VR/AR for Urban Redevelopment
Designing with East Liberty Residents
Interaction Design Rapid Prototyping, Spring 2019
Team Members: Tom Corbett (Faculty Advisor), John Butler (CS/Architecture) Davis Dunaway (Design), Meijie Hu (Design), Dorcas Lin (Design) Jaclyn Saik (Design)
Tools & Techniques: HTC Vive, Unity Game Engine (SteamVR), C#, SketchUp, Rapid Prototyping, Agile PM
The Urban Design Build Studio (UDBS) is working to design and develop an affordable housing prototype for a distressed neighborhood of Pittsburgh (East Liberty) by following a mixed economy model. These neighborhoods are faced with a number of socio-economic issues, and new developments can give rise to concerns such as gentrification. As designers working with UDBS, our efforts were focused on:
Creating virtual environments that accurately represent the design for the housing prototype within East Liberty.
Continuing to design and work with the engagement concepts to turn them into full applications to be deployed in the community
The goal for my team was to create a virtual experience in Unity of the house interior that current residents of East Liberty can walk through and get a better understanding of the benefits that come from the proposed architecture because much of the materials are recycled thereby creating cheaper houses. There's heavy emphasis for us to design with these residents as a major pain point this VR experience addresses is the skepticism current residents might have about the construction of new buildings in their area.
Because of the various components involved in the experience and the major learning curve that we all had to overcome, we decided to split up the work into three categories: teleportation/navigation menu, sensory sounds, and textures.
I specifically worked on the teleportation of the interiors between floors and across rooms with another designer, Davis Dunaway.
Process Familiarizing with Unity & SketchUp
The most difficult part of the process for me was understanding how space translates to 2D form and scaling models/prototypes properly. Eyeballing really didn't cut it this time, so before even touching the empty lot where the proposed infrastructure was to be built, I modeled one of the surrounding houses near the lot as practice. This helped me understand some of the standardized features across all houses like door sizes and brick layout. This model also contributed to the exterior/neighborhood VR experience that another team was working on.
(Surrounding House Model in Sketchup, Surrounding House Model Visualized in Unity)
Spatial Considerations User Movement
I first mapped out a list of general pain points within VR experiences after going through the game tutorials within Unity. Nausea and problems with moving around in the denoted space for longer than 10 minutes were the major issues. With this in mind, we decided to focus our teleportation priorities on:
1. An efficient experience that relays all of the necessary features of the living space in around 5 minutes
2. Scaling options for people of different heights to have the same experience
3. Intuitive wayfinding through buttons
Following these three priorities consistently throughout the design process allowed us to design for inclusivity and accessibility. Because this experience was to be a small part of a greater opening night event all about the proposal, this activity had to first overcome the technological barrier and potential fear of VR before being able to create an experience that can stand alone and be engaging/lifelike.
Spatial Logistics Main Level Floor Plan
The architects in the UDBS team who designed the infrastructure gave Davis and I the floor plans and we decided to break up the main floor depending on how many teleportation points we wanted.
We ended up deciding to move users by section versus by random points in the room because it was less disorienting and people would be able to see the main features of the living room and kitchen space better.
Main Floor (Left to Right): Backyard, Kitchen, Dining Room, Living Room, Porch
Prototyping Testing Teleportation Methods
We imported the floor plans and models into Unity and tested out two teleportation methods: 1) target teleportation by points indicated by a circle and 2) broad teleportation by squared-off sections.
1) The first option allows for more user choice and might require the user to have more experience with VR to navigate through space easily
2) There's less choice for the user
Left: Method One, Right: Method Two
Method One from Front Door to Living Space
Method Two from Dining Room to Living Room
Play Testing Extremities
Next, we wanted to user test and compare the feelings that came from varying eye levels. In the gif below, we set the eye level to the top of the ceiling of the first floor and tested this among our peers. The immediate reaction in this space was the feeling of imbalance and a sense of falling.
Because each of these methods had their pros and cons, we decided to use the block teleportation (method 2) for the main transport mode and method one in order to indicate specific landmark points including stairs. The gif below shows this implementation in the main living room.
Once we were able to test which teleportation methods were easiest for larger movements, we had to decide how we wanted to treat smaller interactions with the house like opening drawers, turning on lights, etc. We first tested this out on a cabinet drawer and found that it becomes rather laggy if a hold-and-pull method was used as the individual drawer would wobble because in Unity, the plane that indicated the interaction was too close in proximity to some items we wanted to be static. We found that holding the trigger and generally pulling the desired direction led to the smoothest interaction.
We decided to implement the hold-and-release for most items in the house. The gif below shows this method used for opening the door along with turning on and off the light in the bedroom.
Feature Considerations Staying Cohesive as a Team
Because we worked on all of the parts like daytime/nighttime toggles, textures, and sounds separately even internally within our team, it was hard to test to see if things worked all together. For instance, we had to change some of the teleportation planes because it interfered with the menu for changing scenes. We also had to meet together to determine at exactly which point the sounds were going to begin and end. Overall, it was a fun experience once we got past the SteamVR learning curve.
Final Walkthrough Video Exterior, Interior, and Toggles (~5 min)
At the end of the semester, we combined out interior teleportation with the other exterior team to create a walkthrough of the East Liberty neighborhood and proposed housing. This video is a walkthrough we did with the UDBS team prior to our presentation pitch at the Carnegie Mellon Meeting of the Minds conference.