For this assignment I implemented three ant types on a shared pheromone grid. The first type is the canonical Langton behavior: if it finds pheromone, it turns 90° one way and removes the mark; otherwise it turns 90° the opposite way and leaves a mark. The second type is a diagonal variant that uses 45° turns instead of 90°, which makes curved, braided paths that feel less orthogonal and more wave-like. The third type uses a timer-reset rule: when it finds pheromone it clears the cell and resets an internal countdown; when it does not find pheromone it leaves one, decrements the countdown, and only turns after the timer expires. This creates delayed pivots and long straight runs punctuated by sudden corners.
Mixing all three types in one field generated richer structures than running any type alone. The 90° ants quickly scaffolded rigid lanes, while the 45° ants crossed those lanes with offset spirals and produced moiré-like seams. The timer-reset ants periodically “lock in” to corridors, then break out and cut across dense regions, which causes repeated cycles of ordering and disruption. Over longer runs I noticed that some areas stabilize into striped highways while nearby regions remain turbulent, almost like phase separation. Visually this gave me a useful contrast between geometric regularity and noisy mutation, which is exactly the emergent behavior I wanted to show.
Project: Run A6 vants simulation