Adaptive multiagent algorithms based upon the behaviour of social insects are powerful decentralised systems capable of solving complex problems. The intelligence of such a system lies not within a single agent but is a product of a network of simple interactions. However, there is no standard way to evaluate the performance of adaptive algorithms due to their dynamic behaviour. This project proposes that such algorithms can be assessed by extensively comparing their performance on an identical problem. Under the context of a mail company environment different artificial intelligent techniques are implemented and evaluated. The project creates a number of strategies to tackle this environment based upon the principles of self-organisation, Markov chains and greedy algorithms. The project also investigates factors that may affect their performance.