The 3D is rendered in real time with Rive’s GPU Canvas. The people and sensors are simulated, so I could test every behavior without a real room. View Models and data binding keep it all editable in the Rive Editor.

Been exploring the Rive CLI, so I built a concept for a kinetic light installation, all in @rive_app 28 fins react to where people stand, how they move and how long they stay. You can design how it behaves, run it live and replay past sessions.