Squads AI is, as the name implies, an AI system based on how squads operate. The most common comparison is a spec. ops. swat squad. This is what my AI is inspired by.
Research and testing
My Squads AI is an ongoing passion project. I want to create a modular AI system which has senses similar to a human being. These senses give the AI certain observatory capabilities. For example vision allows the AI to spot things within it's vision cone. Hearing allows the AI to hear sound through walls (albeit dampened) as well as right in front of it. The combination of both allows the AI to assess certain sounds and visuals as a threat or not. This is just an example that is possible with my current systems in place.
I want to expand by adding a more thorough investigation system through which the squad will communicate with each other.
Once my squad AI is fully functional they should be able to enter an area, investigate it, and clear it. This is similar to a real Swat Squad. They will cover each other from flanking points and will be able to disable threats once found. They will have to work together.
My final testing case for this AI will be a battle royale on a small map in which multiple squads will try to strategically eliminate each other.
My squads AI makes use of sensory modules. In this test you can see my AI utilizing the hearing Module in combination with the Vision module. The source of the sound is obstructed, but the wall in front of the AI doesn't seem to be too wide so it debunks through the sphere where the sound must be coming from. The smaller the size of the sphere, the more confident the AI is.
The squad has investigation capabilities through which the squad moves from objective to objective to determine whether it is a threat, a point of interest or something to ignore. Each node will be flagged so that the other AI will know what it is.