Social Insects

Social insects such as ants, bees, termites, and wasps, provide us with a powerful metaphor for creating decentralized problem-solving systems composed of simple, interacting, and often mobile, agents. The emergent collective intelligence of social insects lies not in complex individual capabilities but in networks of interactions that exist among individuals and between individuals and their colonies.

The daily problems solved by a social insect colony include the following: finding food, building or extending a nest, dividing labor among nest mates, feeding the brood, cleaning the nest, responding to external challenges, and spreading alarms. Many of these problems have counterparts in the business world. The collective behavior of social insects is not only decentralized, it is also flexible and robust. Flexibility allows adaptation to changing environments, while robustness enables the colony to function even though some individuals may fail to perform their tasks.

Swarm intelligence provides a basis to explore collective behavior to solve complex problems. Social insects serve as the inspiration for an alternative way of solving complex problems through their flexibility, robustness, decentralization, and self-organized behavior. This constitutes a basis for exploring the following different kinds of phenomena: how to achieve a collective task assignment without any centralized organization, how a collective representation emerges at a level of the colony without any symbolic representation at the individual level, and how to collectively build a complex organization without a plan. These collective behaviors are characterized by a logical pattern that emerges through simple individuals interacting with each other and their colony. This message is emphasized in the paper that complex collective behavior can emerge from individuals following simple rules (Bonabeau).

What evidence do the authors provide to support each conclusion? Aspects of social insects that make them appealing are that they are flexible, robust, decentralized, and self-organized. Social insects accomplish their work in a collective manner. An individual insect does its own work, but their work affects the entire colony. Individual efficiency and specialization in a particular field makes the whole colony greater than the sum of its parts. This idea is the foundation of "Swarm Intelligence" (Bonabeau).

One such example of swarm intelligence is bees. When food is scarce, the nurse bees will leave the comb and forage for food. Northwestern University devised a system, based on the bee model, for scheduling paint booths in a truck factory. The problem in the factory was that when the booths had to change color, the changes were time consuming and costly. The booths are thought of as bees governed by the following rule: an individual performs the tasks for which it is specialized unless it perceives an important need to perform another function. Although this rule sounds simplistic, in practice it is surprisingly effective. (Bonabeau).

Another useful model of work allocation comes from ants. Ants have several different types of workers in the colony. Each type has a particular and simple duty to accomplish (decentralized), such as the seed harvester who carries food back to the nest and the nest workers who maintain the nest. The interesting thing about them is that if the seed harvester ants disappear then some of the nest workers will change their activities and become seed harvester (flexible and robust). The reverse is also true. Seed harvester ants are like runners transferring a baton in a relay race, but the ants are not stationary, and their transfer points are not fixed. An ant carries the food down the chain until it reaches the next ant, and after transferring the food, it turns back until it meets the previous ant in the chain to receive the next load (Bonabeau). A system known as the "bucket brigade" was developed from this principle, dramatically increasing the efficiency of operations in which work is passed from one person to another. The University of Chicago has applied this approach to order pickers at a large distribution center of a major retail chain. The bucket brigade approach displays all three advantages of swarm intelligence. It allows a work line to balance itself. That is the optimum solution emerges without any intervention by mangers. The system is also flexible and robust, easily adapting to any unexpected surges in demand for particular products.

The foraging ant model has several applications that have been applied to the business world. Foraging ants are able to find the shortest path to a food source merely by laying and following a pheromone trail. This simple yet powerful approach has helped solve several business problems. Researchers at Hewlett Packer have developed a computer program based on foraging ant principles. The program is able to find the shortest path to route calls effectively. Another example is the variations on the ant foraging algorithm that have improved the efficiency of factories. Bios Group, a consulting firm, developed a program that will determine the shortest time to perform a set of jobs given various constraints. For example, if the changeover time for a piece of equipment might be close to an hour, the software avoids using that machinery for numerous, short tasks. Through digital pheromones, the software ants find the fastest schedule that satisfies all the constraints and thus ensures that every job is performed. (Bonabeau).

Details

...