Project · 01 · Unity VR

ARTH.

Adaptive Reuse Thinking for Housing — a VR learning environment.

Unity XR · VR · User studies · 2024 ->

ARTH montage: factory space, object placement validation, and 4E cognition framing

Summary

ARTH (Adaptive Reuse Thinking for Housing) is a Unity-based VR learning environment where learners transform an old factory/warehouse into a new interior program (e.g., office, living room, bedroom, or a full residential layout). The experience is structured as a hands-on adaptive reuse task: users clear existing industrial elements, then iteratively build and furnish a new layout using a digital tablet tool, grid-assisted placement, and three interaction modes (placing, demolition, and editing). Real-time “valid/invalid” feedback (green vs. red) encourages rapid iteration while keeping placements spatially plausible.

A preliminary user study with students from different backgrounds suggested the experience is usable and engaging—overall results were promising.

Unity VR XR Interaction Adaptive Reuse 4E Cognition

Project at a glance

Learning scenario

  • Starting point: an industrial warehouse/factory filled with outdated elements.
  • Goal: transform the space into a functional and stylish residential interior through iterative design decisions.
  • Core activities: site exploration, selective removal, spatial planning, construction of partitions/ceilings, furnishing, and material selection.

My role

  • Designed and implemented the VR experience flow, UI, and interaction logic
  • Led study deployment and data collection (setup, participant sessions, and instrumentation)
  • Synthesized findings and co-authored project manuscripts

Tech stack

  • Engine: Unity (C#)
  • VR: XR Interaction Toolkit
  • Interaction tools: virtual tablet library, laser pointer selection, physics-based constraints

Project video

If it doesn’t load, use the direct link below.

Experience flow

ARTH is structured as a complete, learn-by-doing adaptive reuse workflow, from understanding the existing conditions to testing and refining a new spatial proposal:

1) Orient + read the site

Start inside the original warehouse and explore the existing elements and structure before changing anything.

  • Walk the space at full scale and identify constraints (columns, beams, openings).
  • Use exterior cues (e.g., surroundings and signage) to build spatial orientation.

2) Clear what’s obsolete

Switch to demolition mode to remove outdated industrial components while preserving primary structural elements.

  • Remove non-structural partitions and machinery to reveal design opportunities.
  • Use removal as an explicit “adaptive reuse” decision: what stays vs. what goes.

3) Build + furnish iteratively

Use placing mode (tablet + ray selection) to create layouts, then editing mode to refine.

  • Place walls/ceilings with grid assistance for cleaner alignment.
  • Place furniture with collision checks to prevent unrealistic intersections.
  • Adjust orientation and position through direct manipulation.

Tablet interface + placement feedback

The tablet UI is the “external workspace” inside VR: it reduces menu friction and keeps the learner focused on spatial reasoning rather than tool hunting. A grid system and constraint checks provide immediate feedback, supporting realistic decision-making.

Warehouse before/after: original industrial condition and cleared space with tablet interface
Before vs. after. The original warehouse condition (left) and the cleared design space with the tablet interface for selecting components and materials (right).
Placement validation: invalid placement highlighted red, valid placement highlighted green, and final object placed
Placement feedback: invalid placements highlight in red, valid placements highlight in green, then the object is committed to the scene.

Grid + constraints (why it matters)

  • Structural placement: grid guidance helps align walls and ceilings for cleaner spatial organization.
  • Conflict detection: green vs. red feedback reveals collisions and prevents unrealistic intersections.
  • Lightweight “structural logic”: ceilings can start above walls, but extending them requires additional supporting walls—encouraging planning and load-bearing awareness.

Learning design: 4E cognition as a blueprint

ARTH is designed around the 4E cognition framing—treating learning as something that happens through the body, through action, inside context, and with the help of tools.

4E cognition diagram: embodied, embedded, enacted, extended
4E cognition framing used to guide experience design: embodied interaction, embedded context, enacted learning-by-doing, and extended tools that offload and support reasoning.

Embodied

  • Full-scale navigation, reaching, pointing, grabbing, and spatial manipulation.
  • Physical engagement supports “thinking with the body” in layout decisions.

Embedded

  • Learning occurs inside a realistic industrial shell with constraints and contextual cues.
  • Adaptive reuse is framed as context-sensitive decision-making (site + structure + program).

Enacted

  • Knowledge emerges through action: remove, place, test, revise, and iterate.
  • Immediate feedback makes consequences of decisions visible in real time.

Extended

  • The tablet UI and placement system act as external cognitive supports.
  • Tools reduce memory/coordination load, freeing attention for spatial reasoning.

Outcome

  • A VR prototype that supports an end-to-end adaptive reuse scenario: from exploration and demolition to building, furnishing, and refinement.
  • A reusable interaction toolkit for mode switching, tablet-based selection, grid-assisted placement, and constraint-driven feedback.
  • Study-ready deployment setup; early evaluation feedback was promising.

Links