So far, the most effective way to harvest energy for sensors is by photovoltaic; no other source comes close to that (including mechanical vibrations, piezoelectric, etc.).
TI, ST are among a host of other companies working on energy harvesting and power management solutions for low-power ubuquitous sensing applications. TI had some good presentations on this at last year's Embedded conference in the Silicon Valley.
When I think of energy harvesting, I think of the self winding wrist watch. I don't think it caused any issues for the wearer. Of course, it only contiued to work if you wore it every day. They have battery powered devices that will move the watch back and forth to keep it wound while you aren't wearing it. Kinda defeats the purpose.
As it happens, there's an article on exactly this subject from about a week ago at
Only just seen this....it mentions all the things I suggested above, glad I'm not tooo stupid.... but all of them are a long way off practical use.
Solar's great - does not seem to hassle anything or any one as far as I can see - but time-limited. So to make solar really work - 24 hours - there has got to be a better way of storage. Be that better batteries, supercapacitors, pumping water up into a storage dam during the day and using it to generate hydro power at night, whatever. Anyone fancy running an underwater power cable to/from Australia???
It may be just me, but since I was a little kid I dreamt my glasses could be better than eyes, a heads display, night vision, etc. Something like Geordi had on Star Trek. That OLED from Fraunhofer Institute seems to be a step closer. Maybe by the time my future grandchildren need glasses, we'll be there.
I wonder what kind of trouble energy harvesting is going to get us into once it's ubiquitous. There was a time was timber was thought to be an inexhaustible resource. Same with petroleum. Hydro power and wind power were once though to be environmentally friendly, but hydro messes up river ecosystems and wind messes with birds.
So many things that seem great in small volumes have proven to simply not scale well. If energy harvesting doesn't come with some albatross, then it, combine with increasingly lower powered chips, will help lead to a dramatic increase in the number of application for small scale monitoring and control.
But, that energy has to come from somewhere. Maybe it truly is wasted energy, but maybe when scaled up sufficiently, we'll just find it to be another technology with great promise but a high environmental cost.
Eye tracking device sounds fantastic. I was thinking if there was a interface, like we dont have to press any switch, just merely by thought it happens. i guess this eye tracking is very close to that.