Toward Efficient Resource Management in Buildings
Oct 01, 2010
from 10:00 AM to 11:00 AM
|Where||Engr IV Room 57-124|
|Add event to calendar||
Advisor: Mani B. Srivastava
Friday, October 1, 2010 at 10:00am
Engr IV Room 57-124
Buildings consume approximately 73% of the total electrical energy, and 12% of the potable water resources in the United States. Even a moderate reduction in this sector results in significant monetary and resource savings. The ability to monitor resource consumption at a needed granularity provides consumers and building owners with deeper insight into their resource waste, resulting in improved efficiency. Monitoring resource consumption at a fine granularity, however, is difficult with available technologies because the use of expensive sensors or professional installation of in-line sensors is necessary.
This dissertation first introduces an alternative approach for fine-grained monitoring of resource consumption in buildings. Since resource-consuming end-points (such as appliances, water fixtures) emit measurable signals when they are consuming resources, low-cost indirect sensors can be used for inferring real-time resource consumption. However, indirect sensors cannot be calibrated during manufacturing because of varying ambient conditions and sensor placement. The main challenge is to provide a method that automatically calibrates the indirect sensors, and learns the features of resource consuming activities. We develop a sophisticated model-based optimization and multi-modal sensor fusion framework in conjunction with readings from the central meter of the building. The use of low-cost indirect sensors together with the autonomous calibration and learning algorithms provides desirable characteristics: high-resolution, economic, self-configuring, and scalable.
Furthermore, we go beyond unit or end-point level accounting of resource usage to provide an association of resource usage in real-time with a specific person, when fused with side information that help detect human occupancy. Finally, we discuss a privacy implication of the temporally fine-grained measurement, which advanced electric meters could deliver to utility providers. A privacy preserving metering technique is proposed and evaluated.
Younghun Kim received his B.S. and M.S. degrees in Electrical Engineering Department from Seoul National University, Seoul, South Korea in 2004 and 2006, respectively. He is currently pursuing his PhD with the Networked and Embedded Systems Laboratory and with the Center for Embedded Networked Sensing at UCLA. His research interests include sensor networks, embedded systems, non-intrusive resource monitoring, smart buildings, and their application to environmental sustainability.