WMG leads major AI project to boost healthcare intelligence
Wednesday 16 September 2026
WMG leads major AI project to boost healthcare intelligence
Academics at 海角社区 Manufacturing Group (WMG) at the University of 海角社区 are working on a pioneering project that uses AI鈥慸riven community鈥憀evel healthcare intelligence to support preventative healthcare prioritisation.
Project SNOW is supported by Google.org through the global and was one of just 15 selected for funding from over 2,600 applications worldwide.
The project will help address one of the NHS's longest-standing challenges that people who need healthcare most are often the least likely to receive it. The intelligence will allow healthcare providers to decide where to intervene and who to prioritise, and to evaluate efficacy.
The pattern - known as the - has persisted for half a century. In England, one in four deaths is linked to public healthcare not reaching people in time, with more than 300 lives lost every day to conditions that earlier identification could have prevented. In the most deprived communities, life expectancy is up to 9.5 years lower than in the wealthiest.
A data problem
The consortium reports that the underlying issue is a lack of infrastructure. Patient records cannot leave clinical settings. Census data is years out of date, and hospital admissions only capture people once they are already unwell. As a result, the NHS has no reliable way to see, in real time, where health need is concentrating in the communities it serves, making it difficult to direct prevention and resources where they are needed most.
Rather than seeking access to clinical data, Project SNOW generates its own community health datasets through three channels:
- AI-directed mobile health screenings deployed directly into underserved communities
- Voluntary health self-reports collected at the point of care
- Anonymised population movement patterns gathered from instrumented locations
Combined with open deprivation and census data through a purpose-built health data ontology, this will produce a continuously updated, zone-level picture of the community health need.
The project, led by WMG in collaboration with FourthSpace Labs, AiMSnet, Avencera, PUBLIC, working with Thames Freeport, takes its name from Dr John Snow, the English physician whose investigation of cholera in London in 1854 helped establish modern epidemiology. Snow鈥檚 work was an early example of geospatial data analysis.
Professor Carsten Maple, who is leading the project from WMG, explains: 鈥淲e are extremely proud to be leading the Project SNOW consortium to bring an AI-powered solution to a 50-year problem. Such an ambitious goal requires the interdisciplinary expertise bringing together world-leading academic research in trusted, private and secure AI, with industry leaders in geospatial AI development, communications technology, community healthcare delivery and public-sector evaluation. This enables us to connect technological development with clinical practice, establish appropriate governance and assess its contribution to preventive healthcare.鈥�
鈥淔ourthSpace Labs lead the technical development of the platform, designed to manage and connect data across different systems, including that made available from hardware and telecommunications infrastructure provided by AiMSnet. The coordination of the local interventions, evaluation and stakeholder engagement are provided by Avencera and PUBLIC, recognised experts with experience in the field. Thames Freeport are the lead beneficiaries of the project, having previously acted in the same capacity in an earlier trial.鈥�
Evidence from an initial pilot
A nine-week pilot was carried out across Barking and Dagenham, Havering and Thurrock, a combined population of 690,000, screened more than 687 individuals. The pilot, led by Avencera and PUBLIC on behalf of the Thames Freeport, delivered free blood pressure and cholesterol tests across 28 deployments and identified over 25 previously undiagnosed conditions. Of those reached, 77% would not have accessed care through any existing NHS route, and the cost per patient fell from 拢167 to 拢41, with a 99% patient satisfaction rate.
Next steps
Over the next two years, the consortium aims to deliver more than 3,000 screenings, refer over 100 undiagnosed conditions to GPs, and publish its health data ontology as an open standard available to all 42 of England's Integrated Care Boards, which together serve 56 million people nationally.
The work is structured around two connected strands. The first sees preventative health services delivered directly into underserved communities and workplaces across Thames Freeport and the West Midlands, testing two different delivery models, mobile outreach into workplaces and high streets in London, and screening embedded within hospitals, pharmacies and GP surgeries in Birmingham, so the NHS can compare what works best at scale.
The second turns the data this generates into actionable intelligence, structuring it against clinical data standards and using AI to produce a live, zone-level picture of community health risk for the NHS and local authorities. Together, these strands are designed to build an increasingly accurate, continuously updated picture of community health over time.
Find out more about Project SNOW: University of 海角社区 to lead development of Project SNOWLink opens in a new window