The Emergent Brain

Connecting dots for a global brain.

Posts Tagged ‘Swarm Intelligence

Swarm intelligence for preferred outcomes 1

leave a comment »

If some human social system movements can be 93% predictable, animal social systems and their movement might help in even better and faster predicting certain real-world patterns and outcomes. A whole new branch of research has emerged around decoding and using swarm systems and intelligence in the last two decades. The Economist has a nice story giving an overview on this development focusing on insect, bird and robot behavior.

This basic analysis helps in understanding special decision making processes that are of short- to medium-term focus and can thus be determined by individuals themselves. Goals include route optimization, choice maximization and escape and emergency planning, in short: Movement patterns. Models based on swarm intelligence mentioned in the Economist article include such programs as Ant Colony Optimization and Particle Swarm Optimization that focus on self-reinforcing behavior and choice (some further models can be found in the Wikipedia). Modeling these kinds of behavior into algorithms seems quite easy, as they follow few rules. An older TED Talk by Steven Strogatz highlights the ease and power of synchronization in swarms based on only four rules – follow your nearest neighbor, line up, be attracted and move out in case of danger:

“Mathematician Steven Strogatz shows how flocks of creatures (like birds, fireflies and fish) manage to synchronize and act as a unit — when no one’s giving orders. The powerful tendency extends into the realm of objects, too.”

Watch the video here or below:

Thus, already in the general day-to-day behavior we can see a predictable pattern emerging based on few rules. Part 2 on swarm intelligence will cover higher, long-term cooperative systems of social organization that include generations and personal life roles.

Written by emergentbrain

22/09/2010 at 10:52