Bringing together technical expertise and collaboration to help researchers turn the potential of AI into practical tools for discovery.
Artificial Intelligence (AI) is opening up new possibilities for scientific research—but applying it in day-to-day work can still be challenging.
A recent N8 CIR AI4Science project, funded by the EPSRC, has been working to address this by supporting researchers to overcome practical barriers to using AI. Through a series of targeted, dRTP-led interventions, technical specialists worked alongside research teams to help deploy and integrate AI within real research environments, making it more accessible, usable and sustainable.
The project brought together expertise from across the universities of Durham, York, Sheffield and Oxford, collectively building insight into how researchers can be supported to apply AI in practice.
Case studies
- Evaluating AI Workflows using the Causal Testing Framework on Bede
Farhad Allian, University of Sheffield - “ML Container Toolkit”: a container system for streamlining AI model deployment on Bede/HPC platforms
Ben Thorpe, University of York - Deep-Learning Driven Adaptive Molecular Simulations on BEDE
Nikolai Juraschko, University of Oxford - Streamlining Materials Modeling: A Unified Entry Point for Matbench Discovery Models in CASTEP
Ben Thorpe, University of York - TextToSpeech with OpenAI’s Whisper on Bede
Paul Niklas Ruth, and Samantha Finnigan, Durham University - Bridging the Gap: an onboarding guide for streamlined access to a biological imaging machine learning toolkit
Tamora James, University of Sheffield