AI Unravels Insect Wing Mechanics
350 million years ago, winged insects, or pterygotes, evolved. Their flight mechanics, especially the role of sclerites within the wing hinge, remained a mystery. Michael Dickinson’s team at Caltech used high-speed cameras and neural networks to decode these mechanics, revealing how different muscles control wing motion. Their findings could inspire advanced insect-like flying robots.
About 350 million years ago, our planet witnessed the evolution of the first flying creatures. These early aviators are still around today, and some of them continue to annoy us with their persistent buzzing. While scientists have classified these creatures as pterygotes, the rest of the world simply calls them winged insects. Despite the ubiquity of these insects, many aspects of their biology, especially their flight, remain a mystery to scientists. One of the most perplexing questions is how they move their wings. The insect wing hinge is a specialised joint that connects an insect’s wings with its body. It is composed of five interconnected plate-like structures called sclerites. When these plates are shifted by the underlying muscles, the insect wings flap.
Until recently, it has been tricky for scientists to understand the biomechanics that govern the motion of the sclerites, even with advanced imaging technologies. Michael Dickinson, Zarem professor of biology and bioengineering at the California Institute of Technology (Caltech), pointed out that the sclerites within the wing hinge are so small and move so rapidly that their mechanical operation during flight has not been accurately captured, despite efforts using stroboscopic photography, high-speed videography, and X-ray tomography.
As a result, scientists have been unable to visualise exactly what’s happening at the micro-scale within the wing hinge during flight, preventing a detailed study of insect flight. However, a new study by Dickinson and his team has finally revealed the workings of sclerites and the insect wing hinge. They captured the wing motion of fruit flies (Drosophila melanogaster) and analysed 72,000 recorded wing beats using a neural network to decode the role individual sclerites played in shaping insect wing motion.
The biomechanics that govern insect flight are quite different from those of birds and bats because insect wings did not evolve from limbs. Birds, bats, and pterosaurs all fly with their forelimbs, essentially using their arms to fly. Insects, on the other hand, evolved from six-legged organisms and retained all six legs while adding flapping appendages to the dorsal side of their bodies. The origin of these wings remains a mystery. Some researchers suggest that insect wings evolved from gill-like appendages in ancient aquatic arthropods, while others argue that wings originated from “lobes,” special outgrowths found on the legs of ancient crustaceans, which were ancestors of insects. This debate continues, and the evolution of insect wings cannot fully explain how the hinge and the sclerites operate.
Understanding the hinge mechanics is crucial because this is what makes insects efficient flying creatures. The insect wing hinge enables them to fly at impressive speeds relative to their body sizes—some insects can fly at 33 mph—and to demonstrate great manoeuvrability and stability while in flight. The insect wing hinge is considered among the most sophisticated and evolutionarily important skeletal structures in the natural world.
Imaging the activity of four of the five sclerites that form the hinge has been impossible due to their size and the speeds at which they move. Dickinson and his team overcame this challenge by employing a multidisciplinary approach. They designed an apparatus equipped with three high-speed cameras that recorded the activity of tethered fruit flies at 15,000 frames per second using infrared light. They also used a calcium-sensitive protein to track changes in the activity of the steering muscles of the insects during flight, as calcium helps trigger muscle contractions.
The team recorded a total of 485 flight sequences from 82 flies. After excluding a subset of wingbeats from sequences when the fly either stopped flying or flew at an abnormally low wingbeat frequency, they obtained a final dataset of 72,219 wingbeats. They then trained a machine-learning-based convolutional neural network (CNN) using 85 percent of the dataset. This CNN model investigated the transformation between muscle activity and wing motion by performing virtual manipulations, executing experiments that would be difficult to perform on actual flies.
In addition to the CNN, they developed an encoder-decoder neural network (an architecture used in machine learning) and fed it data related to steering muscle activity. While the CNN model could predict wing motion, the encoder/decoder could predict the action of individual sclerite muscles during the movement of the wings. The accuracy of these predictions was then tested using a tiny winged robot.
The virtual experiments performed using the CNN revealed that the hinge worked in coordination with 12 steering muscles (arranged in four groups) to control wing motion. Dickinson highlighted that the biggest finding of their study was the ability to isolate what different steering muscles do for the flight control of an insect. There are four groups of muscles attaching to four sclerites, and each muscle group has different actions on the pattern of wing motion. It is by controlling these four muscle groups that an insect is able to control wing motion in a very subtle but precise way, allowing for the agile manoeuvres that insects are known for.
Using this information, the researchers constructed a tiny winged robot and used the predictions of the CNN model to test the aerodynamic force it generated. They ran simulations to further test whether the CNN-linked muscle activity to wing motion could generate free-flight manoeuvres like those performed by real flies. The results of these simulations resembled the known behaviour of freely flying flies. When the steering muscle activity was parsed into nodes representing the muscle groups inserted on each of the four wing sclerites, the model’s predictions were consistent with known features of hinge morphology and differences in the insertion patterns of control muscles.
These findings have significant implications. Dickinson and his team hope that their results will allow scientists to deepen their understanding of insect flight and contribute to the development of better and more efficient insect-inspired flying robots. They are currently studying mosquitos and plan to study other insects in the future to understand how the wing hinge evolved in lineages with distinct body plans.
Separately, the flight of mosquitoes has also drawn scientific interest. Within the insect world, mosquitoes have a distinctive flight, characterised by a short wing stroke and a very high frequency of wing beats. Researchers have figured out the physics behind mosquito flight, identifying two mechanisms for generating lift that had not previously been seen in any animal. Much of the aerodynamic force that supports a mosquito’s weight is generated in a manner unlike any previously described for a flying animal.
A small team of Japanese and UK researchers set up a series of eight high-speed cameras to capture every instant of a mosquito’s wing flap from multiple angles. This data allowed them to create a digital model of the wings as they went through a full stroke. They confirmed that a mosquito’s wings beat at a frequency of over 700 Hz, much faster than most other insects, which explains the mosquito’s distinctive whine. The sweep of the wing during the wing beat is less than 40 degrees, less than half the smallest amplitude yet measured for any hovering animal.
The models found that lift is generated by three distinct mechanisms. The first, a leading-edge vortex, is common in other insects. During the downward sweep of the wing, a vortex of air is generated in front of the leading edge of the wing and then loops back over it, creating an area of low pressure above the wing, providing some lift. However, because of the mosquito’s short wing stroke, the wing doesn’t beat downwards for long. Something else must be going on.
One part of that something else is a trailing edge vortex. Normally, this vortex forms at the back edge of the wing and then moves away from it, so it doesn’t generate lift. But the mosquito’s wing stroke is such that, as soon as it stops its downward stroke and reverses its wing upwards, the wing runs into the trailing edge vortex, which reattaches to the wing, providing some lift.
Mechanism three involves what happens when the wing reaches the top of its upwards stroke and starts to reverse course downward. This rotation creates an area of low pressure on top of the wing that adds some lift. The mosquito manages to stretch this out and get much more lift by slowly shifting the axis of wing rotation from the front of the wing to the rear and not rotating the whole wing at once. This exquisite wing control is required to get so much out of a short wing stroke, and no other animal appears to be doing any of this.
These insights into insect flight, from fruit flies to mosquitoes, not only unravel some of the mysteries of their remarkable flying abilities but also hold the potential to inspire advancements in robotics and aerodynamics. Understanding the complex mechanics of insect wings could lead to the development of new technologies that mimic these efficient and agile flight mechanisms.
for all my daily news and tips on AI, Emerging technologies at the intersection of humans, just sign up for my FREE newsletter at www.robotpigeon.be
