Organic Insecticides | Organic Pesticides | EcoSMART Bug Spray

An Inspirational Ant Colony

November 4th, 2009 Posted in Ants & Roaches

Scientists use algorithms to solve problems with computers. Inspired by actual ant behavior, scientists have developed a useful problem solving algorithm.

Called the ant colony optimization algorithm, it mimics the swarm intelligence ants use everyday. It is the most successful algorithmic technique based on ants to this day.

This article will discuss:

• Ant Behavior
• Ant Based Algorithm
• Inconsistencies

Ant Behavior

Ants instinctively know the shortest path between food and the colony. They know to lay pheromones along the path leading to food to help out their fellow ant. They do this without the help of any visual cues, thus no road signs advertising McDonald’s or 7-Eleven.

Ants are also very adaptable to circumstances in their environment. If a path they previously laid out becomes obstructed, they’ll just find a new more accessible route. The ant marks this new path and helps to make this path for other ants successful.

The pheromones they lay on this new path will be stronger and fresher since its new. Ants know instinctively to follow a path with a stronger scent than one with a weaker scent. Once a few ants follow this new path around an obstacle, it becomes positively reinforced the more pheromones on the path.

This all leads to more and more ants following this more efficient and successful path. This kind of problem solving is exactly what scientists are trying to emulate with their algorithm.

Ant Based Algorithm

First developed and introduced by researcher Marco Dorigo in 1992, it began the swarm intelligence research field. Applying the swarm intelligence behavior to algorithm techniques has helped researchers solve more complex problems than ever before.

By observing behaviors in an ant colony, researchers have applied the problem solving technique to help with everything from folding protein to routing vehicles. One problem also solved was the traveling salesman problem. The most common problem in optimization, the traveling salesman problem is finding the shortest distance between a list of cities making sure to visit each city once.

Optimization is finding the most efficient and cost-effective way to do something. Thus the ability to find the shortest paths is what computer scientists are trying to accomplish when they optimize. Optimizing decision making is the main goal of this field of research, but mostly in terms of procedural programs and software on the computer.

By developing software with any variation of the ant colony optimization algorithm, the developer is making a program that is more efficient and successful in completing its objective.

Inconsistencies

The behavior of real ants is hard to completely duplicate because ants are alive and a computer is not, well obviously. Biological cues motivate ants to participate in their social swarm behaviors, while a computer can only merely imitate such actions through programming.

Science today has come pretty close to duplicating the actions of nature, but isn’t as 100% successful as nature can be. Nature isn’t always perfect though, while the actions of a machine tend to run smoothly once successfully programmed. For instance, when an ant first searches for food, it looks at random, not following some efficient optimized path, but kind of just winging it.

Once the ant does stumble upon food is when they make and follow a pheromone path. This aspect of the ant’s behavior is unpredictable and almost impossible for a computer to try and imitate

Nonetheless, the ant colony optimization algorithm has successfully taken successful parts of ant behavior and applied it to real world problems affecting humans. Thus ants and their tiny existence have taught us a thing or two about decision making and problem solving.