Britain’s power networks are too rigid to handle millions of solar panels, heat pumps, electric vehicles, and sudden supply shocks—and the old mathematical methods for running them simply cannot cope with that scale of complexity. This project aims to embed artificial intelligence directly into the operation and long-term planning of the electricity grid. The core problem is that existing AI tools assume the future will resemble the past, but a rapidly decarbonising energy system will look nothing like today’s. The researchers will develop a new generation of AI that can handle incomplete data, vast numbers of interacting devices, and deep uncertainty—allowing network operators to make fast, confident decisions about both day-to-day balancing and multi-year investments. If successful, the work could transform how the UK manages its electricity infrastructure. Instead of preconfigured, brittle systems that risk price spikes or blackouts (as seen recently in Spain and at Heathrow), the grid would become agile and self-adapting. That would mean cheaper, more reliable power for households and businesses, and a faster, safer path to net zero. The project also builds shared data infrastructure and interdisciplinary collaborations, laying the groundwork for the UK to lead globally in agile energy networks.
View original technical description
Today’s power networks are largely preconfigured and rigid, limiting their ability to adapt to a large influx of renewables, low carbon technologies, and emerging threats. Recent events, from price hikes to supply outages in Spain and Heathrow show such rigid systems carry high costs and risks. What is required are efficient and agile networks that allow millions of intelligent devices and subsystems to work together to deliver energy affordability, resilience and net zero. However, the existing analytical approach to network operations and investment seeks optimal solutions and assumes datasets are both perfect and widely available. This approach is not scalable to millions of new devices with incomplete data, nor the growing complexity and uncertainty. The aim of this research is to introduce artificial intelligence technologies (AITs) at scale to support network operators to develop two crucial capabilities, enhancing network agility and flexibility to achieve high speed and accuracy in their decision making to: cope with a vast combinations of possible network links, mass flexible devices and their dynamic interactions; allow energy network operators to make intelligent decisions, both for short-term operations and long-term development, with agility and confidence. Whilst AITs offer promising potential for energy system challenges, much machine learning technology is predicated on the future looking like the past. For the rapid decarbonising energy system, the future will be very different. A simple injection of AITs will not meet the needs of a highly dynamic and rapidly changing energy sector. Key AI challenges in this project are both to drive wide adaption of AI and to develop new generation of AIT developments fit for a rapidly changing system. The objectives of this project are: 1. To lay the foundation to build the critical interdisciplinary collaborations to drive cutting edge research at the intersection of energy networks and AI, to drive wider adoption of AI capabilities and identify fundamental development in novel AITs that are required to address rising complexity, scalability and uncertainty in network operation and investment. 2. To develop research infrastructure to speed up the embedding of AI as a research tool in a fair, collaborative and inclusive way, complementing UKRI’s ongoing work in cutting-edge energy network research. 3. To develop a data sharing infrastructure for increasing the use, access, and awareness of data sets for use with AITs, that come with detailed provenance records to support frontier AI progress and greater use of AI in a broad spectrum of energy network research. 4. To drive the application for agile network operation and investment across local, regional and national scales and over operation and investment time frames. This project will position the UK to be the world leaders in agile network operation and development through bringing together AI Community, laying the foundation for the community to come together to build our strong foundation, transforming energy networks’ efficiency whilst substantially enhancing its ability to withstand future risks and shocks.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know