Is that 12W continuous, or just 12W when the motors are moving the head? It seems to me that it should be possible to run an audio interface and two cameras on 2.5W (theoretical max USB 2.0 power). A DSP performing acoustic echo cancellation and depth calculation and a bright IR emitter could add another few watts, but I don't think that accounts for 12W.

I've got a Kinect at my desk, and the power supply is rated for 12V at up to 1.08A, or ~13W total. My Kill-a-watt shows it drawing ~3W at idle, ~4.4W running just depth, and 4.6W running depth and RGB. Running depth, RGB, and the motor only gets me up to 5W.

The Kinect power supply lists 100V-230V at 0.3A. Using the minimum voltage, that's a peak draw of 30VA. The power supply has a measured power factor of ~0.44, which if maintained to full load means it could draw .44*30=13.2W. Somehow I doubt the power supply is actually 98% efficient at its normal operating range, so in a laptop where a 12V supply is already available, the additional power draw of a complete depth sensing system could be under 4W.

To sum up, I think you could successfully integrate depth sensing into a laptop that already has a camera and microphone array, especially considering that the sensing field will be much smaller. Please correct my estimation if I've made any errors.

Imagine walking up to your desktop and it recognises who you are and automatically unlocks? Apple has been rumoured to be working on something like this.

There are a multitude of ways this technology can be applied, and just because you can't envisage a way to use them - it doesn't make them any less useful.

"No wireless. Less space than a Nomad. Lame."

That technology already exists, and can be used on any computer with a webcam, Kinect would be overkill: http://www.keylemon.com/product/

With depth recognition it would be more difficult to put a picture of someone's face in front of the webcam to unlock the computer.

True, but it is interesting what little soap, toilet paper, and hair can do... http://en.wikipedia.org/wiki/June_1962_Alcatraz_escape#Escap...

Apple TV maybe? Not built-in but as an add-on controller for the device and games/apps?

It would make sense for the current incarnation of AppleTV, as it runs iOS and it's driven by Apple's A4 chip. Technically it could run iOS games, but it's missing a direct intuitive interface (like multitouch on other iOS devices). Kinect-like tech could bridge the gap, and I wouldn't be surprised if Apple came up with something similar.

However, back in June of 2008, the AppleTV was a very different product than it is now. Its innards were more like a bare bones Mac Mini than a iPad. And aside from the architectural differences between Intel and ARM, the App Store was yet to open - Apple could not have predicted iOS becoming such an immensely popular game platform.

I'm not sure how small they can make the Kinect hardware - I doubt Steve would approve of an add-on more than 6 times the size of the Apple TV.

The image on this page looks like it's about a third of the size of the Kinect:

http://www.primesense.com/?p=487