Off-campus UMass Amherst users: To download campus access dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.
Non-UMass Amherst users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Dissertations that have an embargo placed on them will not be available to anyone until the embargo expires.
Open Access Dissertation
Doctor of Philosophy (PhD)
Year Degree Awarded
Other Computer Sciences
The proliferation of smart meter deployments has led to significant interest in analyzing home energy use as part of the emerging 'smart grid'. As buildings account for nearly 40% of society's energy use, data from smart meters provides significant opportunities for both utilities and consumers to optimize energy use, minimize waste, and provide insight into how modern homes and devices use energy. Meter data is often difficult to analyze, however, owing to the aggregation of many disparate and complex loads as well as relatively coarse measurement granularities. At utility scales, analysis is further complicated by the vast quantity of data, which precludes the use of computationally intensive techniques when monitoring hundreds or even thousands of homes.
In this thesis, I present an architecture for enabling smart homes using smart energy meters, encompassing efficient data collection and analysis to understand the behavior of home devices. I consider four primary challenges within this domain: (1) providing low-overhead data collection and processing for many devices, (2) designing models characterizing the energy use of modern devices, (3) using these models to track the real-time behavior of known devices, and (4) automatic identification of unknown devices in the home.
To enable practical smart homes, my proposed architecture combines low-cost, off-the-shelf sensing equipment with a hybrid local and cloud-based processing backend. To analyze data within the environment, I first characterize the basic device types present in today's homes (e.g., resistive, inductive, or non-linear) and distill the essential usage characteristics of each type. Using these characteristics, I construct a set of models that more accurately represents real-world devices than previous simplistic models. I then leverage this modeling framework to track the behavior of specific devices, using a technique that runs in close to real-time and can scale to many devices. Finally, I present a technique to automatically identify unknown devices attached to smart outlets in homes, which relieves homeowners of the need to manually describe devices in order to employ smart home optimizations.
Barker, Sean K., "Model-Driven Analytics of Energy Meter Data in Smart Homes" (2014). Doctoral Dissertations. 156.