The continuing improvements in computer technology and algorithms, and the accompanying improvement in solving complex problems, tend to get a lot of attention. Just yesterday, for example, I posted a note here about a new algorithm for solving linear systems, and have written previously about IBM’s attempt to build software to play Jeopardy!. So it is probably salutary for us to be reminded occasionally that Nature has a few tricks up her sleeve.
According to some new research reported by the University of London, Queen Mary and Royal Holloway colleges, it turns out that ordinary bumblebees manage to solve a complex mathematics problem, even though they are hardly over-endowed in the brain department. When the bees forage, they initially come across desirable flowers more or less at random; but they quickly learn the shortest path that allows them to visit all the flowers, and then to return home. In effect, the bees are solving the Traveling Salesman problem, one of the most carefully studied problems in optimization. It is known to be a very complex problem, in the general case, and in fact to be NP-complete, meaning that the difficulty of computing a solution is like to increase exponentially with the size of the problem. Nonetheless, the bees have cracked it.
The Travelling Salesman must find the shortest route that allows him to visit all locations on his route. Computers solve it by comparing the length of all possible routes and choosing the shortest. However, bees solve it without computer assistance using a brain the size of grass seed.
The research is being published [abstract] in the journal American Naturalist.
Assuming that the results can be confirmed, and studied further, they might lead us to some new understanding of ways to address this class of problems — though that might not be very pleasing to our collective ego.