A new Royal Society paper by our Organisational Members Johnson Matthey shows multiscale modelling moved from a being a research tool to core industrial infrastructure.
The authors highlight that computational modelling is now “indispensable in an industrial setting, fuelled by powerful algorithms and increasing computational speed” and that atomic‑level insights directly shape large‑scale industrial performance. In their words, “understanding reactions at atomic and molecular levels directly affects large-scale industrial processes.”
The paper stresses the need for:
- Integrated modelling–experimental workflows, calling the link between prediction and validation “essential”
- Investment in shared infrastructure (HPC, data standards, interoperable platforms)
- Cross‑disciplinary training to build a modelling‑literate workforce
- Translational funding to move methods from academia into industrial deployment
As expected, machine‑learned potentials are enabling simulations that were previously “intractable owing to prohibitive computational costs.” This opens the door to faster catalyst discovery, better materials design, and shorter development cycles across clean energy, chemicals, automotive, aerospace and more.
The paper’s conclusion strikes a chord with our EMMC community: multiscale modelling “significantly accelerates the design and development of new industrial products.”
