Converging Redundant Sensor Network Information for Improved Building Control

This project is investigating the development and application of sensor networks to enhance building energy management and security. Commercial, industrial and residential buildings often incorporate systems used to determine occupancy, but current sensor technology and control algorithms limit the effectiveness of these systems. For example, most of these systems rely on single monitoring points to detect occupancy, when more than one monitoring point would improve system performance. Phase I of the project focused on instrumentation and data collection. In Phase I, a new occupancy detection system was developed, commissioned and installed in a sample of private offices and open-plan office workstations. Data acquisition systems were developed and deployed to collect data on space occupancy profiles. In phase II of the project, described in this report, we demonstrate that a network of several sensors provides a more accurate measure of occupancy than is possible using systems based on single monitoring points. We also establish that analysis algorithms can be applied to the sensor network data stream to improve the accuracy of system performance in energy management and security applications, and show that it may be possible to use sensor network pulse rate to distinguish the number of occupants in a space. Finally, in this phase of the project we also developed a prototype web-based display that portrays the current status of each detector in a sensor network monitoring building occupancy. This basic capability will be extended in the future by applying an algorithm-based inference to the sensor network data stream, so that the web page displays the likelihood that each monitored office or area is occupied, as a supplement to the actual status of each sensor