HomeAlgorithms & AIAnt Colony Double-Bridge: Pheromone Phase-Space Map

Ant Colony Double-Bridge: Pheromone Phase-Space Map

Watch the Deneubourg double-bridge ant colony experiment as a 2D pheromone phase-space map: a live trajectory of (short-branch, long-branch) pheromone levels over a colored choice-probability field, instead of a 3D diorama of ants on bridges. Tune length ratio, evaporation rate and choice nonlinearity and see the trajectory get pulled toward one axis by pure positive feedback.

Algorithms & AI2DModerate60 FPS📱 Mobile-adapted⇄ 3D version
2d-ant-colony-pheromone-trail-optimization ↗ Open standalone

A colony of ants connects its nest to a food source over two bridges of different length, and no single ant ever measures distance or compares the branches — each simply deposits evaporating pheromone on whichever branch it just walked and, at the next fork, chooses probabilistically toward whichever branch already smells stronger. Instead of watching that unfold as a diorama of ants on bridges, this simulator plots the colony's entire collective state directly as a point in pheromone phase space — the short-branch pheromone level on one axis, the long-branch level on the other — moving across a live-colored choice-probability field. Because trips on the short branch finish faster, that branch accumulates reinforcement faster too, and you can watch the trajectory get pulled off the neutral diagonal and driven into ever more lopsided territory by nothing but positive feedback. Adjust the length ratio, evaporation rate and the nonlinearity of the choice rule to see when the trajectory converges cleanly toward the short-branch axis, when it wanders and stays near the diagonal, and when it can even lock onto the long-branch axis by chance — the same dynamics that inspired Ant Colony Optimization as a search algorithm.

⚙ Under the hood

Watch the Deneubourg double-bridge ant colony experiment as a 2D pheromone phase-space map: a live trajectory of (short-branch, long-branch) pheromone levels traced over a colored choice-probability field, instead of a 3D diorama of ants on bridges. Tune the length ratio, evaporation rate and choice nonlinearity and watch the trajectory get pulled off the neutral diagonal toward one axis by pure positive feedback.

ant colony optimizationpheromone trailstigmergyswarm intelligencepositive feedbackphase spacemetaheuristic

2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install

What did you find?

Add reproduction steps (optional)