SpatialTell

*Independent Research Project

Ongoing*

Full Demo: https://youtu.be/xrT5RJzHc70

SpatialTell is an immersive, VR-based prototype for anchoring spoken knowledge directly onto the architectural elements it describes. Instead of composing through files and menus, users work in-scene: they point to a beam, plaque, or doorway (via controller), speak in natural language, and receive subtle ambient feedback—light/outline cues and brief confirmations—that their statement has been captured and attached to that specific object. Treating authoring as an experience-first interaction, the environment itself behaves like a non-anthropomorphic “space agent” that listens and gently clarifies when needed. Each contribution is stored as a minimal, persistent, object-anchored note so that context is preserved and later retrieval stays tied to place.

SpatialTell reframes heritage work from “viewing content” to “authoring in place,” lowering barriers for non-technical contributors and making each addition visibly live where it matters.

Looking ahead, the same pipeline can support VR education and tourism once a site’s corpus grows, and extend to participatory urban memory, where residents contribute situated stories to their neighborhoods.