation could not reliably solve.

The spinout in November 2025 came with $115 million in seed funding. Scaringe remained chairman of Mind Robotics while continuing as Rivian’s CEO — an unusual dual role that raised eyebrows but appears to have accelerated fundraising rather than slowing it. By March 2026, the company had closed a $500 million Series A. The May Series B adds another $400 million.

Kleiner Perkins led the most recent round, alongside new investors Meritech Capital, Redpoint Ventures, SV Angel, Incharge Capital, A-Star Capital, and Garuda Ventures. Existing backers — Accel, Andreessen Horowitz, Eclipse, Prysm Capital, Bain Capital Ventures, and Greenoaks — also participated.

The Technology: Full-Stack Physical AI

Mind Robotics describes its product as a “full-stack platform” comprising three layers: foundation models trained on industrial data, purpose-built robotic hardware, and deployment infrastructure. The framing is deliberately analogous to how AI software companies have positioned themselves — but applied to the physical world.

The foundation models are the core differentiator. Traditional industrial robots are programmed with explicit instructions for specific tasks; they fail — sometimes dangerously — when confronted with variation. Mind Robotics’ approach trains models on real manufacturing data drawn from Rivian’s production operations, with the goal of producing robots that can reason about novel situations, adjust in real time, and handle the kind of messy physical reality that scripted automation cannot.

“Dexterous, reasoning-intensive manufacturing tasks at scale” is how the company characterizes its target — a phrase that covers everything from precision assembly and quality inspection to material handling in dynamic environments. The company has not disclosed benchmark performance figures, but investors’ willingness to value the company at $3.4 billion in less than half a year suggests they have seen compelling internal demonstrations.

The hardware is built in-house rather than adapted from commercial platforms, which gives Mind Robotics control over form factors and sensor integration but also means it carries capital-intensive manufacturing risks that purely software competitors do not.

Rivian’s Unique Role

The relationship with Rivian is both Mind Robotics’ clearest competitive advantage and its most complex strategic dimension. Rivian is simultaneously the company’s first customer, its primary deployment partner, a significant shareholder, and its most important source of real-world training data.

This tight integration means Mind Robotics has something almost no other industrial robotics company possesses at Series B stage: a large-scale production environment where its systems are operating continuously, generating feedback data, and facing genuine manufacturing pressures. Most AI robotics startups spend years in controlled settings before encountering factory-floor conditions; Mind Robotics is running live.

The risk is dependency. A company whose primary customer is also its majority shareholder and training partner faces questions about how it will serve other manufacturers — competitors of Rivian, in sectors like aerospace, consumer electronics, or semiconductors — without creating conflicts or diluting the special relationship that makes the technology work. Mind Robotics has not publicly detailed its strategy for expanding beyond Rivian’s facilities, though the Series B press materials reference “expanding industrial robotics deployment” broadly.

The Physical AI Moment

The timing reflects a broader industry conviction. Several of the most prominent investors in AI — from Andreessen Horowitz to Kleiner Perkins — have publicly argued that the next major value creation wave in AI will not come from another frontier language model, but from applying existing AI capabilities to the physical world: manufacturing, logistics, construction, agriculture.

The market is compelling on paper. Global industrial automation spending is projected to exceed $400 billion annually by 2028, with robotics representing the fastest-growing segment. Yet penetration of AI-native approaches — as opposed to traditional programmatic automation — remains in the low single digits. The white space is enormous.

Competitors are moving fast. Figure AI, Physical Intelligence (Pi), and Apptronik have each raised significant capital for humanoid and semi-humanoid robots, though their approaches differ from Mind Robotics’ factory-focused, full-stack strategy. Boston Dynamics, now owned by Hyundai, has accelerated its push into industrial AI. And Chinese robotics firms, benefiting from a dense manufacturing ecosystem, are developing capabilities that Western investors are beginning to take seriously as competitive threats.

Mind Robotics’ bet — that the combination of Rivian’s production data, Scaringe’s manufacturing expertise, and top-tier investor backing can produce a defensible industrial AI platform — is large and not without risk. But it is also the kind of bet that, if it lands, could redefine what automation means in an era of genuinely reasoning machines.

What’s Next

With over $1 billion in total funding and a clear first-customer reference, Mind Robotics’ next move is likely to announce additional manufacturing partners — signaling that the platform is not a single-customer solution. The company will need to demonstrate that foundation models trained on Rivian’s electric vehicle production can transfer to meaningfully different manufacturing contexts without starting from scratch.

Scaringe’s track record adds credibility. He raised $12 billion for Rivian before it produced a single vehicle at scale; he has a demonstrated ability to attract capital for ambitious, hardware-intensive bets. Mind Robotics is, by design, a different kind of company — more software-defined, more data-driven — but the fundraising pattern feels familiar.

At $3.4 billion, Mind Robotics is already a significant bet. The question is whether the factory floor — not the cloud — will prove to be the most important AI deployment environment of the decade.