WISP Needs No Battery Or Cable

One of the problems with the Internet of Things, or any embedded device, is how to get power. Batteries are better than ever and circuits are low power. But you still have to eventually replace or recharge a battery. Not everything can plug into a wall, and fuel cells need consumables.

University of Washington researchers are turning to a harvesting approach. Their open source WISP board has a sensor and a CPU that draws power from an RFID reader. To save power during communication, the device backscatters incoming radio waves, which means it doesn’t consume a lot of its own power during transmissions.

The big  news is that TU Delft has contributed code to allow WISP to reprogram wirelessly. You can see a video about the innovation below. The source code is on GitHub. Previously, a WISP had to connect to a PC to receive a new software load.

Of course, RFID tags already grab power from RFID readers. However, a normal RFID tag doesn’t have any processing power or sensor inputs. The WISP has a 16-bit CPU. Development is currently on version five. There is also a version that can interface with the NFC readers common in modern smartphones and other devices.

There are also similar devices that attempt to harvest energy from ambient sources (radio and television broadcasts, for example). One big advantage to not having batteries is that the devices can be placed where they are inaccessible. For example, a building could have sensors poured into concrete that inspectors would power and read wirelessly. Obviously, being able to update such a device wirelessly would be a big benefit since you can’t retrieve the device to connect a cable (or a battery) to it.

Stealing power is not a new idea. We’ve looked at heat harvesting before, as well as a carbon monoxide sensor that steals heat from a hot water spigot. Then again, you can also use the power of the human body (no Matrix required).

10 thoughts on “WISP Needs No Battery Or Cable

  1. I hope we work through the security issues of installing wireless-powered wireless-reprogrammable sensors into buildings themselves. Else we may end up with building ‘pwners’ that hack and reprogram all your sensors to do whatever sort of bidding they want. Even if just to brick them or make them useless, that stuff is still in a very inaccessible location, by design.

    1. That is the problem with these IOT things, once they are compromised how do you find and replace them all? If they are all on the same network and the key is stolen from only one of them there is the risk that the lot need to be locked out until they are updated and what if the compromised device just pretends to update but really just steals the new key too?

  2. Yeah no, there are already NFC chips with CPUs and occasionally IO.

    The real trick would be doing this in the far-field and/or harvesting background irradiation. I suspect a tiny solar panel would be more effective, even indoors.

  3. I just completed a university ‘final year project’ based on the technology. It’s got tonnes of applications, from sensor platforms to control systems. Unfortunately, I don’t think I can share a whole lot more than that until it’s through marking and such.

  4. It’s interesting but it might rely too much on other peoples ‘RF pollution’. It might work great in city A, but not nearly as well in city B, not at all on farm C. If the responders only send when the ‘battery’ is sufficiently charged it might even work on a farm. Only the time between messages will be longer…

    As already mentioned: A small solar cell can do the same and can be more efficient. I’ve seen a few projects that combine more harvesting methods in one product. (wind, sun, temperature difference, RF, etc…)

Leave a Reply to Iestyn Stears Cancel reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.