Learning from the birds and the bees (appeared on L'Express, Mauritius, September 2011)


Austin, the capital city of Texas, is much touted as the “the live music capital of the world”. But what really amazed me on a summer trip there is harmony of a different kind.  Imagine 1.5 million bats emerging from under a bridge at dusk within a half an hour window, on their way for dinner with 30,000 pounds of insects on the menu - a spectacle that attracts 100,000 every year. When Austin’s Congress Avenue Bridge was renovated in 1980, nobody suspected that the crevices in the structure would be ideal large-scale accommodation for a colony of pregnant Mexican Free-tailed Bats on a collective maternity stopover, giving birth before heading back together to colder Mexico.
 
Just like birds, bees, fishes and ants, bats have developed sophisticated social networks and communication systems that can put Facebook to shame, at least in terms of purposeful networking.  What enables such shared spirit, common purpose and distributed problem-solving capabilities? Birds and bees probably know little about combinatorial optimization, but the effectiveness of their self-organization intrigues many.  Such aggregate behavior of large populations, without any centralized control, is studied in a field known as “swarm intelligence”, where apparently very simple rules of action lead to emerging coherent functional patterns and global intelligence. For instance, ants, upon finding food leave a trail of a chemical (called pheromone) that others can follow. So a random food hunt is turned into a complex organization for targeted prospecting, where promising pathways are reinforced and lesser ones die out (when the pheromone gradually evaporates). The end result is amazingly effective and intelligent, and scientists are replicating such models to solve all kinds of problems – like reducing airline delays by improving taxiway traffic at busy airports, where a “colony” of pilots follow simple rules to efficiently produce globally effective queuing solutions. Bird flocking, foraging by honey bees, land animal herding, fish schooling, nest building by termites or wasps, and even bacterial growth are examples of swarm intelligence in nature.

Principles of swarm theory are being increasingly applied to various engineering, design and business problems. The team strategy of ants to detect food (and threats as well) is being used to protect computer networks from malicious programs that can delete data or steal passwords. Thousands of “digital ants” are made to wander through the cyber-world looking for threats, and leaving digital trails. This approach leads to an army of ants that converge to the spot you never suspected, just like in remote areas of your kitchen with food leftovers.

Swarm intelligence models learnt from our six-legged or flying bio-mates are also being used for routing information in telecommunication networks, ticketing and checking-in at airports, simulating crowd behavior, coordinating complex interactions among many robots in minesweeping or search and rescue missions. Computer graphic designers are crafting amazing displays based on the shapes generated by self-organized flocks of starlings or schools of fish. Scientists are also investigating swarm intelligence tactics to control a group of robots to explore remote Mars terrains. In medical nanotechnology, it now seems plausible that a collection of very small nano-robots can cooperate like animals do, to locate and destroy cancer tumors in the body.

Individual insects or animals react to their local environment, without any grasp of the global problem. Nonetheless they effectively influence each other to produce intelligent overall behavior. The sum is indeed greater than the parts. In human market economies, simple rules of self-centered actions of individuals help propel a complex, intelligent and collectively beneficial system, with the blessing of a wise invisible hand.  A key insight is that the absence of a centralized control means no centralized point of failure, and therefore a higher chance of success. Given the failures and dissatisfaction with so many of the centralized governance in human societies, maybe a swarm model is what we need – arguably the design that nature intended. Anyway, it is increasingly evident that the birds and the bees, long invoked to show children the ways of the world, have wisdom in store for adults too.

No comments:

Post a Comment