Profiling spatial and temporal behaviour in sensor networks: a case study in energy monitoring

Wireless sensor networks (WSNs) provide a cost-effective platform for monitoring phenomena of interest at fine spatial and temporal resolutions. In this paper, we consider the application of monitoring power usage in an office environment at the resolution of individual users. A key challenge in this context is how to extract meaningful profiles of user behaviour in the large volume of monitoring data collected by the WSN. To manage the complexity of learning such profiles in this context, we propose a query based model for profiling. This query based model provides the ability to characterize the spatial and temporal occurrences of the power usage patterns of interest. We demonstrate the effectiveness of our query-based profiling model for finding relevant electricity usage patterns in a real life data set of power measurements collected by a WSN deployment in an office environment. To the best of our knowledge, this is the first time such a case study has been made on analysing the power usage of users at such a fine scale in an office environment.