The topic of my thesis is “Hybrid development for mobile augmented reality”. That old chestnut, amirite? If you really want to know, hybrid development is a form of cross-platform software development, where you create only a web application and then wrap it into Android and iOS containers, effectively creating apps for three target platforms.

This method of developing software has proven effective for regular apps, but I haven’t found any project using AR in such a way. That’s why I set out to prove that it is possible. I kinda failed.

Kinda as in, I got it to work on Android and web browsers, but not on iOS. The main problem is that the AR frameworks I tested (AR.js, jsartoolkit, A-Frame, to name a few) use specific browser APIs, which are not available (or don’t work properly) in the iOS WebView used for hybrid apps.

Along the way, I also created a framework for displaying mathematical objects in 3D, a web-based visual editor for AR and hybrid, and a tutorial guide for developers wishing to try this workflow. All that and more can be found on my thesis website.

And if you want to go really in-depth on the topic, you can find my thesis here.

We also got to connect with the principal of St. Pölten Business High School, who we are now working with to set up a pilot test of our math learning app. This event was a huge boost for us and now we’re even more excited for the future of the startup!

The idea we pitched is ARchimedes – an interactive math learning app, which features exercises for all different learning styles, and contents taken from real-world examples. We started working on the idea at the beginning of the year, but changed the business model a lot with the information we learned during our summer school programs.

ARchimedes will also heavily rely on AR and VR exercises, in order to make math more tangible and exciting. To showcase this, we created a small demo scene on exponential growth for the live pitch, which we then showcased in AR. Of course the scene contained an exponentially growing amount of gold coins, to show what would happen if you invested in our startup.

Bein accepted into the incubator means we will now get more support through the next year, which means you’ll definitely get to read more from this project.

We developed a solution which utilizes a regular DSLR camera to automatically take photos of a finished part and transfer those to a computer. This computer uses OpenCV and a few custom-written algorithms to measure the lengths and angles of the edges and determine whether the values are within the error margin. Also, sorry for the edgy puns.

This is how it turned out: