Autonomous Internet of Things in the Built Environment

建筑环境已经被连接的传感器和执行器所渗透，例如可编程恒温器、电表和运动传感器。这些设备的网络被称为物联网（IoT），具有巨大的潜力，可以实现一系列应用，并提供新的创收机会。然而，建筑环境中的物联网尚未实现在资源管理、节能运营以及与智能电网和新兴能源系统无缝集成方面取得重大进展的宏伟承诺。这主要是由于缺乏方便的抽象、适当的计算模型和数据驱动技术。

The built environment has been permeated by connected sensors and actuators, such as programmable thermostats, energy meters, and motion sensors. The network of these devices, dubbed as the Internet of Things (IoT), has significant potential to enable a range of applications and present new revenue generation opportunities. However, IoT in the built environment has not yet fulfilled the grand promise of significant advances in resource management, energy-efficient operations, and seamless integration with the smart grid and emerging energy systems. This is mainly due to the lack of convenient abstractions, appropriate computational models, and data-driven techniques. Autonomous IoT in the built environment can fulfill this promise by providing advanced monitoring, real-time analytics, and automatic control capabilities to buildings. Buildings equipped with this technology will be capable of (a) monitoring numerous environmental facets (states and events), (b) interacting intelligently with the surrounding environment, (c) learning from experiences gained through these interactions, (d) making decisions that are optimal in the long run with minimum human supervision, and (e) adapting to changes in a timely and efficient manner. This research program aims to understand, both theoretically and empirically, the properties of autonomous IoT and investigate how the application of this paradigm to buildings can improve human experience, operational efficiency, and sustainability of the broader built environment including smart cities. The long-term goal of my research is to close the loop between data generation, computation, and control in the built environment. In support of this long-term goal, my short-term research objectives are to 1. develop a distributed operating system that simultaneously accommodates the requirements of real-time decision making and compute-intensive applications; 2. study how diverse environmental facets encapsulating interactions between the building, occupants, and the surrounding environment can be monitored; 3. investigate learning-based control strategies which rely on high-dimensional sensor data to operate the building in an optimal and proactive fashion. The global market of IoT for smart buildings and smart cities is expected to grow from $6.3 billion in 2017 to $22.2 billion in 2026. My work on autonomous IoT would address the key challenges that concern the integration of IoT in the built environment by developing new computational models and distributed systems technology. The contributions of this research program would help to transform the nature of interactions between buildings and the surrounding environment, and would facilitate the adoption of the emerging energy systems in buildings. This research program fits well within the strategic priorities of Canada's tech industry and provides practical training for graduate students in the areas of distributed systems, sensor networking, and machine learning.