Are spiders relatively intelligent

Swarm intelligence: how ants find the shortest route

Solution without central control

Ants don't need a leader to tell you where to go, but rather special rules to follow. The swarm organizes itself. Researchers want to learn from the ants in order to calculate routes or solve difficult tasks.

Scientists have long been convinced that the behavior of animals in swarms has something to do with supernatural perception.

They can now explain the behavior of a swarm scientifically: If the members of a group react together to changes in their environment, rules of interaction develop. Researchers say that a complex, adaptable system is emerging.

In his book "Swarm Intelligence" (2009), the Australian physicist and swarm researcher Len Fisher illustrated this with the applause during a concert. If a viewer starts to clap, he can get many others to clap - until finally the whole audience applauds.

Fisher draws the following conclusion: The forces that emanate from the individual are not linear. The individual can therefore exert a disproportionately large influence on the group.

If concert-goers fall into a certain rhythm while clapping, this is a spontaneous reaction from the audience - and not from the individual audience. Researchers see this as a fundamental property of the swarm: a group of individuals solves a task without central control that a single group member could not manage.

Mathematicians take the ants as a model

There are an estimated ten quadrillion ants on earth. Many of them live in colonies with several million animals. In their search for food, the ants follow a certain principle: They always try to take the shortest route to the food source.

To find it, scouts examine the area around the nest. On their search, they leave behind a scent - a pheromone - to mark the route.

The scout who has found the shortest route to the meal is the first to return to the nest. On the way back he marked the route again. The other ants use this double scent trail to orient themselves later in order to also get to the food.

Mathematicians use the ants' swarm intelligence, for example, when they want to determine the shortest distance between several points.

As long as the number of route points remains clear, the best travel time can be determined relatively easily using a computer. However, this purely computational method becomes more complex the more coordinates are added.

Ants examine Odysseus' journey home from Troy

Fisher gives the following example: Odysseus traveled to 16 islands on his way home from Troy to Ithaca. What is the shortest connection? To get the answer, one would have to compare billions of routes (653,837,184,000). To solve the riddle, researchers use the ants as a model.

Using computer simulations, they send virtual ant scouts on a round trip to all 16 islands. When a scout returns to the starting point, he will indicate the distance that he has traveled. Each section of the route is rated: the shorter the trip, the higher the score in the ranking.

As in the real world, the virtual ants that follow now increasingly use the route with the best ranking. The short routes receive proportionally more points than the long ones. In this way, the ant algorithms provide the best possible route without the computer having to carry out all the arithmetic operations that would otherwise be necessary.

Imitate better solutions

The US researchers Russell Eberhart and Jim Kennedy take the idea of ​​ant algorithms one step further. Their computer program works in a similar way to "an exam where cheating is allowed" (Len Fisher).

First, each ant works out the best solution for itself. Then the virtual animals compare their results and exchange the best solutions with one another. In this way they learn from each other and together they come to the best result.

The research by Eberhart and Kennedy is also relevant for people in everyday life: If someone has found a better solution than we have, it is best to adopt it.