For the final project I will work with Bishka Zareen to replicate swarm behavior exhibited by ants. By focusing on ants’ foraging and feeding techniques we found that ant’s a relatively predictable collective behavior. As they search for food they leave a trail of pheromones that only increase as more ants follow the same path.
An interesting approach would be to see how the behavior of the ants differ according to the initial set up. For example, how would the reaction time and completion rate differ if the ants were spread out within the environment or sectioned together at the beginning?
First, we must procure our “ant”. By using arduino we can create a bot that is attracted to a fixed point. This point will act as a singularity where beings surrounding it will be drawn towards it.
An issue I anticipate running into is organizing the ants around the beacon. This type of coordination will require them to have some sense of the individuals around them as well as being able to gage where the beacon is. Bots who find themselves slotted behind an individual must reorient themselves in order to not have any obstacles between them and the beacon.
A solution to this issue could be the use of infrared sensors. By having the sensors act as locators for the beacon and ants we can have them not only recognize the positioning of the beacon but also recognize the position of their fellow ants.
This simple behavior can be seen when ants are given an obstacle that breaks the scent trail left for them. They collectively work their way around the object while laying down a new trail of pheromones so that trailing ants can follow the new path.
A stretch goal would be to have the ants optimize their trails. In order to achieve this we must procure or develop an optimization algorithm for our bots.
Timeline
5/7 Write Introduction & Materials section
5/8 Prototype & build working “Ant”
5/9 Build & Test
5/10 Test
5/11 Preliminary Findings
5/12 Deadline: Draft I
5/13 Test
5/14 Refine Paper
5/15 Refine Paper
5/16 Deadline: Presentation & Final Draft
5/23 Deadline: Documentation