Collaborative Design Review Sessions in Virtual Reality: Multi-Scale and Multi-User

Shahin Sateei, Mattias Roupé, Mikael JohanssonProceedings of the International Conference on Computer-Aided Architectural Design Research in Asia 2022 · DOI

TL;DR

This paper demonstrates how a multi-scale, multi-user Virtual Reality (VR) system can revolutionize architectural design reviews. By allowing stakeholders to switch instantly between a "bird's-eye" miniature model and a 1:1 immersive experience, the system overcomes common VR disorientation issues and enables non-experts to identify critical design errors that were invisible in traditional 2D drawings.

What Problem Does This Paper Address?

In the construction industry, design reviews are critical for catching errors before ground is broken. However, traditional mediums like 2D plans and 3D models on screens are often difficult for "laypersons"—such as school teachers or facility managers—to fully comprehend.

While immersive VR offers a better sense of scale, it introduces its own set of problems: users often get disoriented in large models and struggle to understand the "big picture" or the functional links between different rooms. Furthermore, VR is frequently a solitary experience, making it difficult for a diverse team (architects, clients, and end-users) to collaborate effectively in real-time. This paper investigates how a collaborative, multi-scale VR environment can bridge these gaps.

Methods

The researchers conducted a case study involving two VR workshops for a new elementary school project in Gothenburg, Sweden.

  • The Tech Stack: The team used three Oculus Rift S headsets running BIMXplorer software. This software allowed for the direct import of IFC (Industry Foundation Classes) files without needing extensive optimization.
  • Multi-Scale Interaction: The core innovation was the ability to switch between two frames of reference:
  • Allocentric (Miniature Model): A "bird's-eye" view where users look down at a miniature version of the building to understand layout and logistical flow.
  • Egocentric (1:1 Scale): A full-immersion view where users "walk" through the building at actual size.
  • Multi-User Collaboration: Multiple participants could enter the same model simultaneously, represented by avatars, allowing them to point out issues and discuss them in real-time.
  • Data Collection: The researchers used video-recorded observations of the workshops and follow-up open discussions to categorize the types of design errors found.

Datasets

As a qualitative case study in construction management, this paper does not use standard machine learning datasets. Instead, the "data" consists of:

  • Architectural BIM Models: Real-world IFC files of a new elementary school.
  • Empirical Observations: Video recordings and transcripts from two workshops (n=8 and n=6 participants).

Key Insights and Ideas

  • The "Hybrid" Map: Interestingly, while end-users often claim they cannot read 2D plans, they used the VR miniature model (which is essentially a 3D version of a 2D plan) with high proficiency for orientation and navigation.
  • LOD over Photorealism: The study found that increasing the Level of Detail (LOD)—specifically by adding furniture—was far more important for spatial understanding than high-end photorealistic lighting or textures. Furniture provided the "purpose and context" needed for users to judge if a space was functional.
  • Scale-Specific Problem Solving: The two scales served distinct purposes. The miniature model was primary for discussing logistical flow (e.g., how students move between floors), while the 1:1 scale was essential for room-specific functionality (e.g., checking if a teacher's desk area felt too cramped).
  • Common Frame of Reference: Multi-user VR acted as a "communication tool" that leveled the playing field between architects and clients, allowing for a more participatory design process.

Key Results

The workshops successfully identified several design issues that had been overlooked in 2D reviews:

  • Spatial Perception: Participants identified "cramped" areas, such as narrow serving spaces in the restaurant and tight layouts in classrooms, which would have been very costly to fix after construction.
  • Sightlines: Users discovered missing glass partitions and realized that certain sightlines (e.g., between the kitchen and restaurant) were obstructed or needed adjustment.
  • Logistics: By using the miniature model, the team identified errors in door swing directions and missing security walls between floors.
  • Qualitative Feedback: One project leader noted: "The design issues we discovered and discussed today were difficult to see let alone discuss about in the 2D drawings we have used up until now."

Limitations and Future Work

The authors acknowledge that this was a single case study with a relatively small number of participants. A few areas for future exploration include:

  • LOD Mapping: Investigating exactly what level of detail is required for different phases of the design process (e.g., when is furniture "enough," and when do you need lighting?).
  • Cognitive Load: While VR is argued to be less cognitively demanding than 2D, more research is needed to see if the "instant switching" between scales causes fatigue or confusion over longer sessions.
  • Causality: It remains unclear if the immersive VR taught the users how to better understand the 3D layout, or if the miniature model is simply a superior representation for overview compared to traditional 2D plans.

Summary updated

Abstract

The use of Virtual Reality (VR) for design reviews in projects is becoming more common in construction. However, the use of VR in these processes has been limited to been used more as a complementary reviewing tool alongside information medias such as 2D drawings and 3D models. Furthermore, immersive VR has been argued to have limitations when it comes to orientation and understanding and reasoning about functional links between physical layouts in a facility. This paper presents a case study of a VR system used during design reviews that support end-users to switch between different representations and scale i.e., miniature model/bird-eye view, and a 1:1 scale experience of the facility. The data gathered, consisted of recorded observation of the VR based design review process and study what type of discussion and design errors that was found during two VR-workshops connected to a new elementary school. The result shows, that by supporting switching between miniature model and 1:1 scale VR experience facilitated spatial orientation and understanding and collaboration across disciplines in the project. The study also show how collaborative immersive VR can be used as an efficient communication-tool during the design process in a real-world project.
BibTeX
@inproceedings{sateei2022collaborative,
  author = {Shahin Sateei and Mattias Roupé and Mikael Johansson},
  title = {{Collaborative Design Review Sessions in Virtual Reality: Multi-Scale and Multi-User}},
  booktitle = {Proceedings of the International Conference on Computer-Aided Architectural Design Research in Asia},
  year = {2022},
  doi = {10.52842/conf.caadria.2022.1.029},
}
Figure