A University of Surrey spinout that uses artificial intelligence to optimise biogas plant operations has won the UK government's flagship AI innovation challenge, taking home a £1 million award.
BiofuelAi, co-founded by Professor Michael Short, has developed an AI-powered decision support platform that gives plant operators a real-time picture of digester activity and predictive guidance on the actions most likely to improve output. Pilot trials have demonstrated revenue increases of 6–10% and profit improvements of 7–13% per site, alongside a 28% reduction in carbon emissions.
The Manchester Prize, awarded annually by the Department for Science, Innovation and Technology for UK-led AI breakthroughs in the public good, was delivered by Challenge Works — part of UK innovation foundation Nesta. Science Minister Lord Vallance said the technology "could supercharge our mission to power Britain with clean, affordable energy."
The platform creates a digital twin of a biogas plant by combining mechanistic models, machine learning and hybrid approaches, enabling simultaneous optimisation of short-term decisions — such as feeding recipes and storage management — and longer-term considerations including feedstock acquisition and digester health.
Professor Short, CTO and co-founder, described anaerobic digestion as "more like brewing than chemistry," noting that what goes into a digester can take days or weeks to show up in outputs — making reliable prediction genuinely difficult. "We spent years developing models that could change that," he said. "The Manchester Prize win matters because it says the science is ready to become a product."
CEO Alan Beesley said the biogas sector remained one of the least data-driven in energy, with many plants still managed through spreadsheets and operator experience. "BiofuelAi changes that," he said. "Winning the Manchester Prize validates the work of an exceptional team and accelerates our mission to make green energy more affordable, more consistent and more accessible."
The company, a spinout from Surrey's AI4AD research project, has attracted more than £1.5m in research funding and is currently onboarding three new sites. It has also signed a UK reseller agreement. Over five years, BiofuelAi projects its platform could deliver more than £500m in client value, and by 2030 estimates it could mitigate 293,000 tonnes of CO2 equivalent per year across the UK — equivalent to heating 133,000 homes.
BiofuelAi is based at the Surrey Technology Centre in Guildford and is actively seeking approaches from plant operators, industry partners and investors.
Biogas AI platform wins £1m prize for greener gas production
