Self-Powered Sensor System Design in Dynamic Low Harvesting Environments

The internet of things (IoT) covers a wide umbrella of applications that range from smart environments, transportation, to even healthcare. However, challenges in interoperability, security and privacy, resilience, and reliability need to be overcome for the full realization of the IoT. To improve reliability, energy harvesting and self-powered operation are promising alternatives to sustainably and efficiently power remote sensing devices by harnessing energy from ambient energy sources while ensuring continuous operation. Although successful demonstrations of self-powered sensing have been presented in the literature, applications with very low energy harvesting levels and high energy fluctuations continue to be a challenge for researchers and designers.

Current works in self-powered sensing have focused on developing more efficient harvesters, ultra-low power electronics, and new dynamic power management strategies. However, the synergistic integration of these individual efforts is necessary to enable self-powered operation in dynamic low harvesting environments. Furthermore, the dynamics of energy harvesting under these conditions are not yet well understood, resulting in not so efficient and/or heavily duty-cycled systems that can miss relevant information. Thus, this work proposes a framework for the design of self-powered sensors intended to operate in dynamic low harvesting environments. The framework integrates three components: energy harvesting profiling, system-level power optimizations, and application-specific quality of information (QoI) metrics. The methods in the framework are implemented in two medical applications as case studies: vigilant cardiac monitoring and continuous respiratory health monitoring.

In the case of energy harvesting profiling, an energy harvesting and data collection (EHDC) system was developed to provide a deeper understanding of energy harvesting dynamics in real-world scenarios. The EHDC platform monitors and records the instantaneous usable power generated by body-worn harvesters, while also collecting human activity and environmental data to provide a comprehensive real-world evaluation of two energy harvesting modalities common to wearable sensors: solar and thermoelectric. Additionally, a mathematical model for piezoelectric cantilevers that correlates fluid flow characteristics with energy harvesting availability was created. The model incorporates principles from fluid dynamics, elasticity theory, piezoelectric science, and circuit design. These techniques aim to provide real-world energy harvesting information under dynamic low harvesting environments to assist in the design and selection of harvesters and low power electronics.

In regards to system-level power optimizations, system power modeling was used as a tool to identify potential variables for power optimizations and analyze their impact in the total system’s power consumption to meet specific power budgets defined by the harvesters or sensing requirements. Similarly, energy storage sizing was conducted as a mechanism to deal with energy harvesting fluctuations and to guarantee the system operation during prolonged energy harvesting droughts. Furthermore, the effect of piezoelectric cantilever shape and size was investigated to determine the best form factor for self-powered fluid flow sensors that operate under non-resonance and sub-Hz conditions. This evaluation was complemented by an assessment of two common energy harvesting circuits for piezoelectric harvesters in dynamic low harvesting environments.

Finally, the definition of application-specific information metrics as an alternative to traditional digital signal metrics to determine the relation between system power consumption and quality of information is proposed. This technique aims to assist in the formalization of specifications for sensing systems during the design process to achieve self-powered operation while providing useful information. The successful implementation of the proposed methods resulted in the demonstration of functional prototypes of self-powered health monitoring systems for the two case studies discussed in this dissertation.