Translating complex molecular research into an intuitive visual narrative through abstraction and 3D animation.

Overview

Much of my practice explores how invisible systems can be translated into a digital representation.

This collaboration with Google DeepMind became an opportunity to apply that approach to a molecular process that cannot be directly observed.

The final film represents months of experimentation, visual research, and close collaboration. What follows is the story of how that visual language evolved.

Interpreting the Science

Google DeepMind's research uses AI to accelerate the discovery of new crystalline materials for applications in energy storage and advanced computing.

The client supplied scientific diagrams describing how atoms organise themselves through different stages of crystallisation. Rather than recreating these literally, they became the foundation for developing an entirely new visual language.

The challenge wasn't to illustrate science, it was to make an invisible process feel intuitively understandable. 

Developing a Visual Language

The earliest explorations were intentionally broad.

Rather than searching for finished imagery, I experimented with behaviours - instability, attraction, transformation and eventually order. Every render became another step towards understanding how abstract science could be experienced visually.

Many of these experiments never appeared in the final film, but each helped shape the interpretation that eventually emerged.

From Chaos to Order

As the project evolved, the visual language gradually became more defined.

The opening sequences focused on unstable molecular formations, constantly shifting, colliding and reorganising. Gradually these behaviours became more structured, allowing crystalline forms to emerge naturally from the apparent disorder.

Refining Material Behaviour

Once the overall visual language felt established, attention shifted towards refinement.

Lighting, materials, composition and colour were carefully refined to balance scientific credibility. Every adjustment strengthened the sense of materiality while remaining true to the research.

Many of these explorations never appeared in the finished film, but each contributed to developing a visual system that felt both believable and beautiful.

Building the Final Sequence

The completed visuals became part of Google DeepMind's communication around AI-assisted materials discovery, helping translate complex molecular research into an intuitive visual experience.