Bees can solve complex mathematical problems which keep computers busy for days, research has shown / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Bees can solve complex mathematical problems which keep computers busy for days, research has shown

Move Thread

LAN_403

Proud Mary

Mon Oct 25 2010, 05:11PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Bees' tiny brains beat computers, study finds

BBC News 20 October 2010

Bees can solve complex mathematical problems which keep computers busy for days, research has shown

Bees can solve complex mathematical problems which keep computers busy for days, research has shown.

The insects learn to fly the shortest route between flowers discovered in random order, effectively solving the "travelling salesman problem" , said scientists at Royal Holloway, University of London.

The conundrum involves finding the shortest route that allows a travelling salesman to call at all the locations he has to visit. Computers solve the problem by comparing the length of all possible routes and choosing the one that is shortest.

Bees manage to reach the same solution using a brain the size of a grass seed.

Dr Nigel Raine, from Royal Holloway's school of biological sciences, said: "Foraging bees solve travelling salesman problems every day. They visit flowers at multiple locations and, because bees use lots of energy to fly, they find a route which keeps flying to a minimum."

Using computer-controlled artificial flowers to test bee behaviour, he wanted to know whether the insects would follow a simple route defined by the order in which they found the flowers, or look for the shortest route.

After exploring the location of the flowers, the bees quickly learned to fly the best route for saving time and energy.

The research, due to appear this week in the journal The American Naturalist, has implications for the human world. Modern living depends on networks such as traffic flows, internet information and business supply chains.

"Despite their tiny brains bees are capable of extraordinary feats of behaviour," said Raine. "We need to understand how they can solve the travelling salesman problem without a computer."

Bjørn

Tue Oct 26 2010, 02:22AM

Registered Member #27 Joined: Fri Feb 03 2006, 02:20AM
Location: Hyperborea
Posts: 2058

A bee got about a million neurons, if every neuron is connected to 1000 other neurons and is updated 40 times a second we get about 80 GFLOPS which is comparable to a modern PC. So the claim is that a bee does computation on an NP-hard problem several thousand times faster than is possible with classical computing using the best known algorithms. Sounds like must have a quantum brain.

Or maybe it is a case of inflating a completely valid but boring experiment so far that it bursts and all the folly spews out all over the place.

Lethal Shot

Tue Oct 26 2010, 03:59PM

Registered Member #1623 Joined: Tue Aug 05 2008, 03:31PM
Location: The Netherlands
Posts: 39

I'm always fascinated by the tremendous feats that can be achieved by neuron cells.
As Bjorn told a bee has 'only' 80 GFLOPS, but with this raw power it can control its body, process image and solve complex things as the salesman problem.
Try finding a computer that does this.

The experiment with the rat brain flying a jet plane also intrigues me.

IamSmooth

Tue Oct 26 2010, 08:12PM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

I'm still waiting for some Shakespears manuscripts from a monkey...

Nicko

Wed Oct 27 2010, 05:20AM

Registered Member #1334 Joined: Tue Feb 19 2008, 04:37PM
Location: Nr. London, UK
Posts: 615

Proud Mary wrote ...

Bees' tiny brains beat computers, study finds
...
"Despite their tiny brains bees are capable of extraordinary feats of behaviour," said Raine. "We need to understand how they can solve the travelling salesman problem without a computer."
...

I'll look forward to the main article - as a beekeeper myself, I've been fascinated by their behaviour since I was a child - In SE London there is an ethnographic museum (Horniman's) which used to have an observation hive with a glass side panel set into the outside wall of the museum. You could sit inside and just watch them for hours...

What puzzles me (!!) about all this is that bees have very complex foraging habits - once on a particular source of nectar, an individual worker will tend to stick to that. What's more, different nectar sources (typically different types of flowers) will have their nectar flows at differing times of the day, so a given worker tends to fly at the same time(s) each day and to the same flowers of a given type.

I'd be very interested to know how accurate the solution they provide is...

Cheers

Bjørn

Wed Oct 27 2010, 06:21AM

Registered Member #27 Joined: Fri Feb 03 2006, 02:20AM
Location: Hyperborea
Posts: 2058

The original paper was not available when I looked for it so who knows what it really says.

In general living creatures does not solve anything, they approximate to a level that is good enough for the purpose. There are two main reasons for that, for one the energy required to get a perfect solution is often much higher than a simple approximation, the other and most important reason is that a perfect solution is usually far less general and is of less practical use in the real world.

The 80 GFLOP number is just a wild approximation, some people have sliced up the brains and counted every neuron and interconnection and some mention numbers as low as 1 GFLOP but since the exact way neurons process information is not exactly known it is not a very reliable number.

It is interesting to make simulations with simple creatures and see how a few simple rules will generate very complex behaviour in a population.

Proud Mary

Wed Oct 27 2010, 08:06AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Here is the citation and abstract of the paper:

Am Nat 2010. Vol. 176, pp. 000â€“000

0003-0147/2010/17606-52035 $15.00
DOI: 10.1086/657042

Travel Optimization by Foraging Bumblebees through Readjustments of Traplines after Discovery of New Feeding Locations

Mathieu Lihoreau,1

Lars Chittka,1 and

Nigel E. Raine1,2,*

1. Research Centre for Psychology, School of Biological and Chemical Sciences, Queen Mary University of London, Mile End Road, London E1 4NS, United Kingdom;

2. School of Biological Sciences, Royal Holloway University of London, Egham, Surrey TW20 OEX, United Kingdom

Abstract:

Animals collecting resources that replenish over time often visit patches in predictable sequences called traplines. Despite the widespread nature of this strategy, we still know little about how spatial memory develops and guides individuals toward suitable routes. Here, we investigate whether flower visitation sequences by bumblebees Bombus terrestris simply reflect the order in which flowers were discovered or whether they result from more complex navigational strategies enabling bees to optimize their foraging routes. We analyzed bee flight movements in an array of four artificial flowers maximizing interfloral distances. Starting from a single patch, we sequentially added three new patches so that if bees visited them in the order in which they originally encountered flowers, they would follow a long (suboptimal) route. Beesâ€™ tendency to visit patches in their discovery order decreased with experience. Instead, they optimized their flight distances by rearranging flower visitation sequences. This resulted in the development of a primary route (trapline) and two or three less frequently used secondary routes. Bees consistently used these routes after overnight breaks while occasionally exploring novel possibilities. We discuss how maintaining some level of route flexibility could allow traplining animals to cope with dynamic routing problems, analogous to the wellâ€known traveling salesman problem.

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob