Science

Understanding Bee Brains

Posted 17 Oct 2006 at 22:52 UTC by steve Share This

Even though modern computer hardware can easily give a robot computational power equivalent to an insect, no robot to date can match the intelligence of an insect. To build smarter robots we're going to have to learn what makes animals smart. A Sydney Herald article describes how Mandyam Srinivasan is trying to do that by studying bee brains. A bee brain has only about 1 million neurons, yet bees "somehow manage to store complex maps of their world, achieve amazing navigational tasks and even pass on information to other bees." They may also find room in their limited brains for implementing basic emotions, as anyone who's seen a group of angry bees can attest. For more on how this research is being applied to robotics, see the ANU Biorobotic Vision Laboratory website.


Cerebral Cortex, posted 18 Oct 2006 at 12:45 UTC by The Swirling Brain » (Master)

I suspect that just as the human nerve system is more than just the grey stuff in your scull cap but additionally your spinal cord, nerves, smart sensors, nerve wiring, etc, bees probably have more nerve cells and programming than just what's between their multi-eyes (if that's possible). Imagine a flying swirling brain (egads!). One mill neurons is probably a lot for a bee, especially since we really don't know 100% for sure how the neurons work (even though some think they know how they work 100% just like our neural net software yet that we just can't seem to get them to work as well as they do). Also, we are not sure how DNA information plays a part in the function of neurons (it actually might be like a hard coded program and a few neurons just interpret that program). There's probably a lot we don't know about brain and nervous function! And of course I don't claim to know either. At any rate, the smarts of bees is probably more than just pure neurons, but probably something else in their body that helps their intelligence along. When you think of Bee Brains, think more of scratch pad memory and slightly less of just pure program space is my guess.

I agree., posted 19 Oct 2006 at 19:50 UTC by marev » (Observer)

Bees have detailed mapping of their worlds,partly eyes but i bet mostly smell and partly getting smacked around if they do not have all the energy that the others have and seem lazy.I have mentioned before that basic little creatures,i beleive,are not the creatures to do in-depth research on or even more clever creatures,the human is the prime evolved mammal on this planet and all should go onwards from where we are only.I know it is said that to understand how we work less complex designs have to be worked out to help but not much i reckon,i.e.humans upwards and onwards to become adapted nice robotics.

Sorry... Bees don't think that way, posted 21 Oct 2006 at 02:41 UTC by Trilithon » (Observer)

I once kept bees. Lots of bees. Though they are extremely fascinating creatures, and capable of things that are extraordinary, they are still just cute little bugs. Bees do not think that heavily. Bees don't map anything visually, with the exception of thier home. Bees don't navigate by landmark, but they do by the position of the sun. Bees do understand time. That is what is most fascinating. Consider... a bee leaves it's hive... flies for 40 minutes,and lands to collect a very high-protien pollen. It spends 20 minutes there. It leaves and flys home... but into a headwind. It takes 90 minutes to get home, since it had a tailwind when it left. How does the bee figure out how to adjust? We haven't even gotten into cross-wind components, keep in mind. I guess my point is, we have no concept of bee-minds... even the mind of the bee is more complex than our understanding can come close too. It is estimated we understand about 2% of honeybee behavior, and as a beekeeper, I would agree.

Matt

Which reminds me...., posted 23 Oct 2006 at 04:18 UTC by TheDuck » (Journeyer)

Now I want to find the info relating to a show I saw on bees. It was probably Scientific American or something like that. They changed the target area for the bee in (iirc) three ways: spatial, colour, and repetition; i.e. the moved things around, then left them but they were all different colours, then made everything the same. The bees definitely either responded significantly to one condition or one condition didn't matter or something like that. Now that I have spent a lot of time providing no information whatsoever, I will say it was a fascinating study. Given that I play with genetic programming for robot control, this reminds me of a great training scenario. Thanks!

See more of the latest robot news!

Recent blogs

16 Sep 2014 shimniok (Journeyer)
5 Sep 2014 mwaibel (Master)
5 Aug 2014 svo (Master)
20 Jul 2014 Flanneltron (Journeyer)
3 Jul 2014 jmhenry (Journeyer)
3 Jul 2014 steve (Master)
2 Jul 2014 Petar.Kormushev (Master)
10 Jun 2014 robotvibes (Master)
10 May 2014 evilrobots (Observer)
2 Mar 2014 wedesoft (Master)
X
Share this page