Skip to 0 minutes and 10 secondsHere we have the ant simulation model of Goss. What we see here is down below, the nest where the ants live. And on top here is food to be found. Now the ants can move to the food, but they encounter a difficult decision. Should they take a left turn, or should they take a right turn? Well, initially they do that by chance, and here the chance of going left or right is 50%, just as over here. And also if they return, they face the same 50-50 choice. But remember that we were thinking about communication through pheromones. What the simulated agents do is, wherever they go, they leave a trail of so-called pheromone.
Skip to 1 minute and 9 secondsActually they influence the chance of other simulated ants to go left or right. So the more ants follow the shorter route, the stronger the pheromone trail will be, and the larger the chance that simulated ants will follow the same route. What do you expect will happen here? What will grow out of this simulation? I Think a moment. Do you think there will be a kind of interesting emergent phenomenon here? Well, let's look and start the simulation.
Skip to 1 minute and 53 secondsHere they go, and in the beginning you see there's a 50-50 distribution of going to the left, to the right. You also see green expressing the concentration of pheromone. Now watch carefully. The ants that took the short route are the first to return. As a consequence, the pheromone trail on the short route will be thicker, attracting more simulated ants. And we see a process of self-organization. We see self-reinforcing process happening, and as a collective, the ants follow the short route. But still there are some ants choosing the longer path, and of course that's important because perhaps in the future, the food is finished and they have to explore new routes to find new sources of food.
Agent based model 1: Pheromone communication in ants
This video guides you through the simulation model about ants that communicate through pheromones.
© University of Groningen