cover andrew philippides

Andy Philippides on insect navigation and ant vision

  • cover play_arrow

    PLAY EPISODE


cover andrew philippides
Season 2012
Season 2012
Description arrow_drop_down

Description

How does an ant with a brain smaller than a pinhead navigate miles of desert using visual memories that would be unrecognizable to a human eye? Andy Philippides reveals the elegant simplicity of insect navigation and why it could outperform GPS-dependent robots in denied environments.

Subscribe for more from the Convergent Science Network podcast series.

Philippides explains why studying ants in their natural environment is essential: laboratory stimuli produce fundamentally different neural responses than the real world. Desert ants like Melophorus bagoti are ideal subjects because they are social foragers that learn routes in a single trial, their behavior gives a direct readout of their nervous system, and researchers can track their entire foraging range. Crucially, ants do not use cognitive maps. Their route memories are insulated by context, so an ant placed mid-path while fed will head home, while an empty ant placed at the same spot will head outward.

The interview dissects the complementary navigation strategies ants employ. Path integration, combining step counting with a polarized-light compass, provides a baseline homing vector but accumulates errors over distance. Ants compensate by deliberately aiming to one side of the nest, much like sailors using dead reckoning would aim to one side of a port. From the very first trip, ants layer visual memories on top of path integration, using the skyline silhouette of trees against the sky as a robust, stable landmark. Philippides describes his group’s visual compass model, where ants store panoramic snapshots oriented toward their goal and recover heading by rotating on the spot to minimize image difference, a strategy supported by observed saccadic scanning behavior in unfamiliar environments.

The computational implications are striking. With poor-resolution compound eyes, minimal memory, and limited processing power, ants achieve remarkably robust navigation. Philippides argues these bioinspired algorithms could serve UAVs, space exploration, and any platform where GPS is unavailable and computational resources are constrained. Taking panoramic images from ant-level positions has already revealed how radically different the visual world appears at ground level, where most of the visual field is sky and ground, and prominent landmarks simply disappear into the background.

Tagged as:

About the author call_made

CSN Podcasts

Both the triumphs of humanity and its most evil deeds have resulted from collaboration. In a time where humanity is required to aspire to the former and minimize the latter, the question arises of how collaboration arises and why it fails. Surprisingly, this phenomenon, so central to who we are, is not well understood. Hence, a collaborative effort is required to understand collaboration in its full biological, psychological, sociological, cultural, and economic complexity and to translate this understanding into operational impact. This series of podcasts is one step toward achieving these complementary goals. The Collaboration Podcast presents interviews with people who are central orchestrators of collaboration in various domains including business, government, science, art, health, sustainability, and the military. The discussions were conducted by Prof. Dr. Paul F.M.J. Verschure and members of the Program Advisory Committee of the Ernst Strungmann Forum on Collaboration (https://www.esforum.de/forums/ESF32_Collaboration.html) during 2021 and had the goal to sketch a map of opportunities, challenges, and obstacles in human collaboration. The forum took place in May 2022, and now we would like to share this series of interviews with a broader audience. The full report of the Forum will be published in 2023 by MIT Press. The podcast was produced by the Convergent Science Network (https://www.convergentsciencenetwork.org/). Context: The stability of social systems depends critically on realizing sustainable methods of “collaboration,” yet how and by which means collaboration is achieved is not clearly understood; neither are the conditions or processes that lead to its breakdown or failure. Collaboration can be understood as cooperation between agents toward mutually constructed goals. Part of the reason for our lack of understanding is that the phenomenon of collaboration is, by nature, a highly multidisciplinary problem, and effective research into its complexities has been difficult to achieve across the broad range of scientific and technical disciplines involved. The need for a fundamental understanding of collaboration, however, has become increasingly important. Not only does humankind demand answers as it attempts to address critical challenges at multiple scales (e.g., climate change, migration, enhanced automation, social and economic inequality), but ever-increasing technological and economic means of interconnecting people and societies are disrupting long-established, familiar patterns of how we interact. Radical technological changes that are ongoing have the potential to reshape collaboration in ways that are currently hard to predict or influence (e.g., by altering configurations in interaction, information creation, and modes of communication). On one hand, such changes could disrupt hitherto stable forms of collaboration by affecting critical communication channels and traditional roles, as can be observed in the rapidly changing patterns in governance, commerce, and social interaction. Conversely, technology could lead to the emergence of novel, successful forms of collaboration that deviate from traditional “hierarchical” architectures. Evidence of this can be seen in areas as diverse as highly automated manufacturing plants, the open science movement, collaborative software repositories, user-centered services, and the sharing of economy-based modes of organization. Without a fundamental understanding of the mechanisms, processes, and boundary conditions of collaboration, it is not possible to evaluate or predict which of these possible scenarios are sustainable or even plausible. The Forum “How Collaboration Arises and Why it Fails” (May 8–13, 2022, Location: Frankfurt am Main, Germany) Chairs: Andreas Roepstorff and Paul Verschure Program Advisory Committee: Jenna Bednar, Julia R. Lupp, Bhavani R. Rao , Andreas Roepstorff, Ferdinand von Siemens, and Paul Verschure

More posts

Timestamp

  • fast_forward00:00:03 - This is the Convergent Science Network podcast. Leading researchers in the domain
  • fast_forward00:00:10 - of neuroscience, brain theory and technology are interviewed by Paul Verscher and Tony Prescott.
  • fast_forward00:00:20 - Hi, this is Tony Prescott for the Convergent Science Network podcast from the
  • fast_forward00:00:25 - Barcelona Summer School on Cognition, Brain and Technology. and I'm talking
  • fast_forward00:00:29 - to Andy Filippides from Informatics at the University of Sussex.
  • fast_forward00:00:35 - Andy, your group works with ants and other insects looking at their behavior
  • fast_forward00:00:43 - in the natural environment and trying to understand the relationship between
  • fast_forward00:00:48 - the environment that they live in and the behavior that they have.
  • fast_forward00:00:51 - Now, why do you think it's important to study animals in their real environment
  • fast_forward00:00:57 - and not just in the laboratory?
  • fast_forward00:00:59 - Well, I think you can't do research in the absence of laboratory experiments,
  • fast_forward00:01:06 - but you have to test the same behaviors in the real world wherever possible,
  • fast_forward00:01:13 - primarily because insects,
  • fast_forward00:01:16 - ants, animals, neurons often respond very, very differently when you have natural stimuli.
  • fast_forward00:01:23 - And so we need to be able to see what the natural behavior is.
  • fast_forward00:01:29 - There are a lot of anecdotal evidence for this, particularly recently with optic
  • fast_forward00:01:36 - flow, where people traditionally use very strong optic flow signals and the
  • fast_forward00:01:41 - models work in a certain way with these very strong optic flow signals.
  • fast_forward00:01:44 - And when they've recently challenged these models with optic flow signals,
  • fast_forward00:01:49 - as an animal would perceive moving through the real world, you get very different results.
  • fast_forward00:01:55 - So would these be experiments with insects?
  • fast_forward00:01:58 - This is experiments. I think they recorded the video from cats moving through
  • fast_forward00:02:03 - the undergrowth, and it's just putting them into sort of optic flow-based models.
  • fast_forward00:02:07 - So with ants, we typically, in the lab,
  • fast_forward00:02:11 - uh we use as blank
  • fast_forward00:02:14 - an environment as possible so that the only um visual objects they can use are
  • fast_forward00:02:19 - the ones that we've put put in um and this clearly isn't a natural um a natural
  • fast_forward00:02:27 - environment so when we're and we typically train into a vertical edge for instance and so um,
  • fast_forward00:02:35 - Whilst we can get a lot of information that way, this is not the sort of thing they usually attend to.
  • fast_forward00:02:41 - So, for instance, Paul Graham's recent work showing that they used the whole
  • fast_forward00:02:45 - panorama wouldn't be...
  • fast_forward00:02:47 - That wouldn't work in the lab because you would have to reconstruct an artificial
  • fast_forward00:02:52 - panorama that doesn't move in the same way as a real panorama does.
  • fast_forward00:02:58 - And primarily it's got a very different distance distribution of objects.
  • fast_forward00:03:02 - So this is looking at the visual navigation capability of an ant. Yeah, basically.
  • fast_forward00:03:07 - So why would people be particularly interested in the ant as a system in which
  • fast_forward00:03:11 - to study navigation behaviour?
  • fast_forward00:03:13 - Well, firstly, because they're very good at it.
  • fast_forward00:03:16 - Secondly, because they're social insects and social foragers,
  • fast_forward00:03:25 - they go out multiple times in the day.
  • fast_forward00:03:28 - So practically, they're very good. they go out to get food to give it back to
  • fast_forward00:03:31 - the colony um and then go out and forage again so you can train them quickly
  • fast_forward00:03:35 - um they learn in one trial they've got lots of interesting behaviors.
  • fast_forward00:03:42 - Um the more interesting reason i think is that they're we'd argue they're.
  • fast_forward00:03:49 - One of the most complex animals you can study in which you can study them over the course of the
  • fast_forward00:03:54 - whole foraging range so it is possible to
  • fast_forward00:03:57 - track an ant when it leap from where it needs its
  • fast_forward00:04:00 - nest to when it finds food and when it comes back and there
  • fast_forward00:04:03 - is some work ongoing in australia to track certain ants um through the course
  • fast_forward00:04:08 - of their whole life until one gets bored but at least from so that one can have
  • fast_forward00:04:13 - a record of that everything everywhere they've been and all the visual experience
  • fast_forward00:04:18 - all the visual um visual input they might have experienced
  • fast_forward00:04:22 - so our ants navigation
  • fast_forward00:04:25 - capabilities are simpler than mammals safe
  • fast_forward00:04:29 - or just specialized well i think
  • fast_forward00:04:32 - they're very good at what they do they're different
  • fast_forward00:04:36 - because they've got compound eyes so there
  • fast_forward00:04:39 - are certain things that uh we would do differently their eyes
  • fast_forward00:04:42 - are certainly a lot worse than ours um they don't
  • fast_forward00:04:46 - use maps and we do use maps
  • fast_forward00:04:49 - i think it's been pretty much categorically shown that they characterically shown
  • fast_forward00:04:52 - they don't use cognitive maps um and
  • fast_forward00:04:56 - so for instance um if you train an ant to go out to find food and then come
  • fast_forward00:05:04 - back it'll come back along the same path um if you then uh place the ant back
  • fast_forward00:05:10 - on that path when it's fed.
  • fast_forward00:05:14 - Um at the middle of the path it's getting it's got exactly the same visual input
  • fast_forward00:05:17 - that it would have got for going out or coming back and it will just return
  • fast_forward00:05:20 - home um if you place it where it's empty it will go out to the to the goal so
  • fast_forward00:05:25 - this kind of indicates that the memories are kind of insulated from each other
  • fast_forward00:05:29 - there's been various quite nice experiments that's shown that that,
  • fast_forward00:05:33 - their memories their root memories are kind of insulated from one another by context and they
  • fast_forward00:05:38 - can't really share across them so by um not
  • fast_forward00:05:42 - having a cognitive map yep they are in some
  • fast_forward00:05:45 - sense limited to doing homing behaviors essentially or
  • fast_forward00:05:49 - or are they able to learn multiple paths they learn
  • fast_forward00:05:53 - multiple paths they learn multiple paths and it's likely if their multiple paths
  • fast_forward00:05:59 - are the same well it's possible our model would be that if they multiple paths
  • fast_forward00:06:02 - to the same food source then it's maybe stored as one root but certainly it would appear that,
  • fast_forward00:06:11 - outbound roots and nest bound roots are insulated from one another and they
  • fast_forward00:06:15 - would be primed by the context of being fed and so they could well be primed
  • fast_forward00:06:19 - by other things um you know they may well forage at different places at different
  • fast_forward00:06:24 - time of day such as bees would um but yeah i think.
  • fast_forward00:06:32 - I mean, the other reason that we study them is because they are specialists.
  • fast_forward00:06:39 - And when they want to return home having found food, all they care about is getting home.
  • fast_forward00:06:46 - And Tom Collett has a very nice phrase where he says that their behavior gives
  • fast_forward00:06:50 - a direct readout of their nervous system.
  • fast_forward00:06:52 - Them and so without doing
  • fast_forward00:06:55 - something invasive and difficult to do in the field you
  • fast_forward00:06:59 - can see exactly well you can see what the ant it
  • fast_forward00:07:02 - thinks it's doing yeah i think that that is really useful because one of the
  • fast_forward00:07:07 - biggest problems when in neuroethology i think of studying an ethology of studying
  • fast_forward00:07:12 - animals in the field is that it's very hard to know what the intentions of the
  • fast_forward00:07:16 - animal are and therefore to have much insight into what it might be trying to do,
  • fast_forward00:07:22 - never mind from what it's actually doing. Yeah, I think that's right.
  • fast_forward00:07:28 - That is primarily all they want to do. So you know exactly what its intention is.
  • fast_forward00:07:35 - There aren't any distractions. And so, yeah.
  • fast_forward00:07:38 - So we know that one of the mechanisms that they use is to do path integration, a kind of step counting.
  • fast_forward00:07:44 - And they also have some kind of compass based
  • fast_forward00:07:47 - on the sun yeah they've got cells and eyes that are
  • fast_forward00:07:50 - sensitive to polarized light um so they have
  • fast_forward00:07:53 - and they need both steps and polarized light to do path integration and path
  • fast_forward00:08:01 - integration in some senses should be enough to go back to the next it should
  • fast_forward00:08:05 - be um but the problem with path integration so in path integration effectively you.
  • fast_forward00:08:11 - If you imagine dividing up your whole path into a series of small vectors,
  • fast_forward00:08:16 - then you simply sum those vectors and minus the sum vector points you back home.
  • fast_forward00:08:23 - Unfortunately, every step you take is going to have some error associated with
  • fast_forward00:08:30 - it, particularly if you're being blown by the wind, which the ants often are.
  • fast_forward00:08:34 - And so the errors accumulate and so
  • fast_forward00:08:37 - when you return home the longer the
  • fast_forward00:08:41 - path you've gone the more you're likely to miss the nest by so
  • fast_forward00:08:44 - ants actually do very similar things
  • fast_forward00:08:47 - to sailors used to do in
  • fast_forward00:08:50 - which if a sailor was trying to head back to their port by dead reckoning they
  • fast_forward00:08:54 - would always aim to one side of the port so that they knew that when they hit
  • fast_forward00:08:59 - the coast they knew which way they had to turn to find it and ants do seem to
  • fast_forward00:09:03 - do a similar thing in that they will go one side of the nest so that they know
  • fast_forward00:09:07 - which side to bias their search by.
  • fast_forward00:09:10 - So path integration is great and is essential for them to find their way home
  • fast_forward00:09:16 - the first time, but from the very first trip back, they will learn visual cues,
  • fast_forward00:09:21 - because envision the stability is in the world.
  • fast_forward00:09:24 - The ants that we study in Australia, Molophorus, Bogotty,
  • fast_forward00:09:28 - they don't use pheromones for navigation irrigation because they they burn off
  • fast_forward00:09:35 - in the uh in the heat and um the cataglyphus would be very much the same um
  • fast_forward00:09:42 - the wood ants we study in sussex in the lab um,
  • fast_forward00:09:49 - pheromones the ground is quite unstable because there's quite a lot
  • fast_forward00:09:51 - of rain um and so pheromones aren't
  • fast_forward00:09:55 - much good um and so again they're primarily
  • fast_forward00:09:58 - visual and i and all ants that can use that have vision do use vision and do
  • fast_forward00:10:07 - seem to prioritize it over the other elements they might use to get home so
  • fast_forward00:10:12 - um ants will have multiple strategies and as you're saying different species,
  • fast_forward00:10:17 - some species may use chemical trials but what's quite common in ants is to use
  • fast_forward00:10:22 - on top of path integration and compass sense some visual memory of the world
  • fast_forward00:10:29 - and that can compensate for errors you might make with your path integration
  • fast_forward00:10:34 - so you'd have two mechanisms that are complementary.
  • fast_forward00:10:39 - Yes well a complementary and there
  • fast_forward00:10:43 - is some evidence that they
  • fast_forward00:10:46 - are used they're used together but it looks more like that they they just come
  • fast_forward00:10:56 - to rely on one they will use the signals from one and then maybe use the other
  • fast_forward00:11:01 - if if the world starts to look different for instance so i'll
  • fast_forward00:11:03 - start using vision once they've learned the path they'll use vision and then
  • fast_forward00:11:07 - if things start to look wrong maybe they'll turn back to path integration or
  • fast_forward00:11:11 - go into a search behavior um,
  • fast_forward00:11:15 - it's very difficult to tease those things apart in experiments in the natural world because um,
  • fast_forward00:11:25 - you can't really put those two things in direct opposition very easily so the
  • fast_forward00:11:32 - ants uh As you said, the ANTAD has a somewhat crude visual system. Yes.
  • fast_forward00:11:38 - What is it that we think that they're attending to when they're using visual cues for navigation?
  • fast_forward00:11:43 - Okay, so some of our recent experiments say that the skyline,
  • fast_forward00:11:50 - Paul Graham and Ken Cheng's work have shown that the skyline,
  • fast_forward00:11:54 - which is the shape of trees against the sky,
  • fast_forward00:11:59 - is sufficient for them to be able to navigate, gate to be able to recover a direction home.
  • fast_forward00:12:06 - And their work also showed that they didn't just use certain key prominent objects.
  • fast_forward00:12:11 - So we think that it's likely to be some version of the,
  • fast_forward00:12:20 - the shape that things make against the sky, which is very easy for them to pick
  • fast_forward00:12:24 - out because they've got UV sensors.
  • fast_forward00:12:29 - So they would, in some sense, encode the shape of the horizon as they were leaving
  • fast_forward00:12:34 - the nest in order to be able to recognise when they're back there?
  • fast_forward00:12:39 - Yeah, that would be the classic snapshot view that you would remember what the
  • fast_forward00:12:43 - world looked like from the nest and a very low-dimensional parametrized version
  • fast_forward00:12:48 - of the world, whether you remember it as an image, whether you remember it as a height map,
  • fast_forward00:12:52 - whether you remember it as in other models as retinotopic positions of significant
  • fast_forward00:13:02 - gaps in the world, those things are very difficult to say.
  • fast_forward00:13:06 - But you remember that and then when you want to return to your nest,
  • fast_forward00:13:09 - you just move to make the world look more like your memory.
  • fast_forward00:13:14 - So that's the kind of hill climbing thing that if I move left and it looks a
  • fast_forward00:13:19 - bit more like my memory, I can keep moving left or something.
  • fast_forward00:13:23 - You just carry on. Yeah, you carry on in the direction you're going while it's getting more similar.
  • fast_forward00:13:27 - But I mean, hill climbing strategies are known to get stuck. They do.
  • fast_forward00:13:33 - So why doesn't the ant get stuck in local minima for its strategy?
  • fast_forward00:13:38 - Primarily because if you're in an area and there aren't any obstacles in the
  • fast_forward00:13:44 - way, if the region that you're in doesn't have any obstacles,
  • fast_forward00:13:50 - then there won't be any local minima.
  • fast_forward00:13:54 - And you can say that from what, from experimental findings? Yes,
  • fast_forward00:13:57 - from some work that Jochen Zyl did originally, where he took panoramic images
  • fast_forward00:14:04 - is from a series of natural environments.
  • fast_forward00:14:07 - And over a cubic meter, there was a clear gradient in this kind of image space.
  • fast_forward00:14:14 - And you can kind of do it mathematically as well. In a sense,
  • fast_forward00:14:19 - it's kind of the inverse of optic flow.
  • fast_forward00:14:22 - That if you haven't got occlusion, then things move in a very predictable way.
  • fast_forward00:14:27 - Now the caveat with all that is that for this to work, you have to be lined
  • fast_forward00:14:31 - up in the same, you have to be yeah you have to be in the same orientation you
  • fast_forward00:14:37 - were for all these images now.
  • fast_forward00:14:42 - Most of the visual homing algorithms this is true and this comes from the original
  • fast_forward00:14:48 - models were based on models of bees and wasps who do line up very precisely
  • fast_forward00:14:53 - in a certain orientation before entering in the nest.
  • fast_forward00:14:59 - And for ants, this is somewhat more difficult. So that is a challenge for the models.
  • fast_forward00:15:05 - And so we've proposed that visual memories might be used in a slightly different
  • fast_forward00:15:09 - way, whereby instead of trying to,
  • fast_forward00:15:14 - you try and recover the direction to your.
  • fast_forward00:15:19 - Yeah you try and recover the direction try and use the images as a visual compass,
  • fast_forward00:15:26 - so with a visual compass what you do is you remember the view from your goal
  • fast_forward00:15:32 - when you're in a certain orientation or more properly you remember the view
  • fast_forward00:15:36 - when you're pointing at your goal and from nearby positions if you rotate on
  • fast_forward00:15:42 - the spot when you're in that same orientation orientation,
  • fast_forward00:15:46 - then you will again find a sort of minimum in this image difference landscape space.
  • fast_forward00:15:52 - And so you'll be able to recover the orientation of those images.
  • fast_forward00:15:57 - Now, that doesn't seem very good for homing. But if you then remember a series
  • fast_forward00:16:02 - of images, when you're pointed at your goal from points surrounding the goal,
  • fast_forward00:16:08 - then you should be able to get back.
  • fast_forward00:16:11 - And we've modelled this and you can
  • fast_forward00:16:12 - then get back from anywhere within a reasonable range around that goal.
  • fast_forward00:16:20 - So how would you know that the ant was using that strategy? Is there some behavioural
  • fast_forward00:16:24 - marker that it's encoded a snapshot?
  • fast_forward00:16:28 - Well, we have observed ants visually scanning the world, or they appear to visually scan the world.
  • fast_forward00:16:37 - So we've got some high-speed recordings, and this is work in preparation of
  • fast_forward00:16:42 - ants, of Malophorus, the Australian desert ant.
  • fast_forward00:16:47 - When it's challenged with a new environment, sometimes just sort of naturally,
  • fast_forward00:16:54 - it has this kind of saccadic motion where it will run along for a bit,
  • fast_forward00:16:57 - then it will stop and it will turn on the spot, seem to pick a direction and
  • fast_forward00:17:02 - then head off again in a straight line and then again turn the spot.
  • fast_forward00:17:05 - And it does more of these scans when it's
  • fast_forward00:17:09 - in an unfamiliar environment and it also seems these scans are directed towards
  • fast_forward00:17:14 - the more familiar part of the environment and some really elegant work by Paul
  • fast_forward00:17:20 - Graham and Ken Chang and Antoine Wistrach so.
  • fast_forward00:17:29 - We at least know that the behaviour is there that would serve to facilitate our model.
  • fast_forward00:17:38 - The other thing we also do see in wood ants is they tend to walk in a sinusoidal
  • fast_forward00:17:44 - path, which again would enable you to behaviourally scan the world as you went along.
  • fast_forward00:17:53 - So as well as having behavioural evidence
  • fast_forward00:17:56 - that points towards this possibility of
  • fast_forward00:17:59 - them encoding snapshots um and this
  • fast_forward00:18:03 - is behavioral evidence from experiments in
  • fast_forward00:18:06 - in in natural environments yeah um you
  • fast_forward00:18:10 - also are doing computational modeling work to actually demonstrate that those
  • fast_forward00:18:14 - hypotheses about the mechanisms could operate to control say a simple robot
  • fast_forward00:18:20 - yep and so um what similarities you think there need to be between the robot
  • fast_forward00:18:26 - on the ant in order to test this model?
  • fast_forward00:18:28 - To test it so that it's a good model of an ant. Yeah, I mean,
  • fast_forward00:18:30 - can it be a very crude robot model?
  • fast_forward00:18:32 - I think that the real difficulty, and this is going to be the sticking block,
  • fast_forward00:18:39 - is to get the eye down to ant level.
  • fast_forward00:18:44 - And that's the real problem. I think we could do, I think we have done some tests on indoor robots.
  • fast_forward00:18:53 - We've got a gantry robot, but that's quite precise.
  • fast_forward00:18:56 - And we've kind of cluttered up the environment the world enough um
  • fast_forward00:19:00 - but we don't need we can do it with any sort
  • fast_forward00:19:03 - of robot platform and we should be doing this in the next year
  • fast_forward00:19:07 - you mean the strong test of actually putting a robot in the australian desert
  • fast_forward00:19:11 - where the ants are and showing that it can yeah i think the first test would
  • fast_forward00:19:15 - be we would do it in natural environments around sussex i think there's no point
  • fast_forward00:19:20 - doing it in the australian desert where it's a bit hot um until we've got
  • fast_forward00:19:25 - it working until we can test whether it's working in, in Sussex when it's a
  • fast_forward00:19:30 - foot above the ground, say. Um, and.
  • fast_forward00:19:35 - Yeah, I think that's our next step. So I think that could be quite a crude robot.
  • fast_forward00:19:41 - We don't really care about speed.
  • fast_forward00:19:43 - The ant has to deal with an uneven ground, which, because it's small, is very uneven for it.
  • fast_forward00:19:52 - And we're not going to be able to recreate that, I think, with any robot we have.
  • fast_forward00:19:56 - So there are certain compromises that come into...
  • fast_forward00:20:00 - We have to take into consideration in building a robot model of the system.
  • fast_forward00:20:03 - But are there, even though it's a crude model, are there already some insights
  • fast_forward00:20:07 - that you've got through this modeling approach that you think you may not have
  • fast_forward00:20:10 - had from a purely experimental one?
  • fast_forward00:20:12 - Um definitely i think the the the thing that we have done i think we've has
  • fast_forward00:20:18 - given us a lot of insight is is taking panoramic images from the positions that ants were navigating in,
  • fast_forward00:20:27 - and being able to see what their visual input.
  • fast_forward00:20:32 - Would have been what the what the input to their eyes
  • fast_forward00:20:35 - would have been it doesn't tell us what they're seeing doesn't really
  • fast_forward00:20:38 - tell us how the image is then being processed but we
  • fast_forward00:20:40 - can tell what the raw visual input is um and um
  • fast_forward00:20:44 - this is following again following on from work of yakinzade sort of
  • fast_forward00:20:47 - pioneered this work um and it is
  • fast_forward00:20:50 - staggering to see um how
  • fast_forward00:20:55 - different the world looks most of the world is sky and
  • fast_forward00:20:59 - a lot of the world is the ground and there's not much of the world carries any
  • fast_forward00:21:03 - kind of a signal for navigation um and so one of the slides we we like to show
  • fast_forward00:21:10 - or one of the images we like to show is an image of the environment where as
  • fast_forward00:21:14 - a human you instantly pick out the house.
  • fast_forward00:21:16 - And when you show it as a panoramic image, the house just kind of disappears into the background.
  • fast_forward00:21:21 - And when you blur it down to insect resolution, you really don't see anything.
  • fast_forward00:21:28 - So that really has given us a lot of insight and
  • fast_forward00:21:31 - has enabled us to do some interesting modelling
  • fast_forward00:21:35 - about the scale over which one single visual
  • fast_forward00:21:38 - memory would serve to allow you to
  • fast_forward00:21:40 - navigate and the insect
  • fast_forward00:21:45 - navigation the is a very collaborative area and lots of other people are doing
  • fast_forward00:21:52 - this sort of methodology and I think it's really helping the field to see you
  • fast_forward00:21:58 - know how different does the world look from this position do I need to I,
  • fast_forward00:22:05 - Would the ant need to have a cognitive map to get back from A to B,
  • fast_forward00:22:09 - or can it simply navigate with a simple strategy?
  • fast_forward00:22:14 - And just getting an idea of what the insect is seeing has been invaluable, I think.
  • fast_forward00:22:24 - Do you think that there are useful ideas people can take from this for designing
  • fast_forward00:22:28 - artefacts, say micro-robots?
  • fast_forward00:22:31 - Well, I think that most people tell me that the SLAM problem is solved and that…
  • fast_forward00:22:40 - That's the mapping problem.
  • fast_forward00:22:42 - The mapping problem. So the autonomous navigation can be solved by probabilistic
  • fast_forward00:22:48 - techniques where essentially you track a large number of features of the world
  • fast_forward00:22:52 - and you integrate that with information about your position and your likely movement.
  • fast_forward00:22:59 - You integrate probabilistically and you localize
  • fast_forward00:23:02 - both yourself and the features of the world in a map and they have um in the
  • fast_forward00:23:10 - DARPA grand challenges they navigate very long distances using these methods
  • fast_forward00:23:15 - however they take a lot of computation awful lot of computation and so they're quite
  • fast_forward00:23:21 - heavy are both with batteries and with power so i think the applications would be for.
  • fast_forward00:23:30 - Uavs unmanned air vehicles anywhere where the sort of power and weight considerations
  • fast_forward00:23:36 - are important so maybe also space exploration um and also the other thing is other places where
  • fast_forward00:23:45 - there is no GPS, a GPS denied environment.
  • fast_forward00:23:48 - If you have a GPS, you probably want to use it. But I think in environments
  • fast_forward00:23:52 - where you want a low cost solution for whatever reason,
  • fast_forward00:23:57 - and you haven't got a global GPS signal, then I think these methods,
  • fast_forward00:24:03 - they're surprisingly robust.
  • fast_forward00:24:08 - And I think is that almost a lesson from studying insect nervous systems for
  • fast_forward00:24:14 - designing technology more generally,
  • fast_forward00:24:16 - that what looks like, what is very robust behavior can be generated using a
  • fast_forward00:24:24 - relatively simple algorithm.
  • fast_forward00:24:27 - It doesn't have to be as sophisticated as what you might imagine at first.
  • fast_forward00:24:31 - No, definitely. Definitely. I think that is one of the great lessons.
  • fast_forward00:24:36 - With simple eyes, very poor resolution eyes, and with a brain that can't do
  • fast_forward00:24:44 - a lot of computation, or can't do very heavy computations, and probably hasn't
  • fast_forward00:24:50 - got a very big memory load.
  • fast_forward00:24:54 - Answer fantastic navigators. I mean, bees go miles with not much bigger brains.
  • fast_forward00:25:03 - And with similar mechanisms or does it get more complex?
  • fast_forward00:25:07 - Oh, I would like to, if I was to speculate I'd say I don't see why not I don't
  • fast_forward00:25:14 - see why not There is a controversy within the field as to whether bees might use cognitive maps,
  • fast_forward00:25:20 - I think the consensus would be that they don't I think they would use similar
  • fast_forward00:25:25 - mechanisms I know several people seem to think that a lot of the long distance homing would be,
  • fast_forward00:25:32 - based on the dormitory, um, I don't see why they wouldn't use visual information.
  • fast_forward00:25:39 - Um, that the horizon, what we've shown when we've shown our algorithms can easily
  • fast_forward00:25:45 - work over a hundred meters in an open environment, um, and you can kind of make
  • fast_forward00:25:50 - the world open by flying up.
  • fast_forward00:25:52 - And the shape of the horizon is a really, really strong queue and a robust one,
  • fast_forward00:25:56 - presumably, but it's not going to move very much.
  • fast_forward00:25:58 - I don't think, um, particularly over their lifetime. time so I don't see why
  • fast_forward00:26:01 - they wouldn't and you kind of only need to match.
  • fast_forward00:26:07 - Gross shape and you can get roughly the way back um i think insects have,
  • fast_forward00:26:14 - always tend to have multiple strategies which makes
  • fast_forward00:26:17 - it to do any one task which makes it complicated
  • fast_forward00:26:20 - when you're trying to study them and which is why they're interesting i think
  • fast_forward00:26:23 - um but they need to because they need to be robust and they need to get home
  • fast_forward00:26:28 - um and so i think that there will always be a combination of strategies so yeah
  • fast_forward00:26:34 - i would like to think that bees and wasps.
  • fast_forward00:26:37 - We know they use vision certainly for the last part of their to find their nest.
  • fast_forward00:26:43 - So I don't see why they wouldn't over a longer distance.
  • fast_forward00:26:46 - Thanks very much for talking to us, Andy. Thank you.
  • fast_forward00:26:48 - Music.
  • fast_forward00:26:53 - The CSN podcast was produced by the Convergent Science Network of Biometrics
  • fast_forward00:26:59 - and Biohybrid Systems, a project funded by the European PN7's Research Framework Programme.
  • fast_forward00:27:07 - For more interviews, recorded lectures or upcoming conferences in the field
  • fast_forward00:27:12 - of biometrics and biohybrid systems, go to csnnetwork.eu.
  • fast_forward00:27:19 - Music.

Be the first to leave a comment

Leave a comment

Your email address will not be published. Required fields are marked *

Convergent PozitivLinia Logo

Exploring the convergence of neuroscience, robotics, and AI through conversations with leading researchers since 2010.

A project of the Convergent Science Network Foundation.

© CSN Podcasts. Developed by IMCreativeWEBC

0%

Login to enjoy full advantages

Please login or subscribe to continue.

Go Premium!

Enjoy the full advantage of the premium access.

Stop following

Unfollow Cancel

Cancel subscription

Are you sure you want to cancel your subscription? You will lose your Premium access and stored playlists.

Go back Confirm cancellation