Investigate the Bowels of a Derelict Spacecraft in this VR Experience

Aberration truly embraces the term “VR Experience”. Designed to be played in a 15’x15’ play space with a wireless HTC Vive, this asymmetrical co-op puzzle game will mess with your mind in more ways than one. Explore the recently resurfaced U.S.S. Innsmouth and work with your fellow crewmate who has remotely accessed the craft to repair its Infinity Drive and uncover the mysteries of its sudden disappearance.

Tricking the Mind to Maximize Space 

When designing Aberration, our team wanted players to feel fully immersed when exploring the rooms and corridors of the U.S.S. Innsmouth. We wanted players to really feel like they were walking around the ship, so early on we decided that the only way for the player to locomote was going to be by physically walking around the playspace. No teleportation whatsoever. This created a unique challenge for me, as the experience designer, to solve, “How do you fit an entire starship into a small VR playspace?”

While we couldn't make the playspace physically bigger than 15’x15’, there were some level design tricks we could employ to make the space feel larger than it actually was. Inspired by the paper “Impossible Spaces: Maximizing Natural Walking in Virtual Environments with Self-Overlapping Architecture”, I decided to design the rooms and corridors of the Innsmouth architecture that intersected itself. As described in the paper, players are generally not able to tell if virtual rooms are overlapping if the overlap is no more than 30%. We used this trick in perception to gain roughly 50ft² of additional virtual floor space. 

Creating Impossible Architecture 

One of the definitions of Aberration is “a departure from the expected”. The team’s narrative and audio designer, Jeena Yin, was already embracing this motif by creating a preternatural atmosphere through the use of mysterious audio logs. How could the level design encapsulate and heighten this uncanny feeling we wanted to convey? 

We were already deploying impossible architecture throughout our game, just in a way that wasn't noticeable to the player. I began to experiment with slowly revealing the weird properties of this architecture as the game progressed. I started by increasing the overlap between the corridors and rooms, pushing just past the 30% mark that the paper had stated as the limit of the unnoticeable. While players weren’t able to articulate that the rooms were overlapping, they were able to realize that something was off about the space they were exploring, creating a strong sense of unease. As the game progresses, the game continues to more overtly reveal its impossible architecture, culminating in an infinitely-looping hallway where the only way to exit is to turn around.

Designing Puzzles 

In parallel to designing the experience of walking around the ship, I was also tasked with designing a puzzle for each room the player encountered. Another definition of Aberration is when two rays of light fail to converge, so we decided to use the idea of manipulating lasers as our primary puzzle mechanic. In each puzzle, the players are given one or more stationary laser emitters and colored receivers. The goal for each puzzle is to redirect the light using a series of mirrors, color filters, and light combiners to activate each receiver. 

Before implementing puzzles directly into the game, we created a series of paper prototypes. We printed the layout of each room onto a piece of paper and used lengths of colored string to correspond to each of the possible laser colors. We would ask playtesters where they would like to place mirrors and then adjust the string accordingly to visualize the result. We tested and refined around ten distinctly different puzzles before choosing four puzzles to implement. When selecting the final four puzzles, we opted for the ones with an easier degree of difficulty. This was because we didn't want players to get bogged down solving a single problem when we wanted the emphasis of the game to be on the exploration of the ship. We knew that the unique part about our game was its preternatural atmosphere and its use of impossible architecture, so we wanted to make sure that that is what the player spent the most time engaging with.

An Active VR Experience Anyone Can Enjoy

Explore desert ruins, solve ancient puzzles, and scale stone spires in your quest for your father’s mythic staff. The Sands Below is all about finding your route and climbing your way to the top, so we wanted to make sure that our climbing experience was both satisfying and accessible. We developed climbing controls that are intuitive enough for a first-time VR user to begin playing with ease, while still being expressive enough to allow veteran players to perform advanced maneuvers to sling themselves to the summit. This VR platformer is sure to be fun no matter what your prior experience with VR is.

Avoiding Virtual Vertigo

The biggest challenge while designing this game was how to make the act of climbing feel both natural and enjoyable while in VR. Normally, when playing a VR game, what you see in the headset is the game trying to mimic the motion your brain is expecting to see in order to prevent motion sickness. The act of climbing, however, necessitates a lot of vertical movement, and because the player isn’t actually able to move up and down in real life, we had to figure out a way to move the virtual camera in such a way that it didn’t cause nausea.

The way we started testing our various ideas was to test them on ourselves. As valuable as blind play testing feedback is, it was important to us that we only put this game in the hands of playtesters once we could say with a high degree of confidence that it would not cause more nausea than any other VR game. We found that rolling the virtual camera and scaling the player’s motion were two of the primary inducers of nausea. We created a controller that when a player grabbed a climbing hold, their virtual hand would become locked to that position. When the player then moved their arm, they would be moving their whole character relative to that fixed position. 

Designing Dramatic Routes

Our game contained 4 levels, a tutorial, and 3 main levels, each with an increasing degree of difficulty. The tutorial served to introduce the climbing controls through climbing routes that were clearly defined. One of these routes created a situation where the player would be forced to launch themselves a short distance to get to the next hold. This helped players quickly establish a sense of how far they can reach between holds. 

The rest of the levels are relatively open, with multiple ways to reach your goal. The first of these open levels establishes a standard for what the climbable surfaces are like. There are no longer predefined holds on a wall, but rather large boulders that the player can grip anywhere. Additionally, the player now has to find their own way up. Once you start climbing, a primary path quickly becomes clear, but so do exciting deviations from the path. In the second level, the risk of falling is increased by the threat of having to restart a long climb. Additionally, a small puzzle is added, giving the player another element to engage with as they climb. The last level puts your climbing ability to the test by forcing you to speed up and optimize your climbing. If you spend too long on one rock, it will be blasted by an enemy, causing you to fall. 

Creating Meaningful Movement

During our early tests of the player controller, we happened to be using a physics-based solution. We found out that if you release a hold while swinging your arm downwards, you could launch yourself upwards almost like a jump. Players could also maintain their momentum and sling themselves to holds that previously seemed unreachable. As cool as this mechanic was, we made sure that performing this technique was never a necessity to complete a level. Since we only had four levels we wanted to make sure that all players could get through the game without getting stuck for large amounts of time. 

That being said, while the technique was not necessary, all of our levels were designed to be able to make use of that mechanic if the player so desired. All of our levels contain shortcuts that utilize this technique. For most players, the real fun of the game begins on their second playthrough, as they will have begun to get a feel for techniques and styles of play that they want to explore in more depth. Allowing for this sort of exploration enables the player to gain a strong sense of mastery over the game and makes it significantly more replayable. 

Grow Plants and Brew Potions in this Engine Building Board Game

In Alchemy Inc, you take on the role of a young Alchemist seeking to become the most prestigious potion maker in the city. Grow plants, buy runes, and craft spells to get the ingredients you need for the most popular potions, but be careful, the more of a single potion you make the less impressed people will be.

Creating a Unique Engine Building System 

In board gaming, engine building is a mechanic that allows players to set up a system where they can efficiently (or not so efficiently) turn one type of resource into another. In our game this manifests as the player designing spell scrolls in order to transmute their magical ingredients. One aspect of engine building that we really wanted to emphasize in our game was every player having their own unique method manipulating the ingredients. Instead of players cards that can change their ingredients in predefined ways, players in Alchemy Inc. buy individual runes and assemble them into spell scrolls. These custom spell scrolls manipulate the ingredients in different ways depending on the runes used to create them and with 24 unique base runes there are endless possibilities for players to assemble their alchemical engine.

Designing a Board Game Remotely 

The Covid-19 Pandemic forced us to reconsider the way we design and playtest games. Alchemy Inc. was developed fully remotely and we had to get creative with our tools in order to bring the game to life. Early on in the process, we used less traditional applications in the tabletop space such as google slides to playtest our early ideas. We later moved to Tabletop Simulator to produce a polished final product. Despite developing the techniques needed to remotely design and test games in response to the Pandemic, the skills learned proved invaluable even as we begin to move back towards normalcy. Tabletop Simulator allows me to produce high fidelity prototypes with custom parts that would take significant time to produce in real life. 

Agents as a Tool for Drawing

This project was an exercise in bringing an algorithmic design all the way from code to a physical drawing. My project specifically focused on creating drawings in collaboration with a group of flocking agents where the user manipulates parameters of the simulation in real time in order to get more engaging and dynamic results.

‍

Rethinking How We Work With Carbon Fiber

Tensionet is a prototype for an adaptive pavilion primarily made out of carbon fiber. This prototype proposes a new method of carbon fiber construction and explores some of the aesthetic possibilities of this new technique. Tensionet demonstrates a weaving pattern based on tensegrity structures that allows two linear elements to be woven together in almost any orientation, creating dynamic and unique structures. For our pavilion proposal we created a general form for the structure using wind speed data gathered at our site location. We then developed an agent based modeling solution to translate that general mesh into a constructible carbon fiber surface based on our weaving technique.

A New Type of Weaving

Our research started by studying Buckminster Fuller's Tensegrity structures as we were inspired by their seemingly gravity defying nature. The idea of having a compression member work in tandem with a tension member seemed like an ideal way to play to carbon fiber's strengths. We began exploring various ways to consistently connect two randomly oriented linear members together with a carbon fiber weave and found that doing so algorithmically was surprisingly simple. The trick was developing a system that could actually manufacture each of these unique weaves efficiently and with a high degree of accuracy.

We set out to develop a custom weaving jig that allowed us to place our two linear members anywhere inside a 24in x 24in x 12in volume. We then used an ABB IRB 120 to robotically weave carbon fiber between the two members. Parts that were needed in high quantities were specifically designed to be compatible with the standard jig so that we could stack them for a higher per-bake yield.

Generating structures Algorithmically 

In order to take advantage of freeform arrangement of linear members we had achieved in our weaving process, we needed to create a system that could loosely arrange the linear members into a desired massing. To do this we created an agent-based system where each agent has its own magnetic field that causes adjacent agents to align themselves in a perpendicular fashion. Additionally, a massing mesh serves as an additional attractor, the strength of which can be varied by mesh color. Over time, the agents are pulled close to the mesh and begin to encapsulate it, creating a textured approximation of the surface.

‍

Creating the Pavilion 

For our site, we chose a small clearing just south of the Fairmount Water Works Trail. This area acts as a natural funnel for wind coming off the river. It is also flanked by several large trees on either side and we decided to use these trees as anchor points for our structure. Using Kangaroo, we created a variety of meshes that were held in tension between the trees. In our selection of the final mesh we prioritized having both enclosed spaces and open canopies so that visitors would have a variety of spaces to inhabit.

After choosing our general massing, we subjected the mesh to a wind analysis using data on Philadelphia wind patterns that we gathered from Ladybug. This data will be used to control the density of our agents when aggregating over the mesh. Areas of the structure that are subject to higher wind loads will become denser while less impacted areas will be more open.

Using the data we gathered through the wind analysis, as well as other general structural analysis tests, we created a mesh that highlighted areas of strong tension and compression. This mesh then served as the massing mesh for our algorithmic assembly program where it was populated with 9,000 individual agents. The resulting structure is comprised of roughly 18,000 unique components.

Explore A Beautiful Papercraft World in this Adventure Game 

When designing the world of Mango, I took a lot of inspiration from the various Paper Mario titles. I had always loved the look of the games and the papercraft world seemed like the perfect fit for the playful premise of a 2D gingerbread man exploring a 3D world. Along with the influence, I also attempted to put my own spin on it. I made everything in the environment 3D including small props such as the flowers scattered throughout the world and designed unique folded paper structures such as the large mushrooms that are near the town square.

A New Role

For most of my previous projects, I had taken on the role of game designer and was responsible for building out and playtesting levels and puzzles. For Mango, however, our small team was short on dedicated artists so we had to work together to restructure our roles. Because I had some prior experience doing 3D modeling in my industrial design work, I volunteered to take on the job of an environmental artist.

Because the other artist on the project primarily worked in 2D, we worked together to plan out an artistic direction that would accommodate both our skill sets. Because we were loosely basing the game on the Gingerbread Man fairytale, we started with the idea of a children’s pop-up book. Over the course of the design process, we started to explore other ideas of papercraft ultimately resulting in the cardboard world of the final product.

Understanding the Context 

The very first thing I did was to return to my game design roots and block out the level. Before I started doing any sort of modeling, I wanted to understand the context in which these assets would be seen. In the initial design document, we had planned to have a level with a lot of depth so that the player would be moving forward and backward in the world just as much as they moved left to right. After some early testing, however, we found that that style of level conflicted with the desired camera angle. This was primarily because navigating forward (towards the camera) left the player feeling blind about what they were going to encounter. I attempted to resolve this in the level design by making the world tiered. Functionally, this just meant tying a series of more horizontal level segments together with stairs or ramps; however, even though the front-to-back moving was minimized, it still wasn't enjoyable. We also explored different possible camera angles that helped resolve the issue, but in the end, we decided we valued the original camera angle more than the forward and backward player movement and decided to go with a more horizontal level design.

Creating the Assets 

After understanding what the final level would look like, I began work on the assets. Coming into the role, I was fairly familiar with 3D modeling. While my primary job on other projects had been game design, it wasn’t uncommon for me to do some prop modeling as well because of my background in industrial design. 

For early modeling, I used SketchUp to get the ideas out of my head and into a rough visualization of what they would look like. After I felt satisfied with my 3D sketch, I moved Maya to create a higher-quality model that was properly UVed. From there, I went to Substance Painter for texturing. On this particular project, the most important detail was also the smallest. In order to get these models to really look like cardboard, I had to make sure that every exposed edge had a texture that mimicked the edge of a corrugated piece of cardboard. I had found through earlier tests that the texture on the colored faces was actually less important to the style than that revealing edge, so I prioritized preserving that feature in all my textures. Lastly, I imported the models into unity, assembled them into the final level, and lit them appropriately.

Achieving Bespoke Production With Robotic Manufacturing 

The use of advanced robotic manufacturing techniques gives designers and architects the ability to design architectural artifacts consisting of unique individual pieces that can be manufactured in a similar time frame to mass produced alternatives. In this project we explored a potential architectural application, a ceiling relief, that's design process could be greatly augmented through such manufacturing technologies. We created a 17 piece proof of concept where each piece is fully unique and produced through the use of an industrial robot arm. The final prototype was assembled and hung from the ceiling to serve as a glimpse into what a fully bespoke ceiling relief might look like.

Derived from Sculpture

For this project we were tasked with creating a bespoke ceiling relief out of styrofoam. We started by studying the work of Naum Gabo and Antoine Pevsner. In particular, we looked at Gabo’s series of Linear Constructions and Pevsner’s series of Developable Surfaces. Both of the series of work highlighted the extremes of what could be achieved by using ruled surfaces to construct geometry. We then translated some of these works into three dimensional volumes that could then be produced out of styrofoam using a 5-axis robotic arm equipped with a hot wire cutter.

After modeling small segments by hand, we fed our models into a specialized CNN to propagate the model into a texture that would become the basis of our ceiling relief. The output was then divided up into sections and toolpathed for cutting. 

Utilizing robotic fabrication  

In order to produce each piece of our prototype, we used a large 5-axis robotic arm equipped with a hot wire cutter. Each piece was individually tool pathed and consisted of 3 to 4 cuts. Each piece started as part of a 2’ x 2’ x 3’ block of styrofoam that was rough cut by the robot into closer approximations of each shape. That part is then registered to the robot using a custom jig in order to ensure a high degree of accuracy for each part. The robot then does a pass over each surface to cut the final shape. The parts were then mounted onto a wood frame in order to add our final lighting elements and to suspend the prototype from the ceiling.

Designing the Lighting Elements  

A key part of our initial relief design was the incorporation of lighting elements directly into the relief itself. We knew early on in the design process that we wanted to take advantage of the light diffusing properties of styrofoam in order to create a lighting effect that was strongly atmospheric but not harsh to viewers. The design of our relief featured recurring stalactite-like features, so we began to explore ways of lighting those features from the inside in order to achieve a glowing effect.

Through our experimentation, we found that in order to achieve the desired effect, the styrofoam could be no thicker than 0.5 inches and that the thickness needed to be completely uniform or smoothly gradated. This created a tricky production problem. Using the robot to cut uniform pieces that thin proved incredibly difficult as inaccuracies in the hot wire cutter became much more noticeable. Additionally, the pieces could not be hollowed out in post production as that would be both incredibly time consuming and yield a bumpy and uneven surface finish.

Taking these constraints into account, I developed a toolpath for the lighting elements that effectively ignored any inaccuracies produced by the hot wire cutter. I tapered the internal wall of the cut inward so that it would cross over the external wall near the top of the piece, the resulting wall thickness gradating from 0.5 inches thick to paper thin. I then would manually trim the top to the desired angle, resulting in our final piece.

Duke It Out at High Speed 

Blade is a fast paced 1v1 fighting game set in cyberpunk-esk future. You play as a Blademaster, a ninja-like android capable of incredible feats of speed and agility. Use abilities like a Grapple Shot and Wall Riding to traverse the battlefield and create the perfect opportunity to strike your opponent.

‍