A complex event processing based framework implementation for ambient intelligence

Abstract

Advances in electronics and wireless communications enable the production of small and wireless development boards with low energy consumption and various sensors. In this study, development boards containing electret microphone, light, temperature, motion, magnetic, optical and RFID (Radio Frequency Identification) sensors are used for collecting data in a classroom. These data are published to the topics created for every sensor in the publish-subscribe type communication software. A complex event processing (CEP) engine listening to these topics analyzes the data by using predefined rule sets and infers high-level information such as “lecture delayed”, “lecture started”, “break given”, and “lecture ended early”. CEP engine also offers some guidance information such as “open the curtains”, “turn on the air-conditioner”, and “open the windows”. The proposed system infers these by examining patterns, relationships and hierarchies within the data as well as applying time windowing, and then publishes these messages to the corresponding topics in the publish-subscibe messaging software. These messages are consumed by mobile phones and web sites for producing customized solutions to the end users.