Supercapacitor Energy Buffering for Self-Sustainable, Continuous Sensing Systems

Continuous, high-data-rate visual sensing and processing in the field is important for intelligent transportation, environmental monitoring, and area / context awareness. Rechargeable batteries in self-sustainable systems suffer from adverse environmental impact and fast aging. Advancements in supercapacitor energy density and low-power processors have reached an inflection point, where a data-intensive visual sensing system can rely solely on supercapacitors for energy buffering. This paper demonstrates the first working prototype of such a system, consisting of eight 3,000-Farad supercapacitors, a custom-designed 70-mW controller / harvester board, and a Nexus 7 tablet com- puter. We leverage the voltage-to-stored-energy relationship in capacitors to enable precise energy buffer modeling (no more than 3% error in time-to-depletion prediction). We find the supercapaci- tor self-discharge (or leakage) to be a minor issue in practice where the operating power significantly exceeds the leakage power. However, accurate energy budgeting must account for the variation of effective capacitance—particularly lower capacitance at lower voltages nearing energy depletion. Precise energy budgeting allows model-driven system adaptation (such as dynamic CPU configu- rations on low-power chipsets) and consequently realizes high, stable operational quality-of-service. Our working prototype has been successfully deployed at a campus building rooftop where it ana- lyzes nearby traffic patterns continuously