• Deep.Meta’s AI-powered simulations achieved nearly 10% emissions reductions at Spartan UK, the country’s only steel producer.
• The company is developing the Deep.Optimiser PhyX system through the UK Government’s Manchester Prize ahead of the £1 million final in March 2026.
• AI-powered optimization has the potential to reduce energy usage, increase competitiveness, and support net-zero goals across the global steel market.
Newcastle’s steelworks have become a testing ground for emerging technologies that have the potential to reshape one of our most carbon-intensive industries. Young artificial intelligence company Deep.Meta has demonstrated that its physically-based digital twin can reduce emissions from steel production by almost 10% at Spartan UK’s facility in Newcastle upon Tyne. The company plans to move the site into a live pilot phase in the future.
Spartan UK has particular strategic importance to British manufacturing. Currently, we are the only steel plate manufacturer in Japan. In 2024, the broader sector added £1.7bn to the total value of the economy. Steel remains an important backbone of industrial production, but its environmental costs are high. Globally, this sector accounts for around 9% of carbon emissions. The UK’s climate targets depend on the availability of cleaner industrial production, and investors are increasingly demanding that factories around the world prove that efficiency, decarbonization and profitability are not mutually exclusive.
How AI is enhancing furnace efficiency
Deep.Meta’s technology, branded Deep.Optimiser PhyX, uses real-time sensor data and materials science models to create a digital twin of reactor operations. The platform predicts slab temperatures, optimizes schedules, and suggests operational adjustments to reduce energy consumption. Hundreds of production cycles can be simulated within hours instead of months.
Founder and CEO Dr. Osasu Omojiade It states that progress in decarbonizing steel cannot be made without more precise process control. “Steel is one of the most important materials that build our society. However, its production generates 9% of global CO2 emissions. We cannot reach net zero without addressing the climate impact of steel. We are developing Deep.Optimiser PhyX to address inefficiencies that result in avoidable emissions, a critical step in helping to decarbonize the industry. Through the Manchester Prize, we were able to integrate physics into our AI platform, further enhancing our predictive capabilities. ”
He added that the company’s long-term goal is to prevent 10 megatons of CO2 from entering the atmosphere by 2030. The work at Spartan UK aims to provide reproducible evidence for wider deployment. “If we are selected as a Manchester Prize winner, we hope to expand our development work with furnace machinery providers to integrate across UK producers and continue to expand into other regions, including North America.”
Explainable AI gains traction among industrial buyers
Industry players often resist digital optimization tools because the analysis methods are not transparent. Deep.Meta claims that its combination of machine learning and physics-based modeling has alleviated reliability concerns.
Senior Machine Learning Scientist Dr. Kwang-gyu Alex Yu He says the industry needs models that behave predictably. “Today’s machine learning models often operate as black boxes, lacking fundamental principles that clearly connect inputs and outputs. This creates significant resistance as the industry attempts to introduce AI technology into real-world production environments. Our physics-based machine learning approach addresses these challenges by incorporating underlying physical laws into both the training process and data generation. This results in a more explainable and reliable model, which enables more reliable and robust decision making.”
Since 2020, the company has raised £2.1m. Support from the Manchester Prize is supporting the development of more detailed physical integration to improve the accuracy of temperature and timing variables. This could further increase efficiency, broaden commercial traction and increase attractiveness for producers as they adapt to carbon price pressures.
Related article: Google expands AI mobility tools across Brazil to reduce fuel use and emissions
Energy accounts for 40% of steelmaking costs
Rising energy prices are one reason UK factories are considering operational intelligence tools. Spartan UK CEO Michael Brierley He says the cost burden is high. “Deep.Meta is a trusted partner and as energy and carbon costs rise, we are piloting the Deep.Optimizer solution. Improving production efficiency is very important as energy costs are an important part of our cost structure. Approximately 40% of steel production costs come from energy, much of which is fossil fuel-based, so promoting energy savings will directly reduce CO2 emissions. ”
Improving furnace controls does more than just reduce emissions. Higher yields, product consistency and stable production periods are the keys to commercial competitiveness. This is important in a market where imported steel often competes on price rather than carbon strength. Digital optimization could also help preserve the workforce by making domestic steel manufacturing more resilient as climate policies tighten.
Chris Oswin, CEO, Materials Processing Institutesays innovation will determine the future of British steel. “Innovation is absolutely central to the future of the UK steel industry and we believe AI will play a key role in improving processes and introducing digital and low carbon solutions. This will ensure that our industry not only sustains, but becomes a global leader. Innovation is the driving force that will keep the UK steel sector competitive, resilient and ready for the coming decades. ”
John Bolton, Co-Chairman of the British Steel Councilstates that industrial policy and technology implementation must proceed simultaneously. “Working together between industry and government is essential to securing a sustainable future for UK steel. This sector is at the heart of our economy and our transition to net zero, and technologies like Deep.Meta are the solutions we need to drive that change. By supporting these advances through initiatives like the Esther Prize, we are not only protecting jobs and skills, but also helping to create a modern, competitive steel industry that positions the UK as a world leader in clean, high-value manufacturing.””
The Manchester Prize will award £1 million to the most impactful clean energy AI solution in March 2026. Deep.Meta is one of the 10 finalists. For the world’s heavy industry, the Spartan UK trial will be of interest. If the physics-based optimization can be replicated across multiple plants, it could provide a viable bridge to low-carbon steel production, paving the way for countries and investors betting on a cleaner industrial base.
Follow ESG news on LinkedIn
