London AI Startup Cuts Cement Emissions With $26M Funding

Imagine software smart enough to run a cement plant the size of a skyscraper, operating at temperatures hot enough to melt rock, all by itself from hundreds of miles away.

That's exactly what Gigaton is doing, and investors just bet $26 million they can scale it worldwide. The London startup announced its funding round led by Plural this week, with backing from seven other firms including Cambridge Enterprise and Clean Growth Fund.

Gigaton's technology is refreshingly different from typical AI tools. They don't build chatbots or analytics dashboards. Their machine learning systems sit directly in the control stack of industrial facilities, physically operating the equipment in real time.

Their first target makes sense. Cement production accounts for roughly 8% of all global carbon dioxide emissions, with about 5,000 plants worldwide burning through billions in energy costs annually. These facilities run at 1,500 degrees Celsius in structures stretching 60 meters long and fifteen stories high, generating terabytes of data every day.

The company started by helping plants analyze their processes. But after years spent in industrial control rooms, they realized recommendations weren't enough. The only way to meaningfully slash emissions was to control the hardware directly.

Now their AI operates entire cement plants autonomously from London. The software adapts to changing conditions, optimizes energy use, and reduces carbon output without constant human oversight.

The Ripple Effect

Cement plants are just the beginning. Gigaton estimates tens of thousands of similar industrial assets worldwide share the same characteristics and could benefit from autonomous control. Steel mills, chemical plants, and refineries all run on machinery that generates massive emissions and consumes enormous amounts of energy.

The company's thesis is elegantly simple. The physical world already runs on machines. Those machines just haven't been upgraded with the software needed to operate them efficiently and cleanly.

CEO Josh Vernon puts it plainly: "We aren't building chat interfaces or agents that analyze and recommend. We sit deep in the control stack and build machine learning systems that actually operate industrial plants."

The fresh capital will fund deployments across the world's most emissions intensive manufacturing processes. Each new facility they bring online could prevent thousands of tons of carbon from entering the atmosphere annually.

For an industry often resistant to change, letting AI take the wheel represents a remarkable leap of faith. But with climate targets tightening and energy costs rising, the economic case for autonomous optimization is becoming impossible to ignore.

The machines that power our modern world are finally getting the intelligence to run themselves cleanly.