In a laboratory in a science park just north of Oxford, biomechanics professor Dr Adrian Thomas is channeling the spirit of da Vinci to create flight inspired by nature.
Leonardo's flying machine never left the pages of his sketchbooks, but Dr Thomas' fluttering drone is almost ready for take off.
"We're right on the edge of cracking it," he says. "We've got the performance that we need in terms of brute force generation out of the wings, and it's all now about refining it to get the endurance longer and the top speed higher and the maximum acceleration higher."
The designs developed by his company Animal Dynamics harness the innate abilities of the natural world to improve motion in the mechanical one. They have already attracted the attention of an eye-catching British client, the Ministry of Defence.
Flightpath of Animal Dynamics
The startup was spun out from Oxford University in 2015 by Dr Thomas and co-founder Alex Caccia, a serial entrepreneur who shared his interest in how the biomechanics of fish enable them to swim at rapid speeds. They developed this curiosity into a machine powered by the same type of flapping propulsion found under the sea.
Their experiments expanded to encompass other aspects of Dr Thomas' decade of studies in the Department of Zoology at Oxford University, researching animal movement on land, sea, and air.
"What's interesting about the way the animal kingdom tackles these problems is that efficiency and performance are completely at the heart of the solutions that the natural world uses," says Caccia.
"There's some fantastic opportunities to learn from that, in particular some very subtle mechanics and fluid dynamics which have not been fully exploited, which can then be used to build machines, robots, and propulsion systems that are highly efficient."
The vehicle closest to bringing their bio-inspired machines to life is a small beast with flapping wings by the name of Skeeter.
Skeeter has been reared in partnership with the Defence Science and Technology Laboratory (Dstl), the research arm of the Ministry of Defence.
The military applications for drones have devastating potential, whether in the form of hi-tech autonomous stealth bombers striking targets independently or as consumer UAVs used by ISIS to maim its enemies from the sky.
Skeeter has been designed to serve a more peaceful purpose: providing portable situational awareness to troops on the ground.
"It's small enough to be carried with the kit that the soldiers carry," says Caccia. "The issue is that as unmanned vehicles get smaller, all the aerodynamic numbers change and they become very unstable, which means that small quadcopters or helicopters can't really operate in windy conditions, which limits their use.
"If there's one out on a mission and they start off in still air, and suddenly a gust comes along and blows it into a bush, you've not only lost your asset but you've lost something that could provide valuable information to the enemy, or you have to send someone to go and fetch it, which is a pretty hairy thing to do.
"There's a considerable interest in having something which is small and discreet but can operate in high winds. Small birds can do it, insects can do it, and there are very good reasons why flapping wing flight, as a fundamental design, is much better at handling turbulent wind conditions than a propeller.
"The challenge that we've been set is to try and design and make something, which has the same sort of gust tolerance and discreetness as a small insect."
The Animal Dynamics team of 12 use highly sensitive measuring equipment to assess the impact of forces on their small and light designs, and have have been working with the Swiss watch industry to develop scaled-down components.
The possibilities have been opened by the development of tiny, affordable, high-quality MEMS microelectromechanical systems (MEMS) by the mobile phone that can integrate a number of different functionalities into one small device, and the increasing miniaturisation driven by the wearables markets.
"This is one of the tectonic shifts in the electronics sphere, which I think is having a huge impact on robotics," says Caccia. "We're able to get hold of very sophisticated sensors and high-performance processing and join them all together to make something that can provide you with the fundamentals of the control system, which can then be used to do things such as what we're doing."
The Animal Dynamics kingdom
They've also been exploring propulsion systems in water based on the fluid dynamics of dolphins and whales powered by flapping fins.
"From a theoretical point of view it's somewhere between 20-30 percent more efficient than a propeller," says Caccia.
"It's got much higher thrust coefficients, which means that it can deliver thrust over a much wider range of speed. If you see a great white shark sprinting to do a breach from 20 metres below, they don't produce a single bubble and yet they breach at 35 kilometres an hour. That should tell you something because bubbles are natures tell-tale sign that you're wasting energy."
Another idea on the whiteboard is a potential land vehicle, but they won't be using a cheetah as their inspiration from nature.
There have been numerous attempts at using "biomimetics" in design, a technique that imitation of nature in human-made systems. It's a term Caccia shuns when describing his company’s work.
"That's not really what we're interested in," he says. "The biomimetic approach would be to make a cheetah because that's the quadruped that runs fastest. I don't think that's the right approach.
"There are also very small creatures that run very fast, and very fast compared to their mass and size. You may well do better by studying them and learning from that and scaling it up. The cheetah exists because of where it is in the food chain, but it's not necessarily the right approach to make something that's fast and efficient."
"What we do is we study what the animal is achieving with the complicated system it's got, and use that as inspiration for an engineering design that does the same thing," adds Dr Thomas.
"The trap that biomimicry has fallen into is situations where people try and copy the details of the animals rather than identifying the underlying principles and then applying those.
"In general, mimicking their movements is extremely difficult. It's mainly difficult because muscles remain a very effective device if you're looking for small, really tiny actuators that can operate with high power to weight ratio and good efficiency. We actually don't have anything to replace the really tiny muscles."
His inspiration for Skeeter has existed for hundreds of millions of years, is just a few inches long, and weighs a fraction of a gram.
"The dragonfly flight motor has, depending on the species, something like 64 independent muscles in there to drive the wings," says Dr Thomas. "That's in a volume of half a centimetre by half a centimetre.
"It's a consequence of the fact that that flight motor evolved from a system that was designed originally for squirming a worm-like organism through the mud. Then it's been co-opted through evolution into becoming a flight motor. You don't need all those muscles to drive the number of degrees of freedom the insect has."
“Animals end up the way they are. Their morphology is often constrained by the evolution history. All the dragonflies, they all spend the first two years of their life underwater being voracious predators in ponds. Then they go and fly and they have to be able to cope with both those sorts of bits of their biology. What we're building doesn't have to do that.
"It can be just focused entirely on the flight side of things. We've got that advantage that we don't have to deal with messy stuff like catching prey and eating them and breeding and defending territories and things. We can just focus on the flight side."
Read next: 22 companies using drones right now
In addition to the dragonfly's biomechanics of flight, it also has outstanding skills as a predator.
“When they take off from a perch, in excess of 90 percent of the time they catch the prey item they've taken off to catch. If you look at a Patriot missile try to shoot down a Scud missile, 90 percent of the time it misses. The dragonflies are hugely over-engineered. That prey capture rate is terrific compared to most other predators, they're fantastically over engineered for it.
“The other big advantage with the dragonflies is they can turn all the motors off and glide, and they're stable in gliding. When we do the control side of things, our control system is much more straightforward, more like a fixed wing aircraft sort of control system.”
The performance of Skeeter is impressive, but there's still work to be done to refine its optimum combination of endurance, speed and acceleration, and the brute force required to increase the efficiency of the drive system. Dr Thomas has just broken the poor drone's wing.
By October they intend to hit an outdoor flight performance target and by the end of next year, they hope to have a commercial product that can be sold to the military. They've also been investigating applications in search and rescue, another sector with a need for equipment that can operate in high winds.
The slow evolution of nature helps robotic motion develop rapidly. After hundreds of millions of years of the dragonfly's existence, Animal Kingdom can now further advance its metamorphosis.
"Dragonflies started to dominate the aerial environment about 300 million years ago, maybe 200," says Dr Thomas. "300 million years of evolution for flight performance has got to give me some advantage."