I just finished my thesis week. My thesis project is a style transfer application which is using machine learning. At ITP machine learning is one of the hottest topics, but still AR & VR are the most popular topic at ITP.
I haven’t checked all of them, but I could find out 2 interest things from my friends’ great projects.
ARKit + Location → StoryTelling tool
In terms of AR, many students are using ARKit which is from Apple and some of them are using location data. The reason why people use location data might be that ARKit still doesn’t allow us to store the position data. Friends’ application are working well which is great. However, what I thought was Zach Lieberman’s Weird Type. When he came to ITP last year, he mentioned to the limitation of ARKit. Especially, about how to store object data. I think there is no way he didn’t think about CoreLocation which is a location library for iOS, but he didn’t mention it. From his response on Github, I guess he cares about privacy, so that the reason why he doesn’t use location data for storing data. In addition he may think about AR’s spatial issue which was caused by Snapchat before.
When I saw friends’ apps, I thought location base is the easiest way to manipulate data with ARKit, but really need to think about the problem which may happen to users and operators. For example, someone will puts something unacceptable such as porn, bad words etc. Current my idea of user created content app is that it’s good as a story telling tool, but creator needs to keep everything safe for uses. So, that kind of apps might be expensive for individual creator/developer.
At the same time, what I thought was that if a company/individual could make an open platform/protocol/agreement for AR with location data, it would be great and a huge business chance. Like Linden Lab for AR.
AR + Physical Computing
There is another interesting type.
These students are using ARKit with Unity for lighting. They showed a great possibility of AR + physical computing. The first application’s device is using Arduino to receive data from the application. In her case, using MKR1000 which isn’t cheap as a micro-controller, but this architect could be a poplar medium. The thing might be security, but as a prototyping it would be great.
The second one is using an existing light bulb like Philips Hue. His application is great because that could add a new fun function to existing light bulbs. I used to used this kind of light bulbs for my class projects and what I understood was basically official applications aren’t appreciated by users. His approach could bring a positive effect to existing apps.
I think I have to think about AR and research on it for my future project lol.