Wireless Sensors: How Not to Replace 1000 Batteries
By Mark Frauenfelder, Tue Aug 17 07:30:00 GMT 2004

Wireless microsensors will be used to monitor everything from nesting conditions for endangered birds to stress-induced cracks in large structures. The big question? How to make batteries for them that last for years.

Ever since I heard the term "smart dust" a couple of years ago, I've been intrigued by the idea of tiny wireless sensors that could be installed in buildings, vehicles and outdoor locations to measure changes in the environment and communicate with one another through ad-hoc networks. Smart dust could be used to monitor stress conditions in a commuter bridge, dangerous leaks in a biotech lab, pollen density in a forest, humidity in an archival library or in countless other places. While these sensors haven't shrunk to the size of a speck of dust yet (the smallest are about the size of a Life Saver), they're already being used in the real world.

Thanks to Moore's Law, it's a given that these microelectromechanical systems (MEMS) sensors and their power supplies will get smaller and cheaper. For example, an Israeli-based company called Power Paper is working on paper-thin flexible batteries that could be used in RFIDs, giving them the ability to transmit a signal on their own. These batteries are actually printed on paper and other flat surfaces, which opens up possibilities for disposable wireless devices and laminated ID cards that transmit identities over longer distances than passive RIFDs.

Running out of Juice

But it's one thing to make a tiny battery, and another to make it with enough juice to run for long periods of time. And that's a problem, thanks to the distributed nature of MEMS network sensors. Imagine having to change the batteries ever month or so in hundreds of wireless humidity and temperature sensors distributed throughout a large facility. It would be a huge maintenance hassle. You want to be able to, in the words of the great product inventor and pitchman Ron Popeil, "set it and forget it."

Researchers are approaching the problem from several different angles. Conservation through power- management techniques is an obvious short-term solution. Many wireless sensor applications, such as humidity detection, might not require continuous monitoring under certain conditions. Each sensor could be configured to sample the air once a minute or even once an hour or day, which could greatly increase the batteries' life.

Keep off the Nanograss

Power management only goes so far, however. Anyone who has switched on a long-unused flashlight can attest to the fact that batteries have a shelf life, and will drain away even when they aren't being used. Lucent and mPhase Technologies are hoping to solve the shelf-life problem by making tiny batteries with a nanomaterial called "nanograss." In its uncharged state, nanograss is hydrophobic, which means it repels water. Nanograss could be used in a tiny battery to separate its liquid electrolyte from its electrodes, giving the battery an extremely long shelf life. When juice is needed, a small voltage is applied to the nanograss. This causes the nanograss to become hydrophilic, and suck the electrolyte down between its blades so it can come into contact with the electrode and provide current to the sensor.

What about situations in which sensors need to be powered continuously for long periods of time? There are a couple of possible solutions -- they could use ultra-long lasting microbatteries, or they could scavenge sources of energy from their environment. At Cornell University in Ithaca, N.Y., researchers are trying to make a nuclear-powered battery with a very long life span. They've built prototype batteries that use a speck of nickel-63 (a radioactive isotope) to vibrate a tiny cantilever. The cantilever could be made from a piece of piezoelectric material, which could supply power to the sensor. Nickel-63 has a half-life of around 100 years, so it could provide power for several decades. Nukes make people nervous, but there's not enough radioactive material in the prototype to cause a mini-meltdown -- it's comparable to the amount found in a smoke detector. Still, researchers acknowledge that they have a perception problem to overcome.

Sunshine and Good Vibes

Much less controversial are the attempts to use solar energy and the strain and vibrations produced by machinery, such as air conditioners and industrial motors, to provide enough juice to run a sensor. Even indoor lighting should provide a photovoltaic microbattery with enough power to keep a microsensor running, provided the sensors are close enough to one another that they can communicate over a faint signal.

Strain (the stretching of a material) can be converted into electricity by stimulating piezoelectric fibers, which in turn, charge a capacitor, which acts as a battery. Microstrain, a manufacturer of microminiature sensors, has developed a prototype system that uses "strain energy harvesting" to provide electrical power for a sensor.

As microsensors continue to get smaller and cheaper, alternative power schemes are starting to look more promising. To bad they won't do anything for the flashlight in your junk drawer.