An Agent Based Model for Health Surveillance Systems and Early Biological Threat Detection

Recent advances in biological sciences have made hostile biological attacks increasingly available to terrorist groups. The success of these threats is based on the assumption that biological attacks are invisible and will not be rapidly recognized as deliberate attacks. Current epidemic surveillance systems use data available from sources such as hospitals, clinics, and medical laboratories to analyze epidemic trends over time and are not capable of detecting such threats in a timely manner. In this paper, we present a multi-agent based system for early biological threat detection. At the micro-level, monitored humans are equipped with personal agents that are responsible for capturing the data transmitted by the wearable sensors, performing basic processing and analysis, and transmitting the collected data to higher-level agents for further processing. At the macro-level, a hierarchy of specialized agents are responsible for collecting, analyzing the transmitted data, and rapidly detecting possible epidemic threats. The experimental results show that the proposed model is able to effectively detect and localize epidemic threats with a various number of simulated humans.