Nine new research hubs located throughout the UK are set to unveil groundbreaking AI technologies as part of a co-ordinated investment by UKRI and Government.

These hubs, with investments totalling £80 million from the Engineering and Physical Sciences Research Council (EPSRC), a component of UK Research and Innovation (UKRI), will be responsible in driving the evolution of AI within the UK to address intricate challenges across various domains, ranging from healthcare treatments to the development of power-efficient electronics. This substantial investment is poised to propel the UK to the forefront of advanced AI research with Newcastle initially focusing on the Healthcare, Transport and Energy Security sectors.

APRIL: AI for productive research and Innovation in eLectronics (APRIL) Hub

Dr Rishad Shafik, Reader in Electronic Systems, will join the leadership team of the National AI Hub APRIL which will generate AI driven tools and methods to revolutionise productivity in electronics. The hub combines expertise in semiconductor technologies, devices, circuits and systems as well in applied data engineering and machine learning algorithms, supported by world-leading academics from Edinburgh, Newcastle, Cambridge, King’s College London, Glasgow, Southampton, Bristol, Sheffield, Bristol, Queen’s University Belfast & Imperial College London.

Overall led by Professor Themis Prodromakis from the University of Edinburgh, the hub will drive innovations through five key research pillars. Dr Shafik will join the leadership team of Pillar 5 which will develop accelerated ‘digital twins’ of low-level devices, circuits and systems using machine learning as a key design automation tool. The aim is to offer a step-changing productivity improvement when designing circuits or systems.

Dr Shafik said: “The electronics and associated semiconductor sectors face major economic opportunities across various application areas such as space, automotive, healthcare, energy, manufacturing, robotics, and the Internet of Things (IoT). Besides these opportunities, accelerated innovations in these sectors are crucial for our national infrastructure. To thrive in this landscape, the UK plc must push the technology boundaries with significantly higher productivity meeting performance, energy efficiency, integration and form-factor needs at a much faster pace. This hub will enable us to address these ambitions and challenges.”

Dr Shafik added: “Today semiconductor microsystems designs go through several iterative and laborious routines. For example, a new machine learning accelerator microchip design using the state-of-the-art CMOS technology could take between 12 to 18 months minimum from conception to fabrication for 3 engineers. If we had access to machine learning models driven by extensive data from technologies and circuits, we can automate the design effectively with many orders of magnitude improvement in productivity.”

Professor Alex Yakovlev FREng, Head of Microsystems Group within the Newcastle School of Engineering and a key investigator of Pillar 5 of this hub, added: ‘’Our group has long-standing reputation of world-class research and development addressing the challenges of today’s complex circuits and systems design and automation. In REF2021, we have demonstrated a 4* impact case study through our industry-grade design automation tool: Workcraft. The tool can automate dependable asynchronous circuits design which is otherwise highly time-consuming. Recently, we have developed international leadership in novel and low-complexity machine learning circuits, systems and applications design inspired by the principles of Tsetlin Machine. The APRIL hub serves as an ideal platform for us to synergise our expertise in one place, facilitating collaboration between design automation and low-complexity machine learning methods.”

Unlocking the potential of AI

Three of the new hubs will address the mathematics and computational research which is foundational to AI, playing a pivotal role in increasing our understanding of new, efficient AI systems. Six of the hubs will explore AI for science, engineering and real-world data which will provide the tools needed to accelerate future AI innovations and advance its application in key areas such as healthcare.

Minister for AI Viscount Camrose, said: “The investment we’re pouring into these new projects is only possible as a result of our pro-innovation approach to AI. The AI Regulation White Paper consultation response we’ve set out today will see us forging ahead with that plan, driving forward the next wave of brilliant AI innovations.

 “These hubs will nurture new, cutting-edge breakthroughs, from healthcare treatments and more power efficient electronics to machine learning and chemical discovery.

 “New projects being delivered by BRAID will also help to define responsible AI in key sectors such as education, policing, and the creative industries, ensuring public trust in the technology as we continue to harness its capabilities.”

Professor Dame Ottoline Leyser, Chief Executive of UKRI, added: “UKRI is supporting researchers and innovators to develop the next generation of AI technologies that will transform our economy and society. The investments announced today will help to deliver the capability the UK needs to realise the opportunities of this transformative technology.

“Through our £1bn portfolio of investments in AI research and innovation, we are supporting the development of new technologies, boosting skills, and accelerating the adoption of trusted and responsible AI.”

The hubs, led by eight universities but working across the whole of the UK, underline the UK’s commitment to maintaining a leadership position in AI research, innovation, and ethical deployment.   

Professor Charlotte Deane, Executive Chair of EPSRC, said: “Artificial intelligence is already transforming our world. EPSRC is supporting world-leading research to unlock its potential and ensure it is developed and used in an ethical and responsible way. Long-term research funding has led to revolutionary advancements to make AI a powerful tool for many applications.

“The hubs will deliver revolutionary AI innovations and tools in a range of sectors from healthcare to energy, smart cities and environment. Only by solving key challenges and improving our understanding of AI will we achieve the increased productivity and economic growth promised by this technology.”  

The investment follows an announcement during autumn 2023 for 12 UKRI AI Centres for Doctoral Training (CDTs). The UKRI AI CDTs will ensure that the UK has the skills needed to seize the potential of the AI era, and to nurture the British tech talent that will push the AI revolution forwards.       

Adapted with thanks from UKRI.

Image: A modern machine learning accelerator layout consisting of intricate low-level interconnects, switching logic and storage devices (courtesy: Mignon AI). New machine learning model driven automation tools could significantly expedite the design process of such an accelerator.