Sensor data fusion for activity monitoring in ambient assisted living environments

We illustrate the PERSONA context-awareness framework applied to a major problem in Ambient Intelligence, namely user activity monitoring, that requires to infer new knowledge from collected and fused sensor data, dealing with highly dynamic environments where devices continuously change their availability and (or) physical location. We describe the Sensor Abstraction and Integration Layer (SAIL), we introduce the Human Posture Classification component, which is one particular context information provider, and finally we describe the Activity Monitor, which is a reasoner that delivers aggregated/derived context events in terms of the context ontology.