We should wire together a cluster of bees only if the stingers can be removed. The spinoff of a bee driven super cluster is honey. But, you'll need to stoke your supercomputer with flowers. Could be interesting.
I was thinking of a way to harness the logic of a bee hive for use as a computing machine. It is not necessary to connect to the bee physically. Containing the bees would be important which could be accomplished with a compartmental tower hive. Since the logic of bee is in the path taken, a visual object path recognition program could determine their paths and then a program, decoder, would translate the logic. (no bees harmed in this process) So each bee is a like a cog processor executing a small program that can be decoded.
It is also important to identify each unique bee. I'm not sure what pattern recognition is needed for this. To another bee, it should bee no problem.
In grad school I was involved with instrumentation to study bumblebees, so they are close to my heart. For a paper on optimal foraging in relation to reward variance, kind of like this one, I build a gizmo with stepper motors driving syringes, which delivered "nectar" into artificial flowers on a plexiglass table. The syringes were controlled by a Heathkit H8 micro (1980!), so we could program in a pattern of rewards. The pattern was either uniform small rewards, or a big reward interspersed with nothing, but with the same average per visit in either case. The preference was always for uniform small rewards, although there was considerable individual variation among bees. They foraged much faster when the average reward was larger. (The same may be true for kids on halloween night!)
Theories focus on the notion of maximizing information with minimum loss of time and energy. Out in real fields, the situation faced by the foragers is complicated by the presence of other foragers and by lots of time and place variation among plants. The analogy with the travelling salesman problem breaks down pretty fast due to the uncertainties, although realistically, a travelling salesman does need to visit cities for information as well as for immediate profit.
Waddington, Allen, Heinrich; "Floral preferences of bumblebees (Bombus Edwardsii) in relation of intermittent versus continuous rewards", Animal Behaviour, 1981, 29, 779-784
iDave
Posts: 252
Comments
It is said some bees are smarter than a dog
http://www.beeclass.com/DTS/world_of_bees.htm
and they have considerable swarming intelligence.
http://en.wikipedia.org/wiki/Swarm_intelligence
.... I can see it now "Humanoido the Beekeeper!"
looking at you avatar, turning that into a bee 'hive' might not 'bee' too far of a stretch.. :smilewinkgrin:
I was thinking of a way to harness the logic of a bee hive for use as a computing machine. It is not necessary to connect to the bee physically. Containing the bees would be important which could be accomplished with a compartmental tower hive. Since the logic of bee is in the path taken, a visual object path recognition program could determine their paths and then a program, decoder, would translate the logic. (no bees harmed in this process) So each bee is a like a cog processor executing a small program that can be decoded.
It is also important to identify each unique bee. I'm not sure what pattern recognition is needed for this. To another bee, it should bee no problem.
Theories focus on the notion of maximizing information with minimum loss of time and energy. Out in real fields, the situation faced by the foragers is complicated by the presence of other foragers and by lots of time and place variation among plants. The analogy with the travelling salesman problem breaks down pretty fast due to the uncertainties, although realistically, a travelling salesman does need to visit cities for information as well as for immediate profit.
Waddington, Allen, Heinrich; "Floral preferences of bumblebees (Bombus Edwardsii) in relation of intermittent versus continuous rewards", Animal Behaviour, 1981, 29, 779-784