AI pair research project implemented as part of a course where I learned other AI techniques such as state machines, behavior trees, steering behaviors, A* pathfinding, or terrain analysis. In this project we explore a technique called "stigmergy", whereby agents in a group communicate by modifying their environment.
Ants do this by leaving behind trails of chemical substances called "pheromones". They're able to sense these, and distinguish between the ones left by ants who just left the nest (blue in our demo) or ants that are carrying food (green in our demo). This means they'll follow the general direction where they sense the highest concentration of pheromones of the type they're interested in, while leaving pheromones themselves. Eventually ants will form paths to/from food sources and the nest, and these will get optimized over time to be as short as possible.
For implementation details, keep reading. However, for a more detailed explanation of how the project works, or some of the existing research we based our implementation on, check out the research paper we wrote, here.