Digital Thread – Theme summary
Artificial intelligence (AI) has transformed the way we conduct materials research. Generative AI models are one example – but more widely, this transformation is driven by a range of innovative digital techniques that leverage large datasets, machine learning, as well as automation and robotics.
The Oxford Advanced Materials Network brings together world-leading experts in this rapidly evolving space, working at the interface of materials science, machine learning, and AI research more broadly. By combining our strengths across disciplines, we are able to tackle challenges that go well beyond what individual research groups or individual departments could achieve on their own.
Opportunities
- From atoms to devices: Oxford researchers are among the pioneers in machine-learned interatomic potential (MLIP) models: atomic-scale simulation techniques that are trained to reproduce quantum-mechanical data. These models enable accurate atomistic simulations on the ten-nanometre length scale and beyond, relevant to real-world devices.
- Learning from spectroscopy data: OxAMN members are closely connected to large UK facilities such as the Diamond Light Source, enabling the integration of high-quality experimental spectroscopy data into machine-learning models. These efforts can help to develop AI tools that are informed by experimental reality and capable of interpreting complex spectral signatures.
- Towards digital twins: Embedded in Oxford’s interdisciplinary research environment, OxAMN members can help to create the foundations for “digital twins”: virtual models that mirror the behaviour of real-world materials and devices. These approaches open new possibilities for predictive design of complex material systems.
