Sensors Online
www.sensorsmag.com


INTELLIGENT SYSTEMS
Table of Contents

Power Considerations for
Wireless Sensor Networks

Two ways to power remote, portable devices are tiny long-life batteries and energy harvesting techniques. The choice depends on the demands of the particular application.

figure
Figure 1. A self-organizing wireless sensor node from Millennial Net incorporates all the required components, except for the sensor, in this form factor.
Power conservation is a critical component of the wireless sensor networks used in hundreds of commercial, industrial, medical, consumer, and military applications characterized by low power requirements and low data rates. Long battery life (typically 5–10 years) is essential in wireless sensor networks (WSNs) where line power is not available or where the sensor nodes are mobile. Because the philosophy of sensor networks is “wireless anywhere,” the size of a sensor node is also critical. Many applications require nodes not much larger than a dime (see Figure 1). In these cases, even AA batteries are too bulky and coin cell batteries are the only option. WSNs are capable of operating on sub-milliampere power, allowing 3 VDC coin batteries to power a sensor node for up to 10 years and beyond, depending on the sampling rate.

But in many WSN applications, using and replacing batteries, even long-lived types, is impractical. Often the sensor nodes are located in unrecoverable locations. Furthermore, the labor and costs associated with changing hundreds, if not thousands, of batteries outweighs the ROI that the sensor network could deliver. In response to these constraints, systems designers and engineers are looking for alternative ways to deliver exceptionally long life to sensor network devices. One method that is gaining traction among designers is the extraction of electrical power from persistent ambient vibrations.

Advances in vibration-based, environmental harvesting have made it possible to provide the power required to operate WSNs and make totally battery-free WSNs feasible. Candidate application areas include machine condition monitoring, HVAC monitoring through duct- or machine-mounted sensors, vehicle tire pressure and temperature monitoring, and the remote colonies of sensors used in aircraft and ships to detect potential problems.

Because these harvesters are so sensitive and efficient, they can generate electricity from vibrations that are barely noticeable to the human touch. Input vibrations are measured in g’s, where 1 g is equal to the acceleration of gravity. (Tapping on a table creates ~0.02 g, or 20 mg vibrations that are detectable by a hand.) One example is the Energy Harvester made by Ferro Solutions.
figure
Figure 2. Continuum Control Corp.’s harvester package, featuring an integrated transducer and harvesting electronics including power conditioning up to 3 V, offers tunable coupling and frequency.
In an environment with vibrations at 28 Hz and 100 mg, this harvester produces a power output of 9.3 mW from a cylinder measuring ~1.8 in. dia. by ~1.8 in. high. Doubling the volume of the harvester will double the useful energy. The power output scales linearly with increased vibration frequency and exponentially with increased g-force. When the electricity thus generated is not used immediately, it can be stored in a super capacitor.

Another example is Continuum Control Corp.’s iPower energy harvesters. These devices, about the size of a pack of gum (see Figure 2), extract electric energy from mechanical vibrations, motion, or impact, and store it for use by wireless sensors or other electronic devices. The technology couples proprietary transducers and circuits to a mechanical system, creating a solution that maximizes power flow from the mechanical to the electrical storage and is specifically targeted at converting micromotion into usable power (see Figure 3).

figure
Figure 3. Continuum’s technology couples proprietary transducers to the vibration source and proprietary circuits, creating a system that maximizes power flow from the mechanical to the electrical storage. High performance comes from exploiting knowledge of the states of the structure and circuitry to control power flow and thus optimize available power and conversion efficiency.

Both technologies provide enough energy to power a wireless sensor node and its attached sensors, and both are viable alternatives to line or battery power for the appropriate sensor network application.

Summary
Environmental energy harvesting from vibration sources promises nearly unlimited life for low-power wireless sensor networks. With even conservative projections putting the number of units to be sold for WSN devices in the vicinity of hundreds of millions each year, suppliers of WSN technologies must do their part by making their products capable of supporting power line, battery, and environmentally powered devices. This cooperative effort will result in savings for the end users and a new degree of freedom for the product designers and engineers responsible for implementing these networks.

Acknowledgments
The author wishes to thank Kevin O’Handley of Ferro Solutions (Roslindale, MA; 978-273-4709, kevin@ferrosi.com) and Nesbitt Hagood of Continuum Control Corp. (Billerica, MA; 978-670-4910, nwh@continuumphotonics.com) for their assistance in preparing this article.

Energy Harvesting for Wireless Sensor Networks
Stephanie vL Henkel

Making machines and structures “smart” allows them to monitor their behavior, report on their current status, and advise of impending problems before catastrophic failure claims lives or property. To do this, the sensors must be able to transmit digital data to a remote receiver. Hardwiring the transmitters to a source of electricity is expensive and time consuming. Replacing batteries, even long-lasting types, can be a never-ending chore. Energy harvesting is an attractive solution to these problems.

figure
MicroStrain’s goal is to combine energy harvesting with low-power sensors and networks to create a completely wireless sensor network that can deployed easily in the field and operate unattended for long periods.

Energy can be scavenged from various ambient sources: solar, wind, thermoelectric, water/wave/tide, vibration, and strain. The strongest candidate for industrial environments appears to be strain. Accordingly, MicroStrain is developing an energy-harvesting scheme (right) based on storing cyclic strain energy by rectifying piezoelectric fiber output into a capacitor bank. When the capacitor voltage reaches a preset threshold, power is transferred to an integrated wireless sensor node. Alternatively, power can be stored on a rechargeable thinfilm battery. Software programming allows one hardware design to operate with many sensor types, including thermocouples, strain gauges, magnetometers, capacitive and inductive sensors, and magnetic, temperature, and humidity sensors.

Advances in single-crystal piezoelectrics have produced materials with 90% mechanical-to-electrical conversion coefficients efficiency and 1.4% operating strain capabilities. The fibers can be directly bonded to a straining element or structure, and even if some or many of the individual fibers get broken, the rest of the bundle will continue to function properly. Moreover, mass production makes these piezoelectric materials inexpensive.

figure
MicroStrain’s network uses addressable sensing nodes incorporating data-logging capabilities and a bidirectional RF transceiver communications link.

The key to making energy harvesting work is to minimize the power required by the entire system—sensors, conditioner, processor, data storage, and data transmission (above). Because the sensors tend to be in sleep mode most of the time, they draw little current. At intervals controlled by a randomization timer, the nodes wake up and transmit bursts of data. MicroStrain’s WWSN ad hoc network architecture allows thousands of multichannel, uniquely addressed sensing nodes to communicate to a central, Ethernet-enabled receiver with extensible markup language data output. Time division multiple-access is used to control communications.

MicroStrain’s development of energy-harvesting techniques has received support from National Science Foundation//Vermont Phase 0 EPSCoR and Navy Phase 1 SBIR, and, currently, from Navy Phase II programs. n

Contact MicroStrain, Williston, VT; 802-862-6629 or 800-449-3878.


MORE!
For further reading on this and related topics, see these Sensors articles.

"Secure Low-Power Operation of Wireless Sensor Networks," January 2004
"Extend the Life of Your Battery," March 2003
"Wireless Temperature Monitoring in Remote Systems," October 2002
"Brains and Brawn—The Power of Smart Batteries," April 2002
"Wireless Sensor Network Technologies," May 2000