Recent developments in energy harvesting and autonomous sensing mean that it is now possible to power sensors solely from energy harvested from the environment. Clearly this is dependent on sufficient environmental energy being present. The range of feasible environments for operation can be extended by combining multiple energy sources on a sensor node. The effective monitoring of their energy resources is also important to deliver sustained and effective operation. This paper outlines the issues concerned with combining and managing multiple energy sources on sensor nodes. This problem is approached from both a hardware and embedded software viewpoint. A complete system is described in which energy is harvested from both light and vibration, stored in a common energy store, and interrogated and managed by the node.
IntroductionWireless autonomous sensors are devices with sensing, processing and wireless communication capabilities. In order to deliver truly autonomous operation, sensor nodes must not depend on an external power source. Conventionally, wireless sensing devices are powered by non-rechargeable batteries which must be replaced when depleted (imposing a maintenance requirement that can prove costly during long deployments), or must support the operation of the node for its complete lifetime.For sustained operation, it follows that nodes should not be reliant on non-rechargeable batteries, but instead generate their own energy in-situ. Energy harvesting is a developing technology area, and prominent technologies facilitate the generation of electricity from light (photovoltaics), vibration (vibration energy harvesting), or thermal gradients (thermoelectrics). The intermittent nature of many environmental energy sources means that viable devices must harvest energy from their operating environment when possible, and buffer excess energy in supercapacitors [1].In general, harvested energy is scarce and often unpredictable. The range of feasible environments for operation can be extended by combining multiple energy harvesting sources on one node, and managing energy resources carefully. In this paper we present, as an example, a system incorporating photovoltaic and vibration energy harvesting. Algorithms to manage these energy resources are introduced, and the embedded software is developed in line with a defined hardwaresoftware architecture [2].