Work

Semi-supervised training for infrastructure mediated sensing: disaggregated hot and cold water sensing with minimal calibration

Year: 2013
Type: dissertation
Abstract: This thesis formalizes and evaluates the use a central sensor for detecting and classifying water usage in a home or apartment building.While the discussion largely explores the theory and implementation of the sensing, the main impact of this thesis is to address a single assertion: we are running out of water.With this sensing technology, my aim is to support initiatives in increasing awareness of water consumption and understanding of the diversity in ways that vast populations consume water.Through my sensing, I classify (1) what type of valve was activated in a home (i.e., faucet, dishwasher, toilet), (2) the temperature of water that the valve uses (i.e., hot, cold, or mixed), and(3) where in the home or apartment the valve resides (i.e., kitchen, master bathroom).This is typically known as disaggregated water sensing or end-use water sensing.In my approach, I use a central sensing point based on two pressure sensors.The sensors monitor pressure fluctuations in the home, providing a time series upon which to classify events.I formalize the feature extraction process using template matching, expertly chosen features, and sparse codebooks.Ultimately I show that few features can be created that generalize across residences, necessitating that each residence go through a calibration process.I formalize this calibration process using semi-supervised techniques such as expert knowledge through a rule-based classifier, co-training, and virtual evidence.I evaluate the system in several studies, including a longitudinal deployment of 15,000 water usages across a 1 month period in five homes.The results show that, with about 30-40 labeled examples, the system is 93% accurate at detecting the type of fixture used, 89% accurate at detecting which fixture in the home is activated, and 81% accurate at also detecting the temperature state of the valve.These accuracies are reported on a test set of 2,500-4,000 water usage events per residence.This thesis, then, has implications in the water sustainability community, activity detection community, and the machine learning and signal processing communities. i. For my childrenIn hope that anyone can leave the world in a better condition than when they found it ii.This thesis represents a body of work over several years in which I devoted my studies and efforts to collecting and analyzing water usage.During that time I have been influenced positively by a number of individuals.I would like to thank my committee for their thoughtful commentary and helpful revision suggestions.They have truly helped me to formulate the probing, scientific questions with which any scholarly work should adhere.Thank you especially to Mari Ostendorf for her suggestions in active learning research.I would like to thank my advisor Shwetak Patel and my co-advisor Les Atlas for their time in reviewing methodologies and their helpful suggestions in making the work more scholarly.Countless brainstorming sessions were pivotal in the exploration of this work.I aspire to be a mentor to other students as you both have been mentors to me.To Shwetak, I have never known someone so creative.I cannot remember how many times I came to you with a problem that I thought was insurmountable, only to leave our brainstorming session with no doubt in my mind that I could solve it.I would like to thank my wife and family for their devotion, love, and understanding while I carried out this work.Chelsea, you are simultaneously what keeps me grounded and what keeps me reaching for the sky.I would like to thank the members of my lab who helped me to build and deploy an array of different sensors in their homes, especially to Tim Campbell, Elliot Saba, TienJui Lee, Gabe Cohn, and Eric Swanson.Without your friendship this research would have been far less enjoyable and orders of magnitude more difficult.Lastly I would like to thank Jon Froehlich for the countless hours that he spent with me collecting, analyzing, and fine tuning the ideas to make this research possible.Jon, you are like a co-advisor to me.I am a better researcher, better
Language: English
Open Access status: green
OpenAlex ID: W2782318465
Cites: 79
Cited by:
Related to: 10