@Smeeh - thanks - a nice start. It would be nice to see some work on some of the commercial solutions out there - I have seen wifi incorporated into "toys" for $30. It would be nice to see a paper on how that is done - start to finish with a bit of information regarding economies of scale and such. This is a revolution - actualy two in having sensors (with increased resolution) in general, but multiplied many times that they will be active on a network.
Big Picture - this will allow us hopefully to cut past some assumptions of humanity although as with almost all science fiction books dictate - many social and political effects and many decisions to be made - something today's politics are illsuited to work with.
Dr. Quine: Good point re support. Can the Comcasts and AT&Ts of the world scale to handle IoT support? There will be whole new apps areas that don't even have a carrier or integrator, opening up new chalenges and opportunities.
I personally agree there will be lots more sensors. I agree lots of sensors will be wireless and will need battery or energy harvesting. Most of these sensors are part of larger equipment and will be connected via wired sensor networks which will provide both power and the digital communication bus. For instance, look at the automobile. A car today has dramatically more sensors than only a few years ago, some of them are wireless, but most are wired using multiple onboard communication networks for drive train and breaks, for climate and comforts, and for information and entertainment. Tire pressure monitoring system is a good example of wireless sensors. The process industries also see increased use of digital network communication for sensors and other devices, and dramatic increase in wireless sensors. These are primarily battery powered. Reducing the power consumption of sensors such that they can be powered by super capacitors whenever the vibration or temperature difference which harvesting relies on is not present is key for power harvesting to gain acceptance. Until then, batteries rule.
A trillion sensors (140 for each of us on earth) sounds exciting until I think about reliability and service. Today I spent an hour on the phone to customer support to deal with a Comcast cable issue and then an hour on the phone to deal with a Carbonite backup issue. One proved to be a billing system bug and the other a hard drive space management bug. In contrast, my grandparents never spent time on the phone dealing with technical support. I think that our first priority should be to build consumer goods and systems that have Six Sigma reliability (and then strive for better). The notion that we are going to deploy more buggy unreliable systems is scary (and Halloween is behind us).
Today, MicroGen's piezo-MEMS vibration energy harvester shown produces 100 to 400 microW/cm^2. In less than 3 years the micro-scale harvester's power density will increase by a factor of at least 8X.
In high volume (100M to 500M units/year) production the wafer-level packaged harvester is projected to have a cost-point between $0.47 and $0.33 each, respectfully. MicroGen's piezo-MEMS harvesters are in 200 mm production at X-FAB (www.xfab.com) in Germany.