Dissertation Finished
Posted April 21st
After a lot of research and thought, I have finally completed my dissertation. The title and abstract are included below. You can download a copy of it here.
Evaluating the current applications of Swarm Intelligence with regard to domestic usage.
by Ben Reynhart
Ants are extremely simple animals, if left alone most would struggle to survive. However a colony of ants is capable of achieving incredible feats such as finding the shortest route to food sources and keeping the nest temperature constant to within a degree. It is not just ants that work together for mutual benefit; many birds, fish and other animals flock together as a defense against predators. The study of this phenomenon and of systems sharing these properties is known as Swarm Intelligence.
The purpose of this dissertation is to investigate Swarm Intelligence (SI), paying particular attention to its current and future applications. To inform this goal, a description of Swarm Intelligence has been established, situating it within the larger field of Artificial Intelligence. Following this, it was necessary to analyze the useful methods and principles that can be extracted from natural swarms. It is then demonstrated how these underlying principles have been developed into two optimization algorithms; Ant Colony Optimization and Particle Swarm Optimization, which are utilized in a number of real-world applications. It was also important to look at the field of Swarm Robotics, which applies similar swarm principles to the field of robotics, allowing the swarm model to be realized in hardware as well as software.
Using case studies of current swarm applications not only indicates the current state of this technology, but also provide a useful source of inspiration for discussing the implementation of a swarm model in different environments, specifically domestic. These findings made it evident that a swarm-based hardware application in a domestic environment has had considerably less discussion and speculation than, for example, a military environment. My concluding argument is that a domestic environment could potentially benefit from a swarm-based hardware application, and it is not unreasonable to initiate discussion thereof.
