FullCAM web transition

FullCAM is the model behind Australia’s national carbon accounting for the land sector. Its outputs feed emissions reporting. Its users are not a general public — they are scientists, consultants and project proponents who have often been running the desktop version for a decade and know exactly what it does.

Moving a tool like that to the web is not a reskin. The desktop application encoded years of expert workflow in its structure, and a browser version that simply reproduced the old screens would have been slower and worse. A version that ignored them would have been rejected outright.

The approach

1. Interview long-standing users about what they actually do in a modelling session, not what the manual says the tool is for.
2. Decompose the desktop application into the tasks it supports, separating the parts that are genuinely scientific from the parts that are legacy interface convention.
3. Rebuild the information architecture around those tasks, then test prototypes against real modelling scenarios rather than invented ones.

Designing for expert users

Expert tools invert most consumer UX instincts. Density is a feature. Progressive disclosure hides things people need. Onboarding is close to irrelevant — the population is small, trained, and returning. What matters is that a long-running simulation is legible while it runs, that inputs can be traced to outputs, and that a wrong parameter is visible before an eight-hour run rather than after it.

The web transition also changed who could reach the model. A desktop install is a barrier; a browser is not. That shifts the design brief from "reproduce the tool" to "make the tool defensible to a wider set of people who will now open it".