[linuxola] Human power for computers

[Theo Schmidt]

Human power has been mentioned several times on this list (Linuxola) for 
powering computers. This is an important subject for me, as I am editor 
of "Human Power electronic Journal" (HPeJ), available at 
www.hupi.org/HPeJ .  This is mostly about human powered vehicles (HPVs), 
but also to some extent about the stationary use of human power.

Human power systems are successful if most of the following conditions
are met:

- effective
- efficient
- comfortable
- practical
- sporty or trendy
- competition
- community
- fun
- traditional
- no or poor alternatives
- rewarding
- inexpensive

Most of these apply to cycling and simple hand tools, and these are
therefore widespread. Especially bicycles are very popular, as shown by
a survey by BBC Radio 4:
http://www.bbc.co.uk/radio4/youandyours/technology_launch.shtml
This is because the above factors are mostly met and especially
"rewarding" is high. If you pedal harder, you go faster, considerably
faster than walking. Also, cooling of the human body is very good; it is
possible to cycle comfortably in 40° temperature (on the level, slowly,
not in high humidity).

Stationary uses of human power are more difficult because the rewarding
and cooling factors are less. I have a little radio which you can crank
for a minute and listen for half an hour. Cranking more than one minute
on the tiny crank is however unpleasant, so I tend to use the radio's
solar cells (one minute exposure to the sun gives two minutes
listening). The crank is however very useful if the battery charge is
exhausted when it is dark. I suspect the well publicised 100$ laptop 
with hand crank will only be acceptable if combined with another power 
source.

Now, properly designed generator pedalling systems are perhaps 100 times
more powerful than the tiny hand cranks on small appliances. However,
they require a sturdy structure, space, and quality components and are
thus not cheap. There is usually no fan, so any appreciable activity
results in much sweating. Such devices are thus so far only used in
emergency situations, by athletes in training, or by people who do not
get enough movement in daily life. (It is hoped that African culture
will not stoop to the northern-western stupid practices of driving by 
car to a fitness center, or using petrol grass mowers in typically tiny
gardens, where a human-powered push mower is better in most respects.)

Two friends of mine have examined human power generators, Designer Ulo
Gertsch and Researcher Andreas Fuchs 
http://www.hta-be.bfh.ch/~fuchs/Transmission/ .  Gertsch also examined 
recumbent pedal generators as a source of earning money. This doesn't 
work, at least not here, because electric energy is extremely cheap and 
food (required for strong pedalling) is much more expensive, apart from 
other things.

So, what basis is there for expecting anybody to use a pedal generator 
in a Linuxola project, except in a temporary emergency situation?

I think it is possible if done cleverly. The first is to not require too 
much power. Although a person can pedal briefly at up to 2 kW (kilo 
Watts, see http://de.geocities.com/manfred43_99/guinness.html ), maximum 
power sustained comfortably by most people will be 50-100 W. This is not 
enough to power a typical recycled Thin Client and CRT Monitor, which 
togther might need 200W. It is ample to power a Laptop or a modern Thin 
Client and LCD Monitor. So, unless we use Laptops or LCD monitors, we 
need more pedallers than computer users. A possible arrangement might be 
three seats with pedals per workstation. Another might be to have a 
whole school class or group drive a large turnstile like in the playpump 
in South Africa: http://www.roundabout.co.za/main.htm .

A further advisable element would be an energy accounting system: 
sensors would measure the energy produced by each person and allocate 
the corresponding amount of computer time. This would introduce an 
element of sport and competition: strong males could compete to maximise 
their computer time or show off to girls, thus utilising the factor 
"rewarding".

Another way would be to run the computers off another power supply, but 
power the monitors with individual generators designed into the computer 
workstation (now earning its name in a double sense!). The monitors 
could use a control circuit to stay active or in standby even when not 
pedalling, but the screen to become visible with minimal peadlling and 
brighter with more pedalling. This would also exploit the "reward" 
factor and would presumably work well and feel good.

Now, the two schemes above don't really exist yet and would require 
quite some engineering, certainly above our present resources. An ideal 
project at a college or university, perhaps.

Theo Schmidt


cc to mailing list futurebike