Government taps AI and machine learning to analyse Test and Trace data
Joint Biosecurity Centre to work with Turing Institute and Royal Statistical Society
Credit: Andrew Milligan/PA Wire/PA Images
The Joint Biosecurity Centre has enlisted The Alan Turing Institute and Royal Statistical Society to lend their statistical-modelling and machine-learning expertise to help predict coronavirus outbreaks and inform the response.
Under a partnership announced last week with the JBC – set up by the government in May to support coronavirus decision-making – the RSS and The Alan Turing Institute, the UK’s national institute for data science and artificial intelligence, will set up a virtual statistical-modelling and machine-learning laboratory to provide independent analysis of NHS Test and Trace data.
The lab will give the JBC a “deeper understanding of how the virus is spreading across the country and the epidemiological consequences”, the Department of Health and Social Care said in an announcement.
“Statistical modelling helps data scientists to predict what the virus might do next, based on what is understood about it already,” it added.
The JBC, which advises on Covid-19 alert levels to inform the pandemic response, has been working with Public Health England to provide real-time analysis about infection outbreaks to national and local government bodies.
The new joint lab will help to identify areas where coronavirus is spreading, to allow the government to respond to outbreaks more quickly.
It will also carry out more in-depth analysis of the factors that have the biggest effect on the number of cases, to support better forecasting of how the virus could spread around hotspots and inform the response.
And it will also examine the effects of interventions to control the virus. “This will build on our understanding of the effect and impact of our responses so far and in the future,” DHSC said.
The JBC has been working with a number of external bodies to provide analysis. Earlier this month, contract documents showed it would use mobile phone tracking data analysed by a telecoms company to determine how many people were visiting shopping hotspots and attractions, and to what extent they are maintaining social distancing, in order to inform its thinking.
Clare Gardiner, director general of the JBC, said the centre has been working with external academic and scientific experts “to ensure our work is underpinned by the best thinking and innovation”.
“Throughout, we’ve sought regular guidance from experts across the UK, including The Alan Turing Institute and Royal Statistical Society, for our continual work on mathematical and statistical modelling to inform the insights we give,” she said.
“By formalising this partnership, we will better support NHS Test and Trace in breaking the chains of transmission of the virus, bolstering our readiness for this winter and beyond.”
Health innovation minister Lord Bethell said the partnership would “give local and national decision makers access to the very best scientific analysis”.
“This information will also be at the fingertips of the wider public,” he said. The lab will openly publish its research, including its statistical methodologies, as research papers, blog posts, and open-source computer code.
Chris Holmes, The Alan Turing Institute’s programme director of health and medical science, and professor of biostatistics at Oxford University, said the partnership would “strengthen knowledge across the wider public health ecosystem”.
“Through independent, open science, rigorous modelling and analysis we will provide further understanding of this issue to the public and wider scientific community.”
UKHSA undertakes workshops and consultancy exercises
Contract floated for provider to assist with existing services and possible new features including real-time data feeds
National statistician says research showing increased faith in government-backed research is ‘encouraging’
Software platform will support collation and analysis of data on humans and the natural world