Optimizing Indoor Environment Quality through Quantifying Human Experience using Ubiquitous Sensing

点击翻译按钮获取中文摘要

Currently, the building sector (i.e., commercial and residential) accounts for more than 25% of the total energy use in Canada [1]. This high energy use combined with the rapid progression of climate change has contributed to a global push for high-performance buildings to curtail energy consumption, by commissioning high-efficiency and highly sensed building control systems such as Heating, Ventilation, and Air Conditioning (HVAC). However, these increasingly air-tight, energy conserving buildings bring dramatic changes to Indoor Environment Quality (IEQ) conditions, which are configurations of IEQ factors including thermal conditions, air quality, acoustic, and visual comfort [2]. These changes pose challenges to human experiences in buildings (i.e., the state of mind that is reflected by our physiological, emotional, and cognitive statuses). Because people spend more than 85% of their time indoors [3], IEQ factors play an essential role in the health and well-being of building occupants. Despite efforts from professional societies and standardization organizations to develop guidelines for improving building design and operation (e.g., LEED), the prevalent method of gauging human experience (e.g., post occupancy surveys) remains subjective and after-the-fact. This research program aims to shift the status quo by deepening our understanding of human experience indoors, through an experimental approach that measures human subjects' physiological responses under various IEQ conditions. I intend to achieve this overarching goal through three short-term objectives: (1) build a laboratory system capable of delivering various configurations of IEQ factors, as well as measuring the environmental states and physiological responses of subjects; (2) conduct human subject experiments and collect environmental and physiological data to quantify the impact of different IEQ factors on human experience; and (3) build machine leaning-based prediction models for subjective ratings of human experience, using collected data, and develop an reinforcement learning-based control algorithm that manipulates IEQ factors to minimize energy consumption while improving human experience indoors. Outcomes from this program can be leveraged by architects, engineers, and facility managers across Canada at various stages of a capital project to improve building design and operation, with the goal of improving sustainability and the health and well-being of building occupants. Highly Qualified Personnel (HQP) involved in this research program will be trained for the marketable skills of experiment design, data collection and management, advanced statistics, and machine learning techniques for data analysis, all highly desirable in the field of civil and mechanical engineering.