Test System to Evaluate Energy- Harvesting Technologies for Wireless Sensors

Abstract

Wireless sensor networks are rapidly growing, both in number and size, and powering their parts with batteries or wires has become difficult. Energy-harvesting is a possible way to power the sensors, which are often located in environments where harvest-able energy sources are present, such as vibration or radio-waves. However, verifying that a energy-harvesting method is capable of sustaining a wireless sensor is not trivial and using conventional lab equipment for testing the viability of a energy-harvesting method on-site is not always possible due to spatial restrictions. In this thesis, a lightweight, but precise test system to evaluate energy harvesting technologies for wireless sensors is presented. We describe and discuss the test system’s specification, from the general characteristic of energy-harvesting, to wireless sensor systems. Furthermore, details about the design are revealed and the results from testing and evaluation are presented. The system designed was capable of drawing controlled power levels and emulating wireless sensors, mimicking their power architecture interface to energy-harvesting systems. This was done with accuracy of within 5% deviation for currents from 1 A up to 250mA, with < 1% average relative accuracy over the entire range, a range that was deemed sufficient to accurately emulate most wireless sensors. The test system turned out small enough to be hand-held and is therefore easily installed in the field. The system turned out to be useful for evaluating a piezo-electric harvesting system, and our analysis indicates that the general-purpose nature of the system should make it useful for evaluating other harvesting technologies, suitable for wireless sensors, such as radio-frequency and electromagnetic energy-harvesting.