Harnessing the power of the ant for technology
If you're looking for the insect that proves most upsetting to the human race, you could do a lot worse than the humble ant. Its tendency to disrupt picnics is particularly irksome given that ants make up around 15% of terrestrial biomass. It apparently annoys senators who are upset with the stimulus money that has been spent on researching our six legged friend. However, the ant could also be just the inspiration we need to solve complex computing problems.
Researchers at Sweden's Uppsala University, in conjunction with academics at the University of Sydney, have discovered that ants are remarkably good at mathematics. A paper published in the Journal of Experimental Biology describes how Argentine ants were made to solve dynamic optimization problems by playing the classic Towers of Hanoi math puzzle.
The ants were made to choose the shortest route between points in a maze to get to a food source on the other side. When the layout of the maze was changed, the ants once again solved the problem by finding the shortest path. It was previously thought that ants, which find the shortest path between two points by laying trails of pheromones, would get stuck in a loop when the walls were changed.
This development has promising implications for the use of software in optimization problems, and it shows how we can draw on algorithms perfected over millions of years in nature to enhance our own ways of doing things in software. Ant Colony Optimization algorithms are already used in network management, but they tend to be used to model static problems. This research shows that ants can adapt quickly and effectively to new parameters. If we can train our software to think in the same way that ants do, we might be able to produce better dynamic network modeling software.
There have been other examples of how software can be adapted using techniques found in nature. For example, scientists have experimented with programming security systems dealing with tasks such as intrusion detection to operate along the same lines as biological immune systems.
Because we still understand relatively little about how nature works, we are still constrained to cursory explorations of its behavior. Hopefully over time, we will become more adept at mirroring its approach to complex problems. Perhaps The Simpsons' Kent Brockman was right: it is time we all welcomed our insect overlords.