Radar Prototyping

The image came first. I was standing in the shower trying not to get soap in my eyes when I saw myself in front of a radar screen. Around me was a control station embedded with different screens and equipment. Beyond the desk, a vast rotating radar antenna that's broken and needs me to drive out to it with a toolbox. Right away I knew what I needed to do.

I dried myself about as well as one could hope for, threw on some shorts, and wrote a message to a friend. I held him hostage while I turned on my screen sharing and started doodling my vision. He was silent and patiently waited for me to explain everything. I struggled expressing the details I could see in my head, but this image was enough to share the concept.

It only took moments before we were both eagerly bouncing ideas back and forth. My friend (who you'll likely meet later) and I worked closely together through the weekend. We knew the core mechanic we needed to focus on was the radar itself: how it works, what it looks like, how we control it, what it detects, what the player's objective is, what the game loop would be, and how we'd architect everything. After a week or so we had some ideas that were good enough to try prototyping.

We started with the basics of the radar. It's essentially a beam that points outward from the radar's face and rotates around its origin. This acted like our central vector (Pink). Then we created 4 more beams that act as the Left/Right (Red) and Up/Down (Green). This gives us a window of visibility.

Now because we're making a video game, we're thinking about performance. In almost all cases, the player would likely never see an aircraft flying in the air. So why render one? We'll go full data mode, which means the ceiling to what we can simulate can be whatever we feel is appropriate.

So lets take a peek into what I mean by simulating our radar data. The image above is essentially just for show. Yes we will have a physical radar array in the game the player can (and will need to) interact with, but our simulation only needs to be represented in terms of code and numbers underneath everything. As you can see below, I establish a basic set of rules that introduces "entities" (aircraft) into the world. We have a "World State" class that exclusively controls the behavior of these aircraft. Our radar, separately, simply rotates every frame and signals an event whenever one of those aircraft were detected within our radar beams. And after some rough UI work, we can see it happen in action:

Both of us knew going in that this strategy would be too simple, but it keeps things moving and lets us have a starting point to build the rest of the game off of. Future goals will be to treat the radar more realistically with the pulsing of the player's frequency, simulating the return frequencies and sending those signals into a "toolchain pipeline" we're currently calling it, but those details will get explained in more detail later. Future posts will cover more details on the world simulation and the player's interaction with it, the game loop, technical radar details, and a lot more.