Typically the standard Perpetuum Vibration Energy Harvester will yield 3-5mW on a pump/motor set and over 90% of "smooth" running machines will give at least 1mW. On trains 10-15mW is readily available. 3mW is typically the average level of power required for an Industrial WSN whether it is high volumes of data at low reporting frequency (eg vibration spectra) or low volumes of data at high frequency (temperature/pressure). Such power levels are readily available from commercial companies like Perpetuum for Vibration Harvesters , Perpetua for Thermal harvesters, and others (SolChip) for Photyovoltaic.
Intelligent Power Modules are available that replace standard battery packs in GE, Emerson and Honeywell products and enable different types of harvester to be simply connected to the WSN.
It is true that a key technology limitation is the availability of high quality energy storage devices that will last as long as the rest of the system but I disagree that marginal improvements in power management efficiency are critical to the successful deployment of EH powered WSN's
Although "energy harvesting" is not new in history and "network nodes" or "dust networks" ideas go back to the cold war years, only in the last few years it has began to mature and reach the consumer market. It is possible that the now widespread debate and definition of the IOT will catapult the evolution and furher the deployment of all the implied technologies. As mentioned by others, one of the crucial challenges is the energy storage which must be quick to capture and retain energy quanta (surges). In my view, the future of this component lies on the still slow evolution of the super-caps.
For those of you interested in getting up to speed in this topic, check my article "Autonomous Sensor Network Nodes" publiched by EETimes' Editor Clive Maxfield in 12/10/2012. It provides a historical account for energy harvestin and sensor network nodes. Of particular importance for the evolution of this segment of the IOT are the technologies developed by Linear Technology and Texas Istruments back in 2009-2010. Both propose solutions for energy harvesting power conditioning. Texas has launched an increasing family of ultra-low power(ULP) microprocessors that consume microwatts of power and awake in a matter of microseconds.
Article is available at
Energy harvesting is now but the way you manage it inside the circuit is the future.
Here we show how a system can be fully energy-driven only woken-up on energy-events whatever is the energy harvested in the environment. It also allows to drastically increase the system life as in a best-effort mode, the battery could never be used.
We have had vibration energy harvesters powering WSN's made by GE, Emerson and others for many years and we have installed many VEH's powering complete systems to monitor bearings and wheels running around on trains in several European countries, Alerts can be sent automatically and the data can be viewed over the Internet. Researchers need to catch up with what is already happening.
Soon we find many wireless sensor nodes in difficult terrain. They are good until healthy. How will we handle situation after they stop working? How will we recover them and recyle them properly so as not not contaminate nature and its eco system?
Looking into environmental aspect may be of utmost importance before we indulge too much on these fancy techniques. Else it is conemporary fashion to spoil environment first and pretend latter to recove situation and make it so called green.