EPSRC Artificial Intelligence Hubs for Real Data and for Scientific and Engineering Research

Funding from the EPSRC to establish up to eight hubs in artificial intelligence (AI) for real data and AI for scientific and engineering research.

The Engineering and Physical Sciences Research Council (EPSRC) has launched this opportunity to establish six research hubs in two distinct areas:

• AI for real data.
• AI for scientific and engineering research.

This major investment will form part of EPSRC’s new strategic delivery plan. The award will support collaborative research programmes that develop new and holistic approaches to tackle working with, and learning from, complex data and datasets.

The hubs will be critical mass investments that are expected to form connections to the wider AI community and research and innovation ecosystem. They will have a core mission and sets of activities and objectives in either AI for real data or AI for scientific and engineering research.

Further details of the two areas supported by this fund are as follows:

AI for real data programme

Through this programme EPSRC is seeking to support collaborative research programmes that develop new and holistic approaches to tackle working with, and learning from, complex data and datasets. Proposals must be multidisciplinary and multi-stakeholder but need not necessarily advance the state of the art in all areas.

Applicants are asked to clearly highlight their chosen area from the following within their proposal:

• Uncertainty quantification.
• Real-time and dynamic data.
• Complex data (federated, heterogeneous, noisy, sparse, and multimodal data).
• Hybrid AI.

While four areas of focus have been specified, it is for applicants to determine the broader areas that the hub will focus on in order to address the above challenges.

Applications are welcomed from all disciplines that are supported by UKRI, including but not limited to:

• Statistics and applied probability.
• Computer science.
• Data science and data engineering.
• Digital humanities.
• Ethics.
• Operational research.
• Applied mathematics.
• Research software engineers.
• Innovation studies.
• Urban and environmental studies.
• Communication studies.

Substantial involvement of end-users such that they co-create the research programme from the beginning is essential to enable the programme to be grounded in a real-world context, and to promote knowledge transfer and expertise.

The majority of the work undertaken should fall within EPSRC’s remit. The end-use demonstrated in the proposal does not necessarily need to be within EPSRC remit, however, the underlying science and research should be, and the results should be generalisable to some degree.

AI for scientific and engineering research programme

Through this programme EPSRC seeks to support collaborative research programmes that enable the development and adoption of new AI capabilities across research domains, driving leadership in focused discipline areas, and, when appropriate, facilitating knowledge exchange of applied AI across disciplinary boundaries. In particular EPSRC is looking to fund hubs which will use the power of AI to transform research across the physical sciences and engineering.

The proposed activity will invest in new interdisciplinary approaches which will bring the power of AI to bear on world leading research. To support this, each AI for scientific and engineering research hub will consist of a consortium of academic and industry organisations and will be expected to work collaboratively with a range of stakeholders across the UK’s research and innovation landscape. Co-creation between the AI community and a science or engineering discipline is essential and should be evidenced in the work packages, leadership team and governance of the hub.

Applicants are asked to indicate in their application how they will address both development of novel AI and contribute to an area of EPSRC funded science or engineering (eg, AI for chemistry, AI for engineering, AI for materials science).

Proposals are expected to be multidisciplinary but need not necessarily advance the state of the art in all areas. Applications which involve large experimental science facilities must demonstrate that they are enabling the development of AI research within universities through their planned programme in order to be eligible to be part of applications.

Proposals are welcomed from researchers in all science and engineering disciplines that are supported by EPSRC, including but not limited to:

• Data science.
• Data engineering.
• Computer science.
• Research software engineers.
• Physical sciences (chemistry, physics, materials sciences).
• Engineering (including manufacturing).

Funding body	Engineering and Physical Sciences Research Council (EPSRC)
Maximum value	£12,000,000
Reference ID	S24544
Category	Science and Technology Engineering and Physical Sciences
Fund or call	Fund