Old Dog, New Tricks: Seeking Metrics for Energy Harvesters as Sensors

Designing batteryless sensors presents many challenges, starting with selecting the right components for a particular application. Every new sensor added to a device can be costly, both monetarily and in energy expense. Recent research has begun to look at using the very harvesters that are already on the device as sensors for various applications. Unfortunately, not all harvesters are created equal and it is not as simple as just looking up current standards in a component's datasheet. In this paper, we propose that new metrics for energy harvesters are needed by the community to reduce complexities in the design process of selecting the optimal harvester to use as a sensor. Using a sampling of 9 solar harvesters, we ran 270 experiments to profile and compare how each reacted to a simple motion event. We also propose and explore a few sample metrics useful in selecting solar harvesters as sensors and discuss their potential impact on different applications.