…I tend to think a lot about heating. A few weeks back I was thinking about networking space heaters to make them capable of out-performing central heating systems (or at least work more cooperatively with them) and today I ran across a post on Hackaday that had an interesting take on the same line of thinking: heat people not buildings .
Once I saw this the thought occurred to me: could beamforming be used to improve upon this design? If you’re not familiar with beamforming it’s pretty fascinating stuff, the first I heard of it was from Mike White describing how it’s using in wireless access points to improve reception without violating signal strength limits.
I haven’t thought through the physics of it yet, so there may be some fundamental problem with the approach, but if I’m right a beamforming technique could address a number of concerns that have been raised about Leigh’s current design. A beamforming transmitter shouldn’t require the physical apparatus that is required by Leigh’s system (in fact, it might not require any moving parts at all).
Without moving parts, it may be able to overcome another concern with Leigh’s design, that is how do you use this technique to provide heat for multiple people or objects. Since a non-mechanical beamforming device can change “focus” orders of magnitude faster than a mechanical armature, heating multiple targets simply means moving quickly between them. It’s also possible that simultaneously focusing on multiple targets is possible, but I’d have to think that through some more (it might still be round-robin, but so fast that it seems instantaneous).
The other area in which I’d improve this design is by using a simpler motion capture system. Watching the demo, I wondered how the Kinect functioned with all the infrared being pumped out of the heating unit. This made me realize that the targets for this system are (or should be) warmer than their surroundings, and furthermore their IR signature could be used to measure their local temperature. With these two bits of information it should be possible to design a tracking system based on measuring radiation instead of movement that would likely be simpler, less expensive and just as (if not more than) effective as the Kinect motion capture system.
This doesn’t itself allow for personal selection of temperature per se (although I don’t know that the Kinect-only system would allow for this either without some sort of visual cue), but preferred local heating could be accomplished using some sort of device or fob held by the user, which would eliminate the need for the visual/radiant targeting system altogether (the signal from the device could be used, in this case it could use existing beamforming data used to target the device’s wifi radio :)