10 AI Startup Ideas Tackling Energy and Climate Challenges
A new wave of British artificial intelligence (AI) innovation is targeting one of the UK’s biggest challenges—cutting energy costs while moving towards a clean, net-zero future. Backed by the UK government through the Manchester Prize, ten pioneering teams have been selected to compete for a £1 million award by developing AI solutions that lower emissions, boost efficiency, and reduce household energy bills.
Each team has received £100,000 in seed funding and access to high-performance computing to scale their ideas. The prize supports the government’s wider Plan for Change, a strategy to drive economic growth and accelerate clean energy adoption.
The Power of AI in the Energy Transition
From forecasting demand to designing better energy systems, AI is quickly becoming a vital tool in the transition to net zero. In homes, AI can help identify heat loss, optimize appliance use, and improve the integration of low-carbon heating technologies. In industry, it enables smarter processes and lower emissions. And across the grid, AI helps balance electricity supply and demand in real time—crucial as wind and solar take a growing share of the mix.
The UK’s science and energy leaders, including Peter Kyle and Ed Miliband, praised the ten finalists for demonstrating how AI can translate climate ambition into real-world solutions—from insulating homes to greening logistics and cleaning up heavy industry.
The Ten Breakthrough Projects
Here’s a look at the ten AI innovations now in the running for the Manchester Prize:
Agent Net Zero (University of Sheffield & AMRC)
A real-time AI sustainability assistant for factories, highlighting carbon hotspots and suggesting energy-saving changes to help businesses cut emissions without sacrificing productivity.BiofuelAi (University of Surrey)
An AI platform that builds a digital twin of biogas plants to optimize production using unpredictable feedstocks—making biofuels cleaner, more consistent, and cost-effective.Carbon Re (UCL & University of Cambridge)
AI software that reduces emissions from cement by optimising kiln operations in real time—cutting fuel use in one of the world’s most polluting industries, with potential to scale to steel and glass.Cavolo (Kale AI)
A smart route-planning tool that helps delivery fleets shift to light electric vehicles, cutting emissions and traffic in congested cities while ensuring goods arrive on time.Deep.Optimiser-PhyX (Deep.Meta)
An AI-powered digital twin for steel production that uses real-time data and physics-informed models to improve furnace efficiency and reduce carbon from heavy industry.DRIVE (Flexible Power Systems)
Advanced AI that manages EV fleet charging schedules in real time, helping operators switch to electric while easing strain on the grid—technology that could scale beyond transport.EnergyWall (Underheat & University of Salford)
AI-customized insulation panels that turn building walls into smart heating or cooling surfaces—ideal for social housing and homes needing low-disruption retrofits.Green Loops (University of Wolverhampton & ABCircular)
Uses machine learning and advanced materials to design high-efficiency solar panels from recycled PV waste—cutting costs, waste, and emissions in one go.Grid Stability Monitor (University of Manchester)
A fast, AI-based system to assess and predict grid stability as more renewables come online, replacing slower simulations and helping maintain reliable power.RaThPAs (Kestrix)
Aerial thermal imaging combined with AI to map heat loss at neighborhood scale, offering Google Maps-style insight into insulation needs and speeding up retrofits.
The Road Ahead
The ten projects will compete for the £1 million grand prize to be awarded in early 2026. To win, teams must show a path to real-world deployment by 2030, delivering measurable energy and climate benefits at scale. From cutting steel industry emissions to designing AI-powered energy walls, these UK innovations show the power of data-driven thinking in building a cleaner, smarter energy future.