New funding to develop AI models that learn individual health signatures from wearable sensor data

Smart Data Research UK, part of UKRI, is funding 15 new Fellowships which address some of the UK’s most pressing challenges, from combating AI-generated misinformation to forecasting demand for electric vehicle charging infrastructure. The Fellowship projects, each worth up to £200,000, cover SDR UK’s four research themes: productivity and prosperity, health and wellbeing, digital society, and sustainability. All projects will start in February 2026 and run for 18 months.

Dr Xiao Gu, Senior Research Associate in the Computational Health Informatics (CHI) Lab at Oxford, has been awarded one of the Fellowships. He will develop AI models that learn individual health signatures from wearable sensor data to detect early signs of clinical decline. Many of us now wear a smartwatch that tracks our daily steps. But what if that same device could quietly and reliably spot the earliest signs of a health problem, long before a crisis sends someone to the hospital?

Using foundation models trained on large wearable datasets, the project will build an AI-driven “Digital Triage Nurse” that learns each person’s health signature and flags early signs of decline, supporting earlier intervention and more personalised care. Like ChatGPT learns patterns from large text data, this model will learn patterns from wearable signals and can be adapted to new health questions with relatively little extra data - especially important for people with multiple long-term conditions.

Dr Gu says, “I’m honoured to join the Smart Data Research UK Fellow cohort. This fellowship will enable me to turn large-scale wearable data into ‘smart’ data that supports ‘smart’ proactive care - from care homes to hospitals, and from the UK to LMIC [Low- and Middle-Income Countries] settings.”

He adds, “The UK’s health services urgently need such proactive tools, as our ageing population means more people are living with multiple long-term conditions, and too often problems are picked up late and dealt with in emergency. ” Dr Gu concludes, “This work is only possible thanks to the generosity, trust, and support of my academic mentors and clinical collaborators, particularly the Computational Health Informatics (CHI) Lab at Oxford, led by Professor David Clifton.”