Central Viewer
Central Viewer brought geological model review into the browser. My work focused on making complex 3D data easier to navigate, interpret, and discuss across teams.
Research the conventions before designing the workflow
Section titled “Research the conventions before designing the workflow”I reviewed common patterns across 3D, CAD, BIM, mapping, and developer tools: project trees, breadcrumbs, toolbars, dashboards, revision histories, object panels, and model navigation. These were not competitor teardowns. The question was why certain patterns recur in technical software, which jobs they support, and where they would or would not translate to Central Viewer.
A cross-application scan of toolbars, model views, navigation, and object controls.
Common patterns were treated as hypotheses to test, not defaults to copy.
This gave the team a shared vocabulary for evaluating options. A familiar pattern was useful only when it helped users orient themselves, understand model state, or act on the right object without adding another layer of technical UI.
Start with the review workflow
Section titled “Start with the review workflow”Desktop modelling tools already supported specialist analysis. The gap was what happened when a model needed to be shared: collaborators moved between the model, screenshots, files, and conversations, while occasional users could struggle to locate the relevant scene or understand what had changed.
The discovery work combined workshops, evaluative review, and usability sessions. I used the findings to focus the design on three connected needs: orienting people in 3D space, making the project structure understandable beside the scene, and keeping discussion attached to the model context.
A workshop with end users and the internal team made the interaction model tangible before the interface was committed.
The synthesis grouped recurring needs across 3D access, structure, slicing, comments, and notifications.
Design for the work around the model
Section titled “Design for the work around the model”I explored how the project tree, scene, object controls, and properties could work together. The main trade-off was density: showing every control at once made the viewer harder to scan, while hiding too much weakened expert control. I worked through panel arrangements and object-selection patterns to keep scene, structure, and action in clear relationship without flattening the underlying geology.
The detailed interaction review exposed where controls applied to selected objects and where they did not.
I also treated collaboration as part of the review workflow, rather than an add-on. Shared scenes and in-context comments were intended to help two people get to the same point of interest and discuss it there; revision visibility helped preserve which model state that discussion referred to.
Validate the interaction model
Section titled “Validate the interaction model”I worked with end users and the internal team through workshops, evaluative reviews, and usability sessions. We examined whether people could find actions, select and inspect objects, share a scene, comment in context, and understand revisions. The testing matrix brought those tasks together so we could see where the interaction model worked and where it still carried too much desktop complexity. The aim was to validate the browser experience as a review tool, not just confirm that a 3D model could render.
Outcome
Section titled “Outcome”The work clarified a cloud-review direction that connected model structure, scene interaction, and collaboration. My contribution was to shape the experience around the work people needed to do with the model: navigate it, inspect it, and discuss it together, while preserving specialist depth.
The model is the centre of the experience; the surrounding workflow makes it reviewable with others.