PhD Defense by Dingtian Zhang
Title: Ubiquitous Self-Powered Ambient Light Sensing Surfaces
Ph.D. Candidate in Computer Science
School of Interactive Computing
College of Computing
Georgia Institute of Technology
Date: April 9th, 2021 (Friday)
Time: 9:30am - 11:30pm (EDT)
Dr. Gregory D. Abowd (Advisor), School of Interactive Computing, Georgia Institute of Technology
Dr. Canek Fuentes-Hernadez, School of Electrical and Computer Engineering, Georgia Institute of Technology
Dr. Thad Starner, School of Interactive Computing, Georgia Institute of Technology
Dr. HyunJoo Oh, School of Industrial Design & School of Interactive Computing, Georgia Institute of Technology
Dr. Sauvik Das, School of Interactive Computing, Georgia Institute of Technology
Many human activities interfere with ambient light in a predictable and detectable way in that our activities implicitly or explicitly block the paths of ambient light in our environment. This dissertation explores sensing of ambient light interference patterns due to human activities as a general-purpose signal at the surface level of everyday objects for activity recognition as well as novel interaction techniques. Using low-cost optoelectronic components and flexible materials, I developed photodetector-based “ambient light sensing surfaces” that can be highly scalable and conformal to our living environments. Through strategic integration with the physical surfaces of everyday objects, they can capture the most relevant ambient light signals and detect events of interest. Emerging optoelectronic technologies such as organic semiconductors (OSC) offer ultrathin and flexible form factors, and amenability to additive mass manufacturing processes that can drive down the cost. With both silicon and OSC-based prototypes, I demonstrated the feasibility of ambient light sensing surfaces that could offer a practical path towards ubiquitous deployment.
This dissertation has made the following contributions:
- A self-powered ambient light sensing platform, OptoSense, which has a flexible form factor to conform to everyday surfaces, and tunable fields of view and sensing ranges. OptoSense has achieved high accuracy in a range of human activity detection tasks while being self-powered by ambient light.
- A design framework that provides design guidelines and inspirations for ambient light sensing surfaces. Based on variations of parameters, it generates a comprehensive set of example applications covering both previous research and those that are uniquely demonstrated in this work.
- A novel concept of a computational photodetector. A computational photodetector has a functional pattern design for sensory signal processing in the analog domain for feature extraction, which reduces the power and latency of the system at scale, and better protects user privacy.
- A design space of computational photodetectors and several representative patterns each addressing different aspects of position, orientation, and motion sensing, demonstrated and evaluated with OSC-based computational photodetectors in a range of human activity sensing applications.