King's robotics and AI specialist helps government upgrade public services

Dr Shan Luo, Reader in Robotics and AI in the Department of Engineering at King's College London, has been selected to spend a year working with UK government teams to build new artificial intelligence tools that improve public services and support national security.

Opened by the Minister for Digital Government and Data, the Open Source AI Fellowship will seek to modernise critical public services and systems across the UK. Delivered by the Alan Turing Institute and funded by a $1 million investment from Meta, Dr Luo will join a cohort of UK AI experts embedded across government departments to create open-source AI for the public good, including models that analyse images and video to help local authorities prioritise infrastructure repairs more effectively.

The fellowship is a key element of the government's wider digital transformation agenda, which aims to bring world-class technical talent into Whitehall and accelerate the safe use of AI in areas ranging from transport to public safety and defence. By focusing on tools that can run securely, including offline or highly sensitive systems, the programme is designed to help public servants make faster, better-informed decisions while providing robust data protection guarantees.

By embracing open-source technology, the fellowship mirrors academic frameworks built on collaboration across communities and furthers the take up of technologies in local authorities.

Reflecting on his fellowship, Dr Luo said: "My Open Source AI Fellowship at the Alan Turing Institute offers a rare opportunity to align advanced AI research with real civil service challenges, while ensuring transparency, reproducibility, and public benefit through open-source development."

This latest appointment follows Dr Luo's recent Engineering and Physical Sciences Research Council (EPSRC) Open Fellowship, awarded in December 2025, which funds his TacDiff project to develop a new generation of tactile robots with a human-like sense of touch.

TacDiff aims to use differentiable simulations and data-driven robot designs to close the gap between laboratory performance and real-world deployment, enabling robots that can safely handle delicate objects and adapt to complex environments in sectors such as manufacturing and healthcare.