Search

With the recent advances in the Internet of Things (IoT), more and more smart devices are entering our everyday life. From light to door, these devices are usually capable of wireless communication and can be remotely controlled. However, most consumer device providers require installation of a separate app on a smartphone or tablet to best control the devices. This incurs a significant time and learning cost for the users. In this dissertation, I propose to use context-awareness to address the problem of how to help end-users discover, connect to and control IoT devices in more intuitive and efficient manners by utilizing low power neighborhood discovery, augmented reality, and machine learning algorithms. This dissertation details the technical solutions to the following challenges: • To identify, access and control individual IoT devices intuitively, I propose a mobile augmented reality system, ArcIoT. In ArcIoT, the user points a smartphone camera towards a device to control, chooses the device by a tap on the screen, and interacts with the displayed interface. • To provide seamless and personalized automation in human-device interaction, I propose rIoT, a system that leverages existing technologies to operate across heterogeneous devices and networks to provide a one-stop solution for device interaction. • Handling conflicts between users is a key enabler of multi-resident smart space. I propose a context-aware conflict prediction framework to facilitate multi-user interactions in smart spaces. Collectively, with these works, we demonstrate the feasibility of mitigating users’ time and learning costs of interacting with IoT devices by utilizing context. Our solutions explore the potentials for a unified system to manage interactions between IoT devices and end-users adaptively and smartly in various situations.