Smart Earring

Low-power Wireless Earring for Longitudinal Earlobe Temperature Sensing

Please refer to the paper for technical details.

Shirley loves earrings. She has a collection of earrings at home, and one day came up with this idea: earrings could be an option for people who would like to track personal health with smart wearables. And she found me, who was a master student in ECE at that time, applied to do research in the Ubicomp Lab.

I like the idea and she seems to be a very nice person to work with, therefore, I joined.

We started with testing

As body temperature is an important vital sign, it can be used as an indicator of physical wellbeing. Measuring body temperature has become more important during the pandemic. Moreover, studies showed that temperature is also associated with emotional states. Longitudinal body temperature tracking would enable a better understanding of both physical and mental health. Therefore, I led the design of a low-power, small-scale earring that can passively sense skin temperature on the earlobe on the earlobe and stream the data to mobile devices wirelessly.

Smart earring that sense skin temperature in natural daily scenarios, like in a meeting or during exercising, and stream the data to smartphone

Considering the constraints of scale, weight, and power consumption of the earring, I explored different temperature sensors, Bluetooth modules, and coin cell batteries to achieve the optimized design. The final millimeter-scale system I designed and prototyped successfully receives accurate temperature data and streams to smartphones with microwatt-level power consumption. See the picture of system design and first prototype below:

Our system design contains two temperature sensors, a bluetooth module and a coin cell battery.

Now, I am actively exploring power harvesting techniques for wearables, including from kinetic and ambient energy sources such as daily activities and light, in order to enable self-powered longitudinal body temperature sensing without increasing the user burden.

The goal of this project is to achieve longitudinal body temperature tracking with a self-powered earring. With the longitudinal temperature data, we would apply machine learning to make further personalized inferences, including fever detection, menstrual cycle prediction, and potentially study the relationship between emotion and body temperature.

Check out the poster presented at UW Paul G. Allen School’s 2022 Annual Research Showcase.