Imperial Researcher Secures £1.8 Million Fellowship for Trustworthy AI in Manufacturing

Dr Nan Li will develop AI systems to enhance decision-making for engineers in complex component manufacturing, focusing on safety-critical applications.

Generic image of an abstract representation of artificial intelligence with glowing lines and a blurred industrial background.
AI

Generic image of an abstract representation of artificial intelligence with glowing lines and a blurred industrial background.

An Imperial College London researcher has been awarded a prestigious £1.8 million fellowship to develop trustworthy artificial intelligence systems for the manufacturing sector.

Dr Nan Li, Reader in Lightweight Design and Manufacturing at the Dyson School of Design Engineering, has received the funding from UK Research and Innovation (UKRI) for a three-year project focused on creating AI that manufacturers can trust. The initiative aims to assist engineers in making faster and more reliable decisions during the design and production of complex components, particularly those made from sheet materials, such as car body panels and aircraft structures.
This fellowship is part of UKRI's £39 million investment to bolster the UK's AI capabilities by supporting AI applications for significant scientific and industrial challenges and attracting top global talent. Dr Li is one of 20 researchers to receive funding through this programme. Her work builds on 17 years of research in sheet-material forming and six years of integrating AI into the field, with her group having developed validated deep-learning and graph-based simulators.
The project is backed by 18 partners from manufacturing, materials supply, software, AI, automotive, and aerospace sectors, who have committed approximately £1.2 million in in-kind support. This collaboration ensures the research directly addresses real industry needs. Manufacturing is vital to the UK's automotive, aerospace, and energy sectors, facing pressure to produce lighter, lower-carbon products safely and reliably. The increasing variability of greener materials makes consistent manufacturing harder to predict.
Dr Li explained that while AI can accelerate design decisions, engineers need to trust it, especially when conditions change. The AIMForming system aims to provide trustworthy predictions with known limits, indicating where its evidence is reliable and where engineers should exercise caution. This approach contrasts with current methods that involve costly and time-consuming simulations and physical testing, often leading to more conservative designs.
AIMForming will integrate component geometry, material behaviour, and manufacturing processes to make manufacturability a design-time consideration. A key challenge it addresses is knowing when an AI model operates outside its training conditions. The advisory system will help engineers act with greater confidence by highlighting when decisions are supported by evidence and when further analysis is needed, potentially reducing development cycles and scrap.
The research also aims to support the broader adoption of new materials and manufacturing processes, benefiting SMEs and contributing to the responsible use of AI in advanced manufacturing. Dr Li will establish an open benchmark and a UK-wide community for AI in forming to facilitate testing and development.
Based on information from the official source: Imperial College London (16/11/2026)