The initial project goal was to create a prototype of a choose-your-own-adventure book with an AR experience. Overall, the project went well, and we met our goal – albeit slightly scaled down from our initial aspirations.

​

Our team (Daisy Chiu, Phil Kinshuck and me) was able to quickly align on our goal for the project and develop a cohesive project plan. Within one week of starting – and as predicted by our professors – we determined our initial scope was too ambitious for the timeframe and pruned accordingly. We are still pleased with the scope of the ultimate product.

​

The team meshed complementary skills, with Daisy as art director, UI art and creating the physical book, Phil taking on building terrain, modeling, animations and animations code, and me tackling project management and coding story progression and user interface.

​

As our team dynamics were strong, the largest challenge for me was figuring out what system to use for user interaction. I was tasked with looking into finger tracking versus image tracking. I spent over a week trying to figure out how to get different APIs to work. Unfortunately, the APIs were either out of date or incomplete to work. I struggled to get anything to work with Unity and the mobile AR API, and since we were on a tight schedule, we as a team decided to use image tracking as our way of interacting with the environment. I then took over programming of image tracking from Daisy. She had put in basic image tracking but was having difficulty getting the Unity collision system to work. I worked with her to get a handle on how the process worked and fixed the collision issues. I then set up a system that would use collision points on the page environment to work with the physical object images. This worked well when we used a flat plane for our environment. The items moved nicely over the surface and collided. I added collision points to the plane and set the system up so it can easily be extended to more pages. After we put the Unity terrain system in, the project had issues keeping track of the physical objects. While the terrain looks good, we need to find out why the physical objects being tracked seem to be under the terrain. I will need to do more research to fix this issue.

​

Daisy made the UI art and worked closely with me to get it in and functional. The UI was complicated as we had to tie in the collision system to turn notifications to the reader based on which object was being used. I added an array system to keep track of what types of collisions were happening and what objects were colliding to simplify the UI system. Once this was complete, the team was able to quickly hook up the models, animation and terrain and we got the program working well in the last few days.

​

The prototype effectively shows how a choose-your-own-adventure book with AR application could work. Given more time, we would sort out the terrain issue and polish things a bit more. I could see this project becoming an actual product if we decide to pursue it in the future.