Optic-flow: from insects to robots

From many behavioural and neurophysiological experiments, we know lots about the use of optic flow by insects for the control on flight and the avoidance of obstacles. Here is a review explaining how the derived control principles can be used as bio-robotic inspiration.

Abstract: “Flying insects are able to fly smartly in an unpredictable environment. It has been found that flying insects have smart neurons inside their tiny brains that are sensitive to visual motion also called optic flow. Consequently, flying insects rely mainly on visual motion during their flight maneuvers such as: takeoff or landing, terrain following, tunnel crossing, lateral and frontal obstacle avoidance, and adjusting flight speed in a cluttered environment. Optic flow can be defined as the vector field of the apparent motion of objects, surfaces, and edges in a visual scene generated by the relative motion between an observer (an eye or a camera) and the scene. Translational optic flow is particularly interesting for short-range navigation because it depends on the ratio between (i) the relative linear speed of the visual scene with respect to the observer and (ii) the distance of the observer from obstacles in the surrounding environment without any direct measurement of either speed or distance. In flying insects, roll stabilization reflex and yaw saccades attenuate any rotation at the eye level in roll and yaw respectively (i.e. to cancel any rotational optic flow) in order to ensure pure translational optic flow between two successive saccades. Our survey focuses on feedback-loops which use the translational optic flow that insects employ for collision-free navigation. Optic flow is likely, over the next decade to be one of the most important visual cues that can explain flying insects’ behaviors for short-range navigation maneuvers in complex tunnels. Conversely, the biorobotic approach can therefore help to develop innovative flight control systems for flying robots with the aim of mimicking flying insects’ abilities and better understanding their flight.”

Serres, J. R., & Ruffier, F. (2017). Optic flow-based collision-free strategies: From insects to robots. Arthropod Structure & Development.

A robot named Cataglyphis

Here is a fun article detailing a robot which successfully completed a NASA challenge. Although not strictly a bio-inspired robot, the challenge was for GPS-denied robots in a retrieve and return task. So the name is very apt. You can find the details here.<onlinelibrary.wiley.com/doi/10.1002/rob.21737/full>

Gu, Y., Ohi, N., Lassak, K., Strader, J., Kogan, L., Hypes, A., … & Watson, R. Cataglyphis: An autonomous sample return rover. Journal of Field Robotics.

When do views beat vectors?

One of the most interesting questions that we can address with insect navigation research is cue conflict and how it plays out in natural real world behaviours. In the case of Path Integration and learnt visual cues, we know that in experienced well trained ants, those two modalities work simultaneously, but that vision “wins” in the case of a 180 degree conflict. Here, Freas and Cheng look at the temporal dynamics of this type of cue integration. They find that multiple visually guided routes within the last 24 hours are needed to give visual cues dominance over PI. This kind of temporal dynamics to the weighting of cues may be part of a suite of mechanisms that ensure robust cue weighting.

Freas, C. A., & Cheng, K. Learning and time‐dependent cue choice in the desert ant, Melophorus bagoti. Ethology.

Motor motifs matched to visual ecology

In insect navigation research we often think of Cataglyphis desert ants as a touchstone or model system that we can use as a reference when comparing other insects. However, there are many species within the Cataglyphis genus and a range of habitat types within which they navigate. The variation within desert ant navigation strategies might actually be quite high and here it is consider how variation might relate to ecology. This paper beautifully shows the value of high-resolution videography by neatly capturing variations in the learning walks of Cataglyphis species from cluttered or barren terrain. Ants from cluttered terrain show very precise movements which result in accurate fixations of the nest. Such fixations are absent in Cataglyphis fortis, which inhabits visually inconspicuous terrain.  This is very suggestive of an accurate visual learning that can drive visual homing when visual information is available. We know that C. fortis can use  visual information, but is seemingly less primed to do so for nest localisation.

P N. Fleischmann, R Grob, R Wehner, and W Rössler (2017) Species-specific differences in the fine structure of learning walk elements in Cataglyphis ants J Exp Biol 2017 220:2426-2435. doi:10.1242/jeb.158147

Steering a ring attractor

In recent years we have started to learn about the circuitry that underpins spatial behaviour in insects. One of the most important findings has been that of a ring attractor circuit in the ellipsoid body of the central complex of insects. This circuit can represent heading and is updated via visual or proprioceptive information. Now, the understanding of how this ring-attractor works is expanding to the associated structures. Green et al. have detailed how specific neurons, in a brain areas connected to the ellipsoid body, are involved in the driving of neural activity in the ring attractor. This opens up the potential for us to understand how sensory input contributes to the neural representation of spatial information.

Green, J., Adachi, A., Shah, K. K., Hirokawa, J. D., Magani, P. S., & Maimon, G. (2017). A neural circuit architecture for angular integration in Drosophila. Nature, 546(7656), 101-106.

It ain’t what you do its the way that you do it.

Nobody would argue with the statement that insects have rich and interesting behaviour repertoires that require (presumably) non-trivial neural circuitry. However, how insects fit into the field of comparative cognition, where we look at cognitive processes across animals as a function of their ecology and phylogeny, is unclear. One approach is to demonstrate the similarity of insect behaviours (that seemingly require cognition) with behaviours of vertebrates (that are assumed to be cognitive). This top-down approach can often lead to a game of semantics. The alternative is a bottom-up approach, focussing on the mechanisms needed for a particular paper.

In this review, Perry et al., take a series of examples and certainly demonstrate the interesting richness of insect behaviour, whilst showing an understanding of how this field can progress. For example, here is a quote from the Discussion section of their paper: “While it is tempting to explore ever more human-like types of cognitive operations in insects and other animals, the field of comparative cognition needs to move on to discover the neural underpinnings of cognition. The same cognitive capacity might be mediated by entirely different neural circuitries in different species, with a many-to-one mapping between behavioural routines, computations and their neural implementations. In fact, before we can understand a cognitive operation as a circuit function we should be wary of rating them as ‘higher’ or ‘lower’ forms of cognition.”

I personally feel like we are getting close to such an understanding of insect navigation. Especially, with interesting behavioural work on ants and new neurobiological findings from flies and other insects.

Perry, C. J., Barron, A. B., & Chittka, L. (2017). The frontiers of insect cognition. Current Opinion in Behavioral Sciences, 16, 111-118.

Real world problem solving

Here is a fun looking paper where ants have been set a series of physical problems (path choice, door pushing and barrier avoidance). Perhaps unsurprisingly, ants (as clever critters) as adept at ‘solving’ these real-world challenges.

The pdf is available here: <http://www.ccsenet.org/journal/index.php/ijb/article/view/67868/36990&gt;

Abstract: “Aiming to know the extent of the ants’ cognitive abilities, we set Myrmica ruginodis workers in four problematic situations. We discovered that these ants could walk round a barrier, by foraging and navigating as usual, using known visual cues. They could walk preferentially on smooth substrates instead of rough ones, but did not memorize their choice. This behavior may be due to the easier deposit of pheromones on a smooth substrate. The ants could establish a single way when having only two narrow paths for going in and out of their nest. This was the consequence of the ants’ traffic and of the distinct pheromonal deposits while going in and out of the nest. The oldest ants needing sugar water could push a door for getting such water. They did so by having the audacity to go on walking, whatever the presence of a door. Such a door is not a tool sensu stricto. Future studies will examine if ants can lean new techniques, can use tools and/or can learn using tools.”

Cammaerts, M. C. (2017). Ants’ Ability in Solving Simple Problems. International Journal of Biology, 9(3), 26.