Gigaton's industrial AI win: cost, carbon and control for heavy industry

Suggested URL slug: `gigaton-industrial-ai-uk-climate-tech-win`

Meta title: Gigaton’s industrial AI win: cost, carbon and control for heavy industry | Clarity Growth

Meta description: London-based Gigaton (formerly Carbon Re) raised $26m to run cement plants with autonomous AI. What its customer results teach growing UK firms about systems that actually pay off.

---

Most “AI success stories” still live in slide decks. Gigaton’s live in control rooms where kilns run hotter than 1,400°C.

In June 2026, the London company — spun out of University of Cambridge and UCL research, and formerly known as Carbon Re — closed a $26 million Series A led by Plural. Existing backers including Planet A Ventures, Cambridge Enterprise Ventures, the UCL Technology Fund and Clean Growth Fund joined, taking total funding past $35 million. The raise is a bet that autonomous process control can cut fuel spend and emissions in heavy industry now, not after a decade of pilot theatre.

For growing UK businesses outside cement and steel, the useful lesson is not “buy an industrial AI platform.” It is how Gigaton earned trust: deep domain expertise, a clear commercial return, operators kept in the loop, and systems that replace brittle manual work instead of decorating it.

The problem was not another dashboard

Energy-intensive plants face volatile fuel costs, messier alternative fuels, and control stacks designed for a simpler world. Cement alone accounts for roughly 8% of global CO₂ — a share that has barely moved despite years of net-zero pledges.

Gigaton’s early product was advisory: recommendations on a screen. Operators pushed back. Their jobs were already busy enough, especially when solid recovered fuel varies in moisture and calorific value. As co-founder Buffy Price told Climate Solutions News, plants did not want another list of set-point suggestions every quarter of an hour.

So the company moved from advice to autonomous control. The platform simulates process behaviour, forecasts each adjustment, then changes parameters such as fuel mix, kiln speed and oxygen — continuously retraining on live plant data. Operators can still take manual control. Oversight stays human. The grind of constant micro-adjustment does not.

Price had expected resistance to AI controlling industrial kilns. Instead, operators were increasingly keen to step back — helped, she argues, by everyday familiarity with large language models changing what “AI” feels like on the shop floor.

Numbers that survive a board meeting

Gigaton reports deployments across ten plants in Europe, Turkey, India, Southeast Asia and South America. Named customers include Adani Cement, Heidelberg Materials, Holcim and Mannok Holdings DAC.

Company and investor reporting puts annual operational savings for an average-size plant at roughly $1.5 million to $3 million, driven mainly by lower fuel costs. Current carbon savings are reported at about 10,000 to 15,000 tonnes of CO₂ per plant per year, with a target of 30,000 to 50,000 tonnes as models expand. Cambridge Enterprise summarises customer outcomes as up to $1–3 million in yearly savings alongside material emissions cuts.

Those numbers matter because they let Gigaton sell outside pure climate branding. Producers can justify the platform on cost alone; decarbonisation is a powerful second benefit. Mannok’s Kevin Lunney puts the operational truth plainly: moving to alternative fuels is “genuinely harder” than running on consistent coal. The upside is large — but only if control-room teams understand and trust the new way of running the plant.

UK roots, global plants

The company plans to grow from about 33 people toward roughly 100 by the end of 2026, and to lift active deployments toward 70–75 plants, extending into steel, glass, mining and petrochemicals — with sector specialists, not generic “AI trainers.” An earlier Innovate UK grant funded steel and glass feasibility work.

Policy is a tailwind, unevenly. The EU’s Carbon Border Adjustment Mechanism began phasing in reporting from January 2025. The UK is still aligning carbon pricing with the European framework. Gigaton is not a substitute for carbon capture: roughly 40% of cement emissions sit on the fuel side it can influence; the rest needs CCS. Stable plants, though, make capture more viable — including at customer sites such as Heidelberg Materials’ Padeswood plant in Wales.

CEO Josh Vernon’s pitch is commercial, not theatrical: costs producers struggle to control, carbon they struggle to reduce, and plants that were not built for today’s fuel mix.

What growing UK firms should steal

Stop decorating broken processes. Advisory dashboards that dump more decisions on busy people usually fail. Automate the repeatable loop; keep humans on exceptions.
Lead with a P&L metric. Fuel cost — or for an SME, quote cycle time, fulfilment errors, cash collection days — beats vague “digital maturity.”
Insist on domain depth. Gigaton’s edge is cement knowledge inside the AI team.
Prove on live systems, then scale. Ten real plants beat a hundred demos.
Keep a human override. Trust accelerates when people can take the wheel — and see why the system acted.

Practical takeaway

This week, pick one operational loop where staff are still the integration layer between tools. Write down the cost of delay, the error rate, and who owns the outcome. Only then decide whether the fix is better integration, clearer visibility, or genuine automation.

Clarity Growth helps ambitious UK businesses turn technology choices into connected operations — strategy, software and automation that remove friction instead of adding dashboards.

---

Sources

Make technology work better.

Clarity Growth helps ambitious UK businesses turn technology and operations goals into practical, measurable outcomes.

Start a conversation

From kiln to control room: how London's Gigaton is turning industrial AI into real cost and carbon wins