Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Computer Science

Committee Chair/Advisor

Jacob Sorber

Committee Member

Bart Knijnenburg

Committee Member

Brian Dean

Committee Member

Long Cheng


Wearable devices have become ubiquitous in modern technology, providing a convenient way for a large number of users to monitor their health and track their daily activities. However, despite their numerous benefits, wearable devices are not without their challenges. One of the most significant issues is the need to recharge their batteries, which can disrupt users’ daily routines and might result in the discontinuation of device use.

Replacing batteries with capacitors and charging them using the available harvested energy from the surrounding environment (e.g., kinetic, thermal, or solar energy) eliminates the burden of recharging or replacing batteries, resulting in convenient, sustainable, eco-friendly, and maintenance-free wearables that can expand health-tracking features to more populations e.g., geriatric and pediatric. However, these benefits come at a cost, as batteryless wearables operate intermittently (based on energy availability), which introduces design challenges and usability limitations that do not exist in their battery-powered counterparts.

In this dissertation, we present some of the challenges associated with the development of batteryless wearable sensor devices and propose solutions that can assist developers in designing such usable devices.

First, existing batteryless energy management techniques are unable to handle fluctuations in energy harvesting that occur while wearable gadgets are in use, which negatively impacts sensor wearables’ performance.

Second, batteryless devices lose power multiple times a second when traditional timekeeping methods like system clocks stop working. Several approaches have been proposed for keeping time during power failures, but they either require additional hardware components for high-precision measurements or rely on existing hardware components for low precision. Third, users of batteryless wearables are not involved in the design process. However, doing so would allow developers to understand users’ perceptions of the inherent trade-offs they face when switching from battery-powered to batteryless devices, resulting in usable batteryless wearables. Lastly, the unavoidable frequent power outages when harvested energy is scarce compromise the accuracy of batteryless daily sensing wearables, and such devices have not been tested for user acceptance in uncontrolled settings.

To overcome these limitations in the existing state of the art of batteryless wearable sensors, we have developed a flexible hardware energy storage technique called Stash that allows batteryless wearable sensors to scavenge energy efficiently under variable and unpredictable harvesting conditions with negligible power overhead and no additional software complexity. Additionally, we have explored a new method for keeping track of time during short power outages with high precision that requires no extra hardware or energy. Also, we have conducted a scenario-based online study of 400 wearable device users to uncover their concerns, expectations, and preferences when switching from battery-powered to batteryless wearables. Additionally, we developed a wristband, batteryless fitness tracker prototype called SolarFit, to assess the feasibility of accelerometer-based step counting under intermittent operation. We studied how this intermittency impacts accuracy and how users perceive it in a study involving 32 participants.

Author ORCID Identifier


Available for download on Tuesday, December 31, 2024