Using a squirrel to control conveyors

blinman.com
First place might not be an option either. The most sophisticated computers on earth have less brainpower than an insect. (So do some of our relatives, but we only have to see them at Christmas.) We are restricted to using the hardware which is currently available. Such as a squirrel. I have a number of grounds for concern:

Squirrel 1) Optimising the squirrel's brain for the control of a medium sized rodent took several hundred thousand years. Given the peripheral nature of the synergies to be exploited between running through trees looking for nuts and dispatching packets of cornflakes to sundry locations in North London, I foresee that a similar development process will be required for the new application. Will there be programme implications?

2) Our major customers, the supermarkets, are at great pains to stress the animal-friendly nature of their operations. Testing products on animals is forbidden. Poisoning animals to grow vegetables is fast becoming a serious social drawback. (Poisoning the animals before growing the vegetables is fine, however, as long as you call it "Organic Farming.") Making products from animals is possible only when the creatures are treated better than long-haul airline passengers or supermarket checkout operatives. Chaining a squirrel to a conveyor support and sticking electrodes into its brain in order to optimise the efficiency of certain distribution channels may therefore present marketing issues.

3) Powercuts, communication breakdowns and some hardware failures are accepted as reasons for non-optimal control system performance. I do not want to be the person responsible for explaining that a countywide shortage of cornflakes was caused by a sexy squirrel breaking into the control room.

 

Why Nigel was moved to defend animal-kind (an email from a colleague):
We bear the responsibility to push boundaries in order to be the best.
Second place is not an option.
A periphery that needs momentum imparting is behaviour prediction and feedback based on non-logical computation. As systems get more complex, 
10 + 01 = 11 does not suffice as a method of control.

The brain of a squirrel is many times more complex than the control system of a high speed materials handling system. If we could only connect a squirrel to the conveyor hardware, it would be able to control it with ease, pre-empt situations, optimise and find short-cuts to convoluted situations almost by instinct, and still have time to sniff the air for squirrels of the opposite sex.
Surely we can aim to beat the squirrel, after all this is the age of liquid cooled silicon computing power that can FEA the entire space shuttle in a fraction of a second, at a cost less than a family hatchback.

I trust your commitment to this cause will produce fascinating and diverse schemes.

This page is pure drivel. There are much better things on blinman.com to look at.