Space-mate : a framework to harmonize occupant well-being and building sustainability
Abstract/Contents
- Abstract
- Knowledge workers, i.e. employees, students, and faculty, are an institution's greatest asset. Nevertheless, studies show that: 1) Poor health, lost productivity and active disengagement of knowledge workers cost the U.S. economy over $1.3 trillion dollars per year. 2) Knowledge workers, like most occupants, spend 90% of their time in buildings. 3) Buildings consume 73% of U.S. electricity and generate 39% of U.S. greenhouse gas emissions. These are key drivers of the Space-Mate study asking an overarching question: How can we harmonize knowledge worker well-being and building sustainable performance? Traditionally, the interaction between occupants and buildings has been represented using occupant comfort models and building performance simulation tools. The purpose of occupant comfort models is to provide neutral indoor environments which reduce negative impacts on the occupant. These models are discrete, deterministic and based on aggregated, qualitative survey data. They do not differentiate between individuals nor consider quantitative occupant psycho-physiological state and well-being variables nor the activities that the occupants are performing. Building performance models consider occupants as building internal loads. Occupant behavior is complex and stochastic, yet current building performance models represent occupant behavior as discrete, deterministic and unchanging in hour-long periods of time. This leads to questionable results and a known gap between the model and the building's actual performance. The Space-Mate study builds on 3 research areas that provided theoretical and practical points of departure: built environment, affective computing, and physiology. Built environment research efforts to monitor and model occupants in buildings have focused on detecting the occupant spatial location, their energy-related occupant behavior, and their subjective awareness of the impact of building indoor environmental quality (IEQ) on their well-being. These models are used as input to simulation tools with the goal of predicting the building performance with an emphasis on energy consumption. These approaches do not consider monitoring, modeling or simulating the occupant variable psycho-physiological state, nor the dynamic interaction between the occupant and the building IEQ, and its impact on building energy consumption. Affective computing monitoring and modelling methods rely on wearable sensor data and implement machine learning algorithms to identify and predict psycho-physiological well-being indicators but do not monitor the built environment surrounding their participants, nor do they provide multi-variable models to represent the participant's composite psycho-physiological state. Physiology research methodologies use fixed or wearable sensors to monitor athletes' physiological parameters. However, their focus is on high-intensity exercise activities. Extant research does not consider low-intensity activities typical of knowledge work. Consequently, their existing artificial intelligence models are inaccurate for these types of activity levels. To address these challenges, I formalized the following research questions: RQ1: How can we dynamically track occupant spatial, temporal, psycho-physiological states? RQ2: How can we model variations in the occupant state? RQ3: How can we model the interaction between the occupant state and the building state? RQ4: How can we simulate the dynamic interaction data flow between the occupant and the building towards harmonizing occupant well-being and building performance? The Space-Mate study leverages a three-step methodology, Monitor-Model-Simulate. The units of analysis were one occupant inside one interior room of the building. The occupants were volunteer participants in the AEC Global Teamwork course at Stanford University. The data was collected during their weekly 2-hour project meetings. The Monitor step of Space-Mate consists of a concurrent occupant psycho-physiological state and building indoor environment data collection which uncovers the variability of both the occupant well-being and building interior environment as a function of the occupant activities. I developed an instrumentation and data collection protocol and collected ten data types—five occupant psycho-physiological variables, three building room variables, and two context variables. Two correlated and synchronized occupant-building datasets were collected in 2017 and 2018, for a total of 33 meetings. Towards gaining an understanding of the occupant-building datasets, I developed interactive big data analysis and visualization applications to detect variability and patterns reflecting the duration and transitions of states for the occupant, the building room and the context variables. The Model step used the correlated and synchronized occupant-building datasets for the development, training, testing, and validation of four analysis models: 1. An Occupant State Classification model defines and classifies a finite set of occupant psycho-physiological states through unsupervised and supervised machine learning classification. 2. An Occupant State Transition model implements a Markov chain model to represent the transition probabilities between occupant psycho-physiological states. 3. An Activity Duration and Transition model predicts the duration of an activity during the meeting and represents the transition probabilities between activities with a Markov chain. 4. A Building State Transition Neural Network model implements a neural network time-series model to predict the variations in the building indoor environment state as influenced by the occupant psycho-physiological state. The Simulate step integrates the four analysis models into a novel Space-Mate framework. The Space-Mate framework is based on a 6-step iterative cycle that represents the dynamic interaction data flow between the occupant and building collaborating towards positively impacting the occupant psycho-physiological state in support of the occupant activity and managing the building sustainable performance. Space-Mate presents an innovative perspective by assigning agency to both the occupant and the building, enabling them to work together taking into consideration their independent goals, constraints, and preferences. The Monitor-Model-Simulate methodology and the Space-Mate framework provided the road map of the Space-Mate study and implementation of the Space-Mate simulation prototype presented in this dissertation. To test and demonstrate the feasibility of the Space-Mate framework, I developed a prototype implementing the 6-step Space-Mate simulation. The simulation results indicate that the Space-Mate decision-making framework can lead up to 35% improvement in occupant well-being. When the occupant and building work together, similar improvements in occupant well-being can be achieved with up to 65% less energy consumption than when they work independently. The development of Space-Mate led to contributions to knowledge, technology, practice, and the three areas of research that informed the points of departure for this study. The contributions propose a protocol for concurrent occupant and building instrumentation and data collection, correlated and synchronized occupant-building datasets, a Big Data Analysis and Visualization application to study any combination of time-series data, and visualize content in context; four analysis models, the Space-Mate framework, and the Space-Mate simulation prototype that is customizable and scalable. Findings from the data analysis and visualization indicate: 1) Patterns of data variability, 2) Loss of variability and increase in error result from data aggregation, and 3) Data variability by individual and by activity. In closing, Space-Mate limitations and future research directions are discussed.
Description
| Type of resource | text |
|---|---|
| Form | electronic resource; remote; computer; online resource |
| Extent | 1 online resource. |
| Place | California |
| Place | [Stanford, California] |
| Publisher | [Stanford University] |
| Copyright date | 2019; ©2019 |
| Publication date | 2019; 2019 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Author | Grey Rodriguez, Flavia Cristina | |
|---|---|---|
| Degree supervisor | Fruchter, Renate | |
| Degree supervisor | Law, K. H. (Kincho H.) | |
| Thesis advisor | Fruchter, Renate | |
| Thesis advisor | Law, K. H. (Kincho H.) | |
| Thesis advisor | Bhargava, Sumit, (Clinical Associate Professor) | |
| Thesis advisor | Fischer, Martin, 1960 July 11- | |
| Degree committee member | Bhargava, Sumit, (Clinical Associate Professor) | |
| Degree committee member | Fischer, Martin, 1960 July 11- | |
| Associated with | Stanford University, Civil & Environmental Engineering Department. |
Subjects
| Genre | Theses |
|---|---|
| Genre | Text |
Bibliographic information
| Statement of responsibility | Flavia Cristina Grey Rodriguez. |
|---|---|
| Note | Submitted to the Civil & Environmental Engineering Department. |
| Thesis | Thesis Ph.D. Stanford University 2019. |
| Location | electronic resource |
Access conditions
- Copyright
- © 2019 by Flavia Cristina Grey Rodriguez
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...