Saturday Science and Math Academy (118)/Proposal

Proposal
http://mode.lanl.k12.nm.us/get_proposal1112.php?team_id=118 Solving real world problems using ant-based algorithms

The purpose of our project will be to use ant-based models to study and analyze the effectiveness of different decentralized systems. We will use the Netlogo programming language to model how an ant colony collects food. This model will incorporate knowledge of how existing ant colonies operate. We intend to program two different behaviors seen within ant colonies: density recruitment and site fidelity. Density recruitment behavior involves the drawing of pheromone trails, which act as indicators to other ants. These pheromones are drawn based on the density of surrounding food. Site fidelity, in contrast, is a behavior where ants do not recruit others, but return to the location of previously found food.

We plan to program three different behavior sets. Sliders will control individual aspects of each behavior, one set will operate on a site fidelity behavior, and one will use density recruitment. We plan to record the average rate of food collection and the time needed for all the available food to be collected.

This information will then be used to analyze real world problems. For example, ant colony recruitment operates in a similar manner to the human immune system. The model of a decentralized system is also is seen in the use of networking. Instead of managing data as one controlled structure, it is split into packets of information. Additionally, this behavior may be implemented through the concept of bomb defusing and other robotics concepts. One large centralized robot may be more capable than many smaller robots, but the small robots are capable of working together to produce a more effective result.