Enabling Virtual, Cloud-Based Sensors in Assisted Living Environments

Abstract

In this work, we combine the field of cloud computing with assisted living to utilize an improved activity classification. The proposed system addresses the challenge of enabling heterogeneous, cloud-based sensors in small home environments. With a semantic and model-based approach, we seamlessly integrate Web services as virtual sensors and, therefore, increasing the accuracy of human activity classifiers. We solve the most important compelling issue of interoperability, by using generated wrapper classes and semantically unify all heterogeneous sensor events within the system. Our approach is evaluated with a test installation by means of the requirements for Sensor Web Infrastructures combined with the one for assisted living environments.