Fall 2017

3D Interaction Designer
Project RE_ is a large-scale architecture project that results with the building of several affordable homes in East Liberty—a low-income, mid-gentrification neighborhood in Pittsburgh, Pennsylvania. 

I am working on this project throughout this whole semester. This part of the project was to help architects and urban planners by developing tools to aid them with their designs.

Context Visualier—

Contextual understanding is essential for an architect. The architect must be able to visualize how their designs will fit into the space, how they will be hit by the sun, etc. The context visualizer was the main tool to help the architects.

Contextual visualizer in action in Unity 3D.

Choosing VR—

We chose Virtual Reality as the method to portray this context because of its capacity to portray spatial relationships as well as produce a level of immersion that allows one to think about the space as a real space. To create this effect, my team would have to recreate the environment (houses, streets, etc) surrounding the building lot in a way that feels similar to the actual site.

Prototyping with the Google Cardboard VR Viewer.

Initial Site Capture—

For our initial site capture, I used photogrammetric techniques to create a rough model of the street surrounding our lot.

Dense point cloud of the model I made of the site to view in VR. This was made with Agisoft’s Photoscan Pro.
Ethan, an Urban Build Architect, enters VR for the first time with a rough prototype on Cardboard VR. After the initial sense of wonder, he was able to critically asses his design.

This is a flythrough that I created of some proposed models placed in a photogrammetric model.

Shifting to Unity and Modelling—

Ultimately, the limitations of Photogrammetry and the dense geometries were too expensive for the computer to process large scenes. Also, the trees and other complex trees would be distractingly terrible in VR. Therefore, we shifted to creating simple models of all the different elements in a two-block radius, and handmaking the environment in Unity itself.

I personally did not do any of the modelling—that was done by the talented Anna, Monica, and Scott. My task was to take their assets (as well as assets I scoured from the Unity Asset Store) and aggregate them into a singular, cohesive environment. This included the implentation of VR tookits, C# scripts, and live post processing.

Screenshots of Anna’s models created in SketchUp.

This is a scrollthrough of the Unity scene I created to view the buildings in VR.
This was taken during my team’s checkpoint.  Viewing the models in Virtual Reality allowed this architect to understand it at a greater capacity, and discuss concerns and excitements to a larger extent.

What I learned—

The main point of this visualizer was to create an environment that would accurately emulate the feeling of the real-world place. This required many qualitative decisions such as wind, flickers in the lights, time of day, etc. I learned how to create a cohesive environment that would portray a lifestyle.

In terms of hard skills, I was able to learn how to use Unity 3D at a far greater capacity. Because I was the only one in my group with any functional Unity experience, I learned how to use the platform quickly, while concurrently teaching my peers how to use it. I also learned how to prototype in VR much more quickly, as well as understand certain limitations of the platform. VR should be viewed in 60fps, which can be a steep performance unless one understands the optimization methods. I’m getting there.

Email  /  Instagram  /  Goodreads  /  Medium  /  Vimeo