Prototype: does a rendered-splat route view read as navigation? #12
Notifications
Due Date
No due date set.
Blocks
Depends on
#14 Write the Session 1 spec (pose to splat rendering)
lars/splat-indoor-navigation
#3 Research: WebGL2 splat viewer performance inside iOS WKWebView
lars/splat-indoor-navigation
Reference: lars/splat-indoor-navigation#12
Reference in New Issue
Block a user
Question
Make something cheap and concrete to react to, and judge whether the chosen presentation actually works as navigation.
The decision was a rendered splat view — the photorealistic model drawn from the user's pose with the route overlaid, no camera passthrough. That is the product, and its usability is unvalidated. The open risk: a photorealistic render of a corridor you are physically standing in invites direct comparison with reality, and every artifact and every centimetre of pose error becomes visible. Passenger-facing navigation may want something deliberately more abstract.
Build a rough version — a splat view with a route overlay, a "you are here" indicator sized by confidence, and the next instruction — and react to it. Compare against at least one alternative treatment (stylized/abstracted rather than photoreal, or a hybrid with a 2D deck map) so the choice is made against something.
Link the prototype from this issue. The output is a decision about the presentation, not shipped code.
Part of the wayfinder map #1.