We created a virtual reality museum in VRChat which can be accessed by players in VR or viewed on YouTube through the perspectives of content creators giving viewers a museum tour. This VR Museum contains posters and art exhibits related to sleep science.
We wanted to create an immersive experience that lets people learn about sleep science in a new way. Being able to combine a handful of our shared interests in design, museum exploration, art, education, VR, and sleep science was an added bonus.
Dimitri and I took Prof. Matthew Walker’s class on Psychology of Sleep at UC Berkeley. One of the course requirements was a ‘Sleep Outreach Project’ — the goal is to help educate the public about sleep in a creative way, using all sorts of possibilities for choice of medium, engagement, and content delivery.
Physical museums are geographically restricted and people must travel to them in order to experience the exhibits. An additional problem is that scientific concepts can be dense and unengaging for several people. People also often have different learning styles: while audio input might be one person’s preference, another person could be a visual learner.
Creating a VR experience could help improve education by creating memorable experiences that cater to all sorts of learning styles for people. Approaching education in this novel way allows for experiences that would otherwise not be possible in the physical world: for example, making a brain 3D model light up for different region activations for different stages of sleep.
We had several lecture and section notes, readings, and video recordings that were covered over the course of the semester. After revisiting the course material, we narrowed down on the topics we found most interesting and synthesized the relevant information. Discussing which topics would translate well into different 3D assets and exhibit pieces also helped us highlight certain sections as we filtered through the content.
Our goal behind doing a Berkeley brand audit was to design the museum experience that would reflect it as a space, albeit virtual, belonging to UC Berkeley or at the very least serving as an extension to it. We set out to better understand the design system behind the Berkeley brand and made note of the colors, typography, graphic elements, and photography among other things.
As museum enthusiasts, this was the perfect opportunity to take a little break (which in this case was also relevant to work) and visit the Legion of Honor Fine Arts Museum in San Francisco. We spent hours soaking up the paintings, interiors, and beautiful statues that inspired us to include sculptures.
Slightly closer to home, we went onto campus and visited the Valley Life Sciences Building and Doe Library: both of which are architectural marvels that we are fortunate enough to call part of our UC Berkeley campus. This visit involved browsing halls of informational displays and mini display models that were tailored to the Berkeley brand guidelines.
In order to study the affordances and interactions that a VR museum could have, we also visited various user uploaded VRChat rooms that came up when we searched for words like ‘museum’, ‘exhibition’, and ‘art gallery’. We made note of elements that enhanced the experience such as audio tours, any specific buttons that changed day/night modes, and ideas that came to us while we walked through these spaces.
We discussed various approaches that we could adopt for this project. By not setting any initial constraints, our ideation session yielded all sorts of results: some feasible and some extremely ambitious. Given the time constraints, we had to prioritize our existing technical know-how and skills over our ideal scenarios that required us to learn new software and design processes.
In an attempt to finalize our project ideas, we made an impact-effort chart. Some of our ideas, such as the sleep deprivation hallucinatory experience, involved a steep learning curve which we could not devote ourselves to given the amount of time we had. As we had been planning on working on a collaborative VR Design project for some time, this seemed like the perfect opportunity and we were able to finalize the decision to build a VR Sleep Museum almost instantly.
Next, we shifted our attention to brainstorming ideas about specifics for the exhibitions: their interactivity, possible visualizations, the types of assets that could be included, and the scope for VR to enhance education around specific topics. We took factors such as availability of pre-existing 3D models, animation requirements and their difficulty level, and educational impact into consideration.
To scope the feasibility of certain ideas we had, we mapped our ideas on another impact-effort chart. Posters and statue art pieces emerged as the optimal solution as they could still be experienced by our video viewers, were easy to implement, displayed more content, and did not involve a learning curve for new interactions in the VR environment.
In order to zero in on the visual elements and overall aesthetic direction we wanted to take our VR Museum in, we curated mood boards with images of various Unity store assets as well as items that we found particularly usable or interesting from our in person and virtual reality based contextual inquiries.
At this point, we looked into how we could upload our museum to VRChat. After digging through several online forums and obscure documentation, we came to a scary realization: in order to be able to upload avatars and worlds to VRChat, a user must have spent 50+ hours on the platform and have befriended enough other VRChat users to have a certain trust ranking.
We had to increase our logged hours on VRChat, fast. Luckily for us, our experiment (quite on brand for a sleep science project, I must say) with stuffing a pillow into Dimitri’s Oculus Quest 2 to keep it running VRChat paid off. We were able to make up the remaining required hours and get world upload permissions within the next two days.
After exploring various set options ranging from more modern-looking art gallery style architecture to museum kits with marble and Greek statues as adornments, we settled on a showroom hall environment from the Unity asset store. Not only was this a cost effective option that would save us hours of 3D modeling and asset creation time, but it also most closely resembled what an extension of the UC Berkeley campus space might look like due to its striking similarities to Anna Head Alumnae Hall, Berkeley.
Next, we shifted our attention to our findings from the Berkeley brand audit and the posters we viewed in the Valley Life Sciences Building to start crafting our own set of informational displays for the museum. We took note of various stylistic elements and attempted to maintain consistency. This led us to create two general templates for posters, one square and one rectangular, that we then populated with sleep science information that suited specific formats best.
Finally, we looked into various models that could be added to the display cases in order to draw people in by further indicating the concepts being discussed in various posters. We included a full sized bed, fast food, planes, and a few other assets.
Creating a logo for the museum and adding in the UC Berkeley seal (after spending several hours editing, texturing, and mapping it to make it work) to the center of the hall served as our finishing touch!
We are extremely pleased with the outcome. The hotdog avatar was also a huge hit in amusing several viewers and helping them enjoy the learning experience even more by keeping them entertained and engaged.
Working on this project was nothing short of delightful because it allowed me to combine so many things that I’m passionate about. However, there were days where technical hurdles made it incredibly hard to keep moving forward and caused some frustration. That being said, it felt extremely rewarding when we surpassed the errors and got everything to work. I learnt that I am a determined, can tackle projects without much external guidance, and learn new skills on the go.
Given more time, I would have loved to explore VR animations and additional affordances like audio tours. Embellishing the museum with smaller details such as entrance/exit signs, directions, carpets, and other finishing touches would also help make it look more realistic. If I had more time, I would have loved to implement the ideas we had for making the exhibits interactive (such as placing electrodes on a dummy’s head or the brain region activation model).
I learnt that Dimitri does not let scary Unity errors (or anything really, for that matter) intimidate him. When he sets his mind to something, he is not afraid to spend hours scouring heavy technical documentation to complete the task at hand. Not only does he ensure that everything is complete, but he also never compromises on quality and will go above and beyond to make everything pixel perfect. Working with him has helped me become a better designer and teammate.
Completing this project was a very satisfying, rewarding experience. I realized that I am good at unconventional problem solving, combining different concepts to come up with actionable solutions, and pushing myself to experiment. However, dealing with heavier, technical problems like tackling dozens of errors to upload our project to VRChat is not my strong suit. I am working on improving in this area by staying focused and having more trust in myself.