Hidden Markov model as a framework for situational awareness

In this paper we present a hidden Markov model (HMM) based framework for situational awareness that utilizes multi-sensor multiple modality data. Situational awareness is a process that comes to a conclusion based on the events that take place over a period of time across a wide area. We show that each state in the HMM is an event that leads to a situation and the transition from one state to another is determined based on the probability of detection of certain events using multiple sensors of multiple modalities - thereby using sensor fusion for situational awareness. We show the construction of HMM and apply it to the data collected using a suite of sensors on a Packbot.