Wearable sensors are being widely used for health monitoring. These wireless sensors are envisioned to operate on scarce harvested energy resources. In addition to the hardware and software constraints arising from the form-factor and low energy operations, there are safety requirements such as avoidance of physical injury. The design implications of these requirements are non-intuitive and may involve the estimation of human physiological dynamics. The physical impact of a sensor operation can be controlled by appropriate design of multiple sensor components such as the processor, radio, and optimization of data algorithm. For example, the risk of thermal injury to tissue can be reduced by limiting the sensing frequency, the computation power, and the radio duty cycle of the body-worn sensor. Hence, it is a challenging task to trace back a cause of a physical impact to hardware and software design decisions in a sensor. This talk will focus on a non-linear optimization framework to generate sustainable and safe sensor configuration while considering interactions of the wearable sensors with the environment. This methodology is demonstrated using three case studies: a) continuously monitoring ECG sensor sustained by body heat, b) thermally safe network of implanted sensors, and c) infusion pump control algorithm to avoid hypoglycemia.
Priyanka has obtained her Ph.D. in Computer Science and Engineering in Dec 2015 from Arizona State University. Her research was focused on safety, security, and sustainability of IoT systems for healthcare using mathematical modeling and simulations. Currently, she is working in Intel Corporation as an IoT software engineer. Her current research at Intel involves product innovation through the integration of deep learning and computer vision on Intel IoT platforms. She also provides advanced technical training to customers on the Intel IoT technology through in-person workshops and architect their IoT solutions.