University of Glasgow Engineers Offer Support to Manufacturing AI Research Hub
Engineers from the University of Glasgow are lending their support to a new research hub which aims to bring the benefits of AI to the UK electronics industry.
The AI Hub for Productive Research and Innovation in Electronics, or APRIL, is one of nine new artificial intelligence research hubs across the country that will deliver next-generation innovations and technologies.
The new hubs are supported by £80m in new funding from the Engineering and Physical Sciences Research Council (EPSRC), part of UK Research and Innovation (UKRI).
APRIL will develop AI tools to accelerate the development of key components such as new semiconductor materials, integrated circuits, complex microchip designs and system architectures – leading to faster, cheaper, greener and overall more power-efficient electronics.
APRIL will be led by Edinburgh’s Regius Chair of Engineering, Professor Themis Prodromakis. Professor Vihar Georgiev will lead the University of Glasgow’s contribution to APRIL and he will be supported by Professor Hadi Heidari and Professor Bo Liu, all of them from the James Watt School of Engineering.
The University of Glasgow team will contribute to APRIL’s simulation and modelling and AI-driven IC design work package, helping to develop sophisticated ‘digital twins’ capable of simulating the performance of key semiconductor production systems under a range of different conditions as well as realising effective and efficient IC building block design automation. These models will leverage data gathered through extensive automated testing routines and will inform new device design and fabrication methods. The AI-driven analog/RF IC design techniques will obtain considerably higher performance and robust designs compared to human experts using a much shorter design time.
By leveraging advanced AI in this integrated framework, APRIL aims to greatly boost the rate at which new semiconductor technologies can mature to the point they are practical for adoption in real-world designs.
APRIL will be structured with five main pillars spanning the semiconductor innovation chain: materials discovery, device design, system and circuit design, and testing and verification, and modelling. Professor Georgiev will lead the modelling work, which will provide a link between all the work packages.
Cross-cutting these are five key capabilities that will enable the delivery of the pillar tasks: data collection, AI model selection, AI model training, system integration, and data security.
The matrix structure formed by the intersection of the innovation chain and capabilities will help promote the efficient division of labour and create opportunities for multidisciplinary collaboration.
Professor Vihar Georgiev, Professor of Nanoelectronics and leader of DeepNano Group at the University of Glasgow, said, “This hub is a very exciting opportunity to nurture new, cutting-edge breakthroughs for the electronics industry not only in the UK but also globally. I am pleased to be part of the APRIL research hub and lead the modelling activities, which have the potential to boost the UK’s productivity and innovation in an increasingly valuable sector of industry.
“The University of Glasgow’s contribution to the project builds on decades of experience in semiconductor research and design at our specialised facilities, including the Glasgow Conputaional Engineering Centre (GCEC).
“I am looking forward to working with my colleagues on this exciting project to harness the power of AI to benefit the semiconductor sector.”
Researchers from the University of Glasgow are also involved in supporting the University of Bristol-led e AI for Collective Intelligence (AI4CI) hub, which received support in the same round of funding.
Glasgow researchers are also leading the “iREAL: Inclusive Requirements Elicitation for AI in Libraries to Support Respectful Management of Indigenous Knowledges” project, funded by the Arts and Humanities Research Council (AHRC), announced alongside the funding for the hubs.