UC Berkeley

Wearable sensors that allow continuous, real-time, and point-of-care monitoring of body state have many advantages over invasive, hospital-based tests. Wearable sensors that probe the chemical state of the body by measuring analytes in sweat have generated particular interest in recent years, due to sweat’s rich composition and ease of access. Preliminary studies have revealed that sweat composition and secretion rate may indicate blood concentrations or health conditions including dehydration, metabolic conditions, and stress, enabling physiological health to be tracked dynamically without painful, disruptive, and discrete tests. However, accessing sweat for measurement poses unique challenges. Sensors must limit evaporation and mixing of old and fresh sweat, must be robust under the mechanical rigors of prolonged on-body wear, and must be expanded towards low sweat volumes and secretion rates in order to collect sweat without needing to exercise. Careful interpretation of sensor measurements must also be performed to account for cross-dependencies between measured parameters and to develop personalized and universal correlations between sweat measurements and established health metrics.

In this dissertation, I discuss the vision of wearable sweat sensors for personalized health monitoring and present technological progress towards realizing this goal. Chapter 1 introduces background and challenges of sweat sensing, as well as key considerations and components of wearable sweat sensors. Chapter 2 addresses high-throughput fabrication of electrochemical sweat sensors via roll-to-roll processing. Chapter 3 discusses the integration of microfluidics to enable dynamic sweat capture and analysis of both composition and sweat secretion rate, with a focus on sweat that is secreted via exercise or local drug-based stimulation. Chapter 4 moves beyond these disruptive modes of sweat generation to present a platform for rapidly accumulating and analyzing sweat at rest, via functionalized gloves that dramatically inhibit evaporation. Chapter 5 presents a complementary system using microfluidics to dynamically monitor sweat at rest, including while sitting and sleeping. Throughout these chapters, small-scale human subject trials and studies highlight the potential of wearable sweat sensors for personalized health and fitness monitoring applications. Chapter 6 concludes by discussing work that must be done to enable sweat sensing to become a viable and ubiquitous tool for health monitoring outside the lab.