United Kingdom’s Innovation Clusters | Where are they and how do they thrive?
Visualising the distribution and strength of UK’s innovation clusters
Introduction
Both the Labour and Conservative governments have stated their desire for the UK to become a global hub for innovation by the year 2035. Former Prime Minister Rishi Sunak envisioned the UK as a “scientific and technological superpower”. However, examining how this goal is being operationalised by the government is crucial to evaluating policy effectiveness.
This article explores how innovation is being funded in the UK. It specifically examines how funding for highly prioritised sectors such as AI and Data is distributed around the country. A better understanding of the distribution of funding helps inform policy decisions to better develop centres of excellence for different sectors. For example, the following visualisation shows how Manchester is a hub for AI and Data, whereas Oxford is a hub for biotechnology. We undertook cluster analysis based on projects funded by the government to understand the dispersion of innovative activities across the UK.
Visualisation
Cluster analysis is a method of grouping together data points that exhibit similar characteristics. In our case, it means that spatially co-located projects within the same sector form a cluster.
In alignment with the government priority areas, the study considers four major sectors: AI and Data, Clean Growth and Energy, Aging, Health and Society, and Advanced Materials and Manufacturing. The visualisation below shows the location and strength of innovation hotspots. It also provides project-level information on title, funding and sector.
Source: Department for Science, Innovation and Technology (January 2023)
Analysis
Finding 1: Innovation funding in London remains highly dense and diverse across sectors compared to other regions
In our study, projects within London form tight-knit groups across all sectors. Of the total £141 million spent on innovative projects between 2016 and 2023, City of London-based projects accounted for £78 million, while the remaining £63 million was distributed across other regions. In addition, the relative advantage of the London ecosystem is indicated by how labour productivity is 26.2% above the UK average (ONS 2024)
Finding 2: There is a prevalence of dense clusters around universities
The UK is renowned for its research-intensive academic institutions. In 2021–2022, research and innovation at Russell Group universities alone generated over 250,000 jobs and £37.6 billion in revenue. Through extensive professional and academic networks, specialised technology development facilities, and incubators, UK universities boost innovation.
In line with this, the map shows dense clusters around the University of Cambridge, Oxford, LSE and UCL. Interestingly, universities also exhibit specialisations in certain sectors. For example, surrounding the University of Bath and the University of Manchester are clusters of AI-related activities. Health and life science-related work is relatively denser at the University of Oxford, University of Cambridge and Cardiff University.
Finding 3: Catapult centres are diversified across the UK
Visualisation adjusted for scale of funding
Catapult Centres are non-for-profit Innovate UK-supported hubs that provide research facilities, technical expertise and business support. The Catapults are uniquely/strongly equipped to bridge the late stages of innovation — where challenges such as scaling, compliance, and commercialisation arise.
At present, there are nine Catapult Centres across 56 locations in the UK. The visualisation shows that much of Catapult funding is directed outside of traditionally high-growth areas outside the Golden Triangle (London, Cambridge and Oxford) showcasing an intentional investment in regional innovation.
Catapult Centres have a unique funding structure consisting of three types: core grants (Innovate UK), collaborative R&D (capped at 30% of total project costs as the remaining 70% must come from other sources), and commercial revenue. Hence, they are able to bridge the gap between research and commercialisation, especially for high-risk projects, unlike traditional funding options such as direct grants and loans.
However, there is a need to improve the match between private and public funding. The House of Lords Science and Technology Committee report (2022) suggests that unless Catapults are scaled up further by the government, they risk not receiving sufficient private funding that aligns with government targets.
Conclusion
The distribution of innovation funding across the UK reveals three key findings. Notably, London dominates the public funding landscape, hosting diverse innovation clusters that span multiple sectors and technologies. This concentration of funding aligns with our second finding - there is a strong correlation between university locations and specialised innovation clusters. The relationship suggests the natural emergence of knowledge-based niches, where academic expertise drives specialised innovation. The strategic placement of Catapult Centres aims to foster innovation hotspots evenly across the country. These specialised technology and research centres have received significant public investment, reflecting a strong effort to diversify the nation’s innovation ecosystem.
As the UK works towards its goal of becoming a global innovation hub, sustained and targeted funding for innovation projects across all stages will be essential. Continued efforts to develop and expand innovation clusters, along with leveraging the knowledge and resources of universities, will be crucial in increasing innovation capacity.
Alongside these key findings, the interactive visualisation enables interested parties to explore their area and examine specific projects that have been funded by UKRI.
Appendix on Methodology
In the data from the UKRI, only projects from 2016 onwards are considered to reduce firm-level noise. Each data point has geographic coordinates derived using postcodes and a sector code that allows for spatial and attribute-based clustering.
The Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN) algorithm is used to identify spatial innovation clusters. This classifies regions based on the density of firm locations. Areas with high concentrations and large coverage of firms are grouped into clusters using the Excess of Mass method, while less populated regions are marked as 'noise', ensuring that only densely co-located firms form clusters. Clusters are then combined with their corresponding sectors, which are identified with different shades of purple.
The sector clusters are based on two criteria,
The projects must be spatially co-located
The projects must be in the same sector.
A final condition is that the minimum size is set at 50 firms to prevent over-clustering and ensure that identified clusters represent significant masses of innovation activity.