Coming Soon: the Laptop You Power by Typing

Multiplying a couple of estimates together is hardly a deep dive into mathematics, it’s just basic numeracy. Energy harvesting is a topic where too often wishful thinking overrides realism. I have shut down projects where regrettably that realism was absent.

My figure was for the oldish laptop I’m currently using – it’s what the power gauge is reading. I could imagine the power consumption being reduced to a tenth or less, say to that of a PDA-type device, smart phone or MP3 player. This would make energy harvesting easier: now we are only three orders of magnitude away, i.e. we need to type at an average of 1,000 keystrokes per second. It helps.

However, reducing the power requirement now makes it easier to achieve satisfactory operating life using a battery: in our example, it increases battery life by a factor of ten, from 4 hours to 40 hours. A further factor of ten reduction in consumption gives a battery life of 400 hours; somewhere near this point power stops being a major issue!

Kindle-type devices with passive E Ink screens already have a battery life in weeks if not months, with no need to type continually on some Keyboard of Death.

Anyway, eReaders and iPads don’t have keyboards.

The difficulty with the laptop application is that there is not enough power going into the keyboard by typing, however it is harvested. Our fingers simply don’t deliver enough power, just a millijoule or so per keystroke. And the lower the power consumption of the computer, the more attractive (i.e. smaller, cheaper, longer lifed) batteries become.

This is the bind in which energy harvesting technology so often finds itself. On one side, the energy available in the environment tends to be rather too scarce to be useful, that is, it is too widely dispersed and therefore has to be gathered over too large an area or volume to be convenient and cost effective.

On the other hand, if the power requirement is lower, then batteries start to look more attractive. Usually they can deliver more power at less cost in a smaller package. But they do need to be recharged or changed periodically.

There are definitely niches where energy harvesting works very well, though they are quite rare. Wireless sensors are an example, as there is no wire to power the device. But the power consumption of modern chipsets can be so low that the sensor may have served its purpose or become obsolete before its battery is flat.

Anyway, the abstract refers to ‘integrated microscale energy scavenging systems’; ‘micro’ i.e. millionths, very little. I’d guess that early applications for the thin piezoelectric films that RMIT University is developing are more likely to be as integrated sensors for motion detection and other sensing, and maybe for generating a few microwatts for very low-powered circuits.