Maximising effectiveness of distributed mobile observation systems in dynamic situations

The trend in modern day observation systems is towards distributed (often mobile) systems that are able to automatically adapt themselves in dynamic situations. They have to make most of their resources to maximise the system's effectiveness, all at reasonable cost. Currently, there is no formal framework of reasoning to design and implement the mechanisms for the required adaptivity. In this paper, utility based reasoning is combined with the Markov Decision making Process (MDP) formalism to come to such an approach. Its ramifications are discussed, followed by several cases in which this approach has been applied successfully.