Intern Q&A: How to make PIE
Our summer interns worked on a very cool project to make a Virtual Reality prototyping tool called PIE. They learned that making PIE isn't that easy.
Intern Q&A: How to make PIE
Alec Hill, a biomedical engineering grad student from UW-Madison, and Alex Upadhyaya, an Industrial Designer from the Milwaukee Institute of Art and Design, were half of the 2018 summer intern team at Delve.
Along with two other Badgers: biomedical engineering grad student Luz Maria Neira, and mechanical engineering student Laura Kramer, they were tasked with developing virtual reality into a tool for Delve's product development process.
The goal was to create an impactful tool that was useful and adoptable within the company. The interns took this open-ended idea and came up with a tool that will enhance prototyping and concept review. It’s called PIE - Prototype Interaction Environment. This system utilizes HTC Vive VR technology along with the game development software Unity to attach virtual 3D models to low-fidelity physical models by using a tracking sensor. This allows designers and engineers at Delve to physically interact with computer-generated models, making concept review more tangible, and eliminating the need to do extensive physical prototyping.
Throughout the rest of this post, Alec and Alex answer questions about their process, results, and perspective on the project.
How much experience did you have with VR before beginning this project?
Alec: Absolutely zero. The first time I tried on a VR headset was my first day here.
Alex: Yeah, same here. I had no idea of the possible implication of VR in industrial design, let alone how it worked.
What was the hardest part of developing PIE?
Alec: The coding, hands down. Before starting here, I knew nothing about C# or how it worked. Getting the code to work properly and have it not break the rest of our work as things got more complex was a real challenge. Getting the orientation and position of the tracked objects right every time, regardless of how they were modeled, or how the tracker was placed, was also an endeavor. It took some tricky mechanics to get it just right every time.
Alex: For me the hardest part was more abstract – I knew that we would be able to accomplish a working tool, but I was more skeptical of our ability to create a clear repeatable methodology for the tool. It took a lot of trial and error to figure out how everything we developed could be repeated on a variety of projects by a variety of people. It seemed like we wrote and rewrote the workflow hundreds of times.
How many times did Alec drop something?
Alex: This was a running joke all summer, because Alec seemed to drop something all the time– usually something important. My approximation, 759 times.
Alec: Yeah, I’m surprised I didn’t break more things to be honest.
How do you think your differing backgrounds and departments (industrial design, mechanical and electrical engineering) helped complete this project?
Alec: It really makes a difference to have a diverse skill set on your team, no matter what you’re doing. For our sake, all of our skills were put to good use for different parts of the project. I sort of gravitated to the “under-the-hood” work, like the scripting and the mechanics, while Alex excelled at coming up with compelling ways to present our work and make it more applicable to the scope of our project. Even though the other mechanical engineering intern, Laura, was out in San Fran, and the electrical engineering intern, Luz Maria, was busy with other projects, they had a lot of great ideas to bring to the table. They were able to provide a different point of view that really helped us hone in on one project idea. We wouldn’t have gotten this far without them.
How did you decide on PIE for your project?
Alec: We came up with a lot exciting and unique ideas. Ultimately, though, we decided to focus on an idea that aligns more closely with the work that we do here: prototyping. We saw huge potential in VR being used as a tool that could possibly eliminate the need for fabrication and help speed along the design process by allowing industrial designers and engineers see their work instantly, to scale, and with context. VR also allows us to switch quickly between possible variants of our prototypes and assess them in real time.
What skills did you have to develop to complete this project?
Alec: Getting SolidWorks, Unity (our VR software), SteamVR (the plugin we’d use to get the VR system working with Unity), and the hardware to all work in harmony proved to be quite the feat. To get everything to cooperate, we had to learn a lot about how gaming software works, how different 3D file types are constructed, and how to make things behave realistically in VR.
What project did you use to exemplify PIE?
Alex: We developed a mock project to exemplify PIE in use. Delve does a lot of work in the medical field, so we wanted to play off these types of projects but in a fun way with our Beerstorming cart. A wheeled cart is also a great way to show off PIE because it is the right scale for interaction and can be wheeled around the space.
How do you see this tool being used at Delve?
Alec: It would be awesome to see us using PIE to efficiently progress through our ideations in an innovative way. Down the line, it’d also be great to see it being used to showcase what prototyping we’ve done in VR to potential clients.
Why did you complete an experiment on size variance?
Alex: One of the main benefits of using PIE for concept review is that it allows multiple concepts to be reviewed with the use of just one physical model. This saves both time and resources. However, this brought up interesting questions regarding the needed accuracy between the physical model and the virtual model. We started to ask ourselves whether the user’s mind fills in the gaps between what they are feeling and what they are seeing. Ultimately, we wanted to be able to determine how much inaccuracy between the CAD model and the physical model is acceptable– when does the experience becoming jarring?
What did the experiment reveal?
Alex: We ran 36 Delve employees through testing and the collected data revealed distinct preferences. When shown two models the users chose the larger option 80 percent of the time. This is regardless of whether an accurate model was included or not. So, what does this mean? People prefer when what they see in VR aligns with or exceeds the boundaries of what they’re holding. In other words, the mind can more easily adjust when the scale in VR is increased as opposed to decreased.
Why is this experiment important?
Alex: Understanding the way people’s minds work in relation to VR is super important if we– or the larger design community– plans on evolving workflow and process as VR technology evolves. In terms of PIE, our experiment gives our designers and engineers guidelines to follow when implementing our tool. For example, if you are using one physical model for three varying concepts, there is going to be inaccuracy. Our experiment results drive how those inaccuracies should manifest in order to have the most comfortable and believable experience for the users.
How do you see VR changing the design community?
Alec: It might sound outlandish, but I think that one day most of the work designers and engineers do will be in virtual reality. Think about 3D modeling in VR. A little into the future, you’d able to pinch and pull different surfaces and models or have thorough and intuitive user interfaces that appear at the wave of a gesture. You’d get an instant sense of scale and context for your creations and be able to traverse every corner of your model easily, as well as see how mechanisms would work in real-time. All of this could be done inside the computer, and little to no fabrication would be needed. Sure, we’d still need to prototype things in real life at some point, but I think there’s a lot to be said about the potential for VR to expedite the design process.
Alex: I agree about the infinite possibilities of VR in design, however I believe the gap between VR technology and using it within the design process is significant and will close slowly rather than quickly. VR technology is evolving but I don’t think it’s close to being as intuitive and efficient as the design process requires. We are at the point where the technology is more cumbersome than convenient and time-saving. For example, sketching in 3D within VR is great– it’s fun and opens a whole new dimension (literally) for ideation – however it will never replace pen and paper.