Thursday, October 19, 2006
Yesterday I attended a seminar by Professor Mandyam V. Srinivasan, who has just won Australia's top science award, the Prime Minister's Prize for Science, with a prize money of $300,000. The award was given for his research into the insect mind which has helped redefine robotics.
The following is a description of the seminar
___________________________________________________________________________________
Speaker: Professor Mandyam V. Srinivasan
Centre for Excellence in Vision Science, Research School of Biological Sciences, ANU
Title: "More than a Honey Machine"
Abstract:
Anyone who has watched a fly make a flawless landing on the rim of a
teacup, or marvelled at a honeybee speeding home after collecting
nectar from a flower patch several kilometres away, would know that
insects possess visual systems that are fast, reliable and accurate.
Insects cope remarkably well with their world, despite possessing a
brain that carries fewer than 0.01% as many neurons as ours does.
This talk will describe research aimed at understanding the
mechanisms underlying visual perception, navigation, learning, memory
and ìcognitionî in honeybees. Opportunities for incorporating insect-
based principles into the design of novel, autonomous robots will
also be discussed.
___________________________________________________________________________________
The hour-long seminar was extremely interesting and stimulating. It was divided into two sections, the first about his group's research into how bees navigate and communicate, and the second about how they used the results in computer vision and robotics.
I find the first part in particular very interesting. Prof. Srinivasan's group was intrigued by how insects, and in particular bees, determine the distance of a landmark because unlike humans and other animals, they do not have stereoscopic eyes. With very small brains, how these insects make their way home or towards a food source and navigate around obstacles is truly amazing.
The first thing they discovered was that bees use the speed at which a landmark moves as they fly to determine how far the landmark is from them. We can also judge how far an object is from us if we are moving. Imagine that we are driving. We know that a tree that quickly becomes larger in our view is close to us, whereas the hill which doesn't appear to move at all is far from us. That's also how bees determine the distance. What this also means is that they can't determine the distance if they are not moving.
Another experiment that they did was to determine how bees can fly exactly in the middle of a tunnel. The way they do this is basically by balancing the image flow rate from both sets of eye (remember bees, flies, insects have groups of tiny eyes on each side). To test this hypothesis, they purposely move one side of the tunnel wall at a certain speed while keeping the other side of the tunnel fixed. If they move one wall, say wall A, at a certain speed in the direction opposite to the motion of the bee, they would find the bee flying closer to the other wall, wall B. Vice versa, if they move wall A in the same direction of the bee motion, they find that the bee will fly closer to wall A. The reason is that in the first situation, the opposite motion of wall A causes the bee to think that wall A is closer to it, due to the higher image flow rate. Therefore it steers away from wall A and fly closer to wall B. Whereas in the second situation, wall A moves in the same direction as the bee, so the bee thinks that wall A is farther away, therefore it moves closer to wall A as it flies through the tunnel.
They also find that bees like to maintain a certain speed while flying based on the image flow rate. In the same tunnel where they did the above experiments, the researchers first determine the speed of the bee using cameras. They then move both sides of the wall at a certain speed in the opposite direction of the bee motion. And guess what? The bee slows down. When they move the walls in the same direction as the bee, the bee speeds up. They discovered that the relative speed betwen the bee and the motion of the wall is approximately the same.
The bees' sense of direction is extremely good. You've probably heard of the saying 'to make a bee line', which is quite true. Once the bees find a food source, they can head straight to their hive. They also find that these bees communicate with the others using a 'wiggle dance', where they do a certain 'dance' motion. The duration of the dance determines how far the food source is from the hive. This relationship is fairly linear. And they can also communicate the direction of the food source in this dance.
The main reason why I enjoyed this seminar was that the experiments and ideas that they come up with were very simple, yet very effective. But of course, Prof. Srinivasan has been working in this area for about twenty years. But this sort of applied research is what I prefer, making a new invention or an application based on what you have discovered in your research.
The following is a description of the seminar
___________________________________________________________________________________
Speaker: Professor Mandyam V. Srinivasan
Centre for Excellence in Vision Science, Research School of Biological Sciences, ANU
Title: "More than a Honey Machine"
Abstract:
Anyone who has watched a fly make a flawless landing on the rim of a
teacup, or marvelled at a honeybee speeding home after collecting
nectar from a flower patch several kilometres away, would know that
insects possess visual systems that are fast, reliable and accurate.
Insects cope remarkably well with their world, despite possessing a
brain that carries fewer than 0.01% as many neurons as ours does.
This talk will describe research aimed at understanding the
mechanisms underlying visual perception, navigation, learning, memory
and ìcognitionî in honeybees. Opportunities for incorporating insect-
based principles into the design of novel, autonomous robots will
also be discussed.
___________________________________________________________________________________
The hour-long seminar was extremely interesting and stimulating. It was divided into two sections, the first about his group's research into how bees navigate and communicate, and the second about how they used the results in computer vision and robotics.
I find the first part in particular very interesting. Prof. Srinivasan's group was intrigued by how insects, and in particular bees, determine the distance of a landmark because unlike humans and other animals, they do not have stereoscopic eyes. With very small brains, how these insects make their way home or towards a food source and navigate around obstacles is truly amazing.
The first thing they discovered was that bees use the speed at which a landmark moves as they fly to determine how far the landmark is from them. We can also judge how far an object is from us if we are moving. Imagine that we are driving. We know that a tree that quickly becomes larger in our view is close to us, whereas the hill which doesn't appear to move at all is far from us. That's also how bees determine the distance. What this also means is that they can't determine the distance if they are not moving.
Another experiment that they did was to determine how bees can fly exactly in the middle of a tunnel. The way they do this is basically by balancing the image flow rate from both sets of eye (remember bees, flies, insects have groups of tiny eyes on each side). To test this hypothesis, they purposely move one side of the tunnel wall at a certain speed while keeping the other side of the tunnel fixed. If they move one wall, say wall A, at a certain speed in the direction opposite to the motion of the bee, they would find the bee flying closer to the other wall, wall B. Vice versa, if they move wall A in the same direction of the bee motion, they find that the bee will fly closer to wall A. The reason is that in the first situation, the opposite motion of wall A causes the bee to think that wall A is closer to it, due to the higher image flow rate. Therefore it steers away from wall A and fly closer to wall B. Whereas in the second situation, wall A moves in the same direction as the bee, so the bee thinks that wall A is farther away, therefore it moves closer to wall A as it flies through the tunnel.
They also find that bees like to maintain a certain speed while flying based on the image flow rate. In the same tunnel where they did the above experiments, the researchers first determine the speed of the bee using cameras. They then move both sides of the wall at a certain speed in the opposite direction of the bee motion. And guess what? The bee slows down. When they move the walls in the same direction as the bee, the bee speeds up. They discovered that the relative speed betwen the bee and the motion of the wall is approximately the same.
The bees' sense of direction is extremely good. You've probably heard of the saying 'to make a bee line', which is quite true. Once the bees find a food source, they can head straight to their hive. They also find that these bees communicate with the others using a 'wiggle dance', where they do a certain 'dance' motion. The duration of the dance determines how far the food source is from the hive. This relationship is fairly linear. And they can also communicate the direction of the food source in this dance.
The main reason why I enjoyed this seminar was that the experiments and ideas that they come up with were very simple, yet very effective. But of course, Prof. Srinivasan has been working in this area for about twenty years. But this sort of applied research is what I prefer, making a new invention or an application based on what you have discovered in your research.
Comments:
Post a Comment
