AR Prototype Game

29/10/2025

During my time at PhantomTech, we were toying with the idea of creating a collection of AR game experiences to showcase the capabilities of the tech we were developing. The idea was a simple FPS base-building, tower-defence game where you would fend off waves of zombies trying to reach you and destroy your base. Little came of it in the end since the company decided to shift focus away from building games, but I did manage to put together a small prototype that shows off some of the work we'd done.

To make this happen, I had to pull together a lot of the systems we had been working on. AR takes a toll on the processing power of the device. While you unfortunately can't see this happening in the video, the device is actively scanning the world and creating a physics mesh the game can interact with. The game needed to run smoothly while also handling all of the processing required to scan, mesh, and generate collision data for the world in real time. I then had to build the game within whatever performance budget was left.

The game concept was simple: place defences to keep yourself alive against hordes of enemies swarming you in AR. Wave-based enemies are a useful way to stress test a game system, as you can keep pushing the limits of how many enemies can exist at one time. With a basic setup, we were able to get hundreds of enemies spawned at once and get a real look at what an AR game could be.

While it may have been a simple whitebox prototype, it served its purpose: demonstrating what kind of processing power we had to work with, what kinds of game experiences we could create, and how they might feel to play.

I would've loved to develop this further, fleshing out the game mechanics and incorporating other technologies we had available, such as our occlusion shader. One thing was pretty clear at the time, though: we first needed to improve the performance of our meshing technology. It was promising, but not quite ready for the kind of game we wanted to build.

---