New sensor technology may give us new ways to look at the world, but sometimes making existing high end technology readily available can have a much bigger impact. Originally developed with funding from the NSF and the US Department of Homeland Security, radio tomographic imaging is moving into the home.
Radio tomography makes use of mesh networks of radio devices to detect motion. In a mesh network each radio in the network is in contact with every other radio, forming a web of links between the devices. The strength (attenuation) of the radio signal along each of these links can be measured, creating a “radio image” of the environment. Here the radios themselves are the sensors.
The way this image is constructed is similar to the Computed Tomography (CT) scans used by hospitals for imaging — where X-rays are used to build up a detailed picture of the body — or Doppler Tomography which is used by astronomers to create images of accreting binary star systems.
However due to the (relatively) small number of transmitters and receivers in the network, and since the line-of-sight distance between them is in most cases unknown, the “radio image” built up isn’t particularly useful. However as people, or other things, move around the space the attenuation of the radio signal between the mesh networked radios changes. Any movement in the space is immediately obvious.
However unlike a PIR sensor it’s not just movement that is detected. Since any moving object in the space will intersect with more than one ‘path’ between the radios in the network, the position of the person (or object) in the space can be found — in a way this works in a similar manner to the cell tower and Wi-Fi positioning used everyday by our cellphones.
For some years this sort of technology has been the preserve of the military but, based off their original work while at the University of Utah using off the shelf ZigBee networks, the founders of XANDEM have now packaged radio tomography up for the home user.
While not hackable at the hardware level, at least not without breaking your warranty, the system is software hackable as it comes with a REST API. That means you can write your own software to talk to hardware — the XANDEM web application is entirely built on top of the same API — bringing real-time motion and tracking information into your project. There are also hooks to make the system notify third party endpoints, like IFTTT’s Maker Channel, when motion is detected.
Importantly, and unlike a lot of the Internet of Things, the XANDEM system is useful without its optional cloud service. While you can connect it to the cloud, it doesn’t have to. Your project can talk directly to the XANDEM hardware. Which, at least to Makers, is going to be crucial.
While XANDEM is marketing this as a home security system to replace cameras — which especially for home use can be seen as very invasive of your privacy — this system isn’t just limited to security applications.
For instance, this system could be used to detect and map movement of a robot, or even a large number of robots. Instead of individual computerised vision systems on each robot, the XANDEM system could provide vision-as-a-service to a collection of robots moving through the monitored space. It could serve as a building block for cooperation between robots in a swarm, because by using it, not only does an individual robot know where it is, but it also knows where the other robots in the swarm are, and any humans that stray into the space.
If you’re interested in taking a look at the XANDEM system it’s currently crowdfunding on Indiegogo, and has almost reached its funding target with still more than two weeks to run on the campaign.