Biomimicry yields agile wing flapping drones with IoT potential

Until now, flying drones or robots have been inspired by fixed wing aircraft or helicopters rather than flying insects. But that could be about to change for smaller sizes at least, thanks to research taking advantage of AI-based technologies. While being inspired by insect counterparts, emerging drones with flapping wings also promise to help clear up remaining puzzles over the aerodynamics of insect flight, as well as opening new IoT applications in remote monitoring and surveillance. Their main advantage lies in smaller size since making working propeller-based drones on that scale would be prohibitively expensive per unit.

The new designs also score on agility, through making use of insect-like aerodynamic capabilities. At least two significant steps have been taken in flapping winged drones this year by separate groups either side of the Atlantic. One from the University of Washington, called RoboFly, was one of the first flapping winged robots untethered to any wires, incorporating a processing unit but not a battery. About the size and weight of a toothpick, this is powered by a laser beam converted into electrical energy by a miniature onboard circuit.

However, this would have limited utility and so a flapping robot developed by Delft University of Technology (TU Delft)’s Micro Air Vehicle Laboratory (MAVL) with Wageningen University & Research, both in the Netherlands, shows greater promise at present. This is modelled on the fruit fly, chosen because it is used widely for research and therefore readily available for study. But the robot has been scaled up in size to make it more practical for manufacture and more readily applicable to potential applications such as sniffing for gas leaks.

While a typical fruit fly flaps its wings 200 times a second, this robot’s wings beat over 10 times slower at 17 per second. With a wing span of 33cm and weight of 29 grams this is larger than any flying insect, and yet its creators believe it can yield further insights into insect flight. This is because it can still perform many of the feats of fruit flies, including angled turns to evade potential predators and also hovering, which is common to many other insects, as well as humming birds and a few small bat species.

A key point here is that maximum size of flying insects is constrained not so much by aerodynamics but scaling laws of biology. Wing surface area increases with the square of the animal’s length but body mass by the cube, so the amount of energy required to sustain flight, and the amount of power for lift off, scales disproportionally. This is why no large animal can ever fly. The amount of food required would require a stomach larger than the animal and wing span would also have to be out of proportion.

The rules of physics also apply to machines, but humans are free to select their means of propulsion as well as the materials for construction. Electric motors can provide greater power for a given size, with the challenge then being battery life. The MAVL machine can fly 1 Km at up to 25 Km/h, or can hover for up to five minutes which is almost as energy sapping.

The aerodynamics of hovering are now quite well understood, involving a combination of wing flapping frequency and orientation tuned to generate tiny vortices in the air that provide just the right amount of uplift to keep the insect or robot stationary. Evolution has equipped small insects such as fruit flies with the physiology and wing structure to achieve hovering as well as acrobatic flight, requiring little training to perfect.

Until recently the processes were hard to study and required some enlightened guesses to elucidate. But now robots can emulate these processes with the huge advantage that their operators can tweak parameters such as wing frequency and orientation, as well as structure during the design stage. As a result, there is now a platform for studying processes that are still not well understood at the level of micro vortices and small air movements, such as some of the aeronautics performed by insects either when under attack, or chasing prey in the air as some species can.

Meanwhile though these flapping robots inspired by insects can be developed for a variety of novel applications, such as more granular surveillance in agriculture. This is where the IoT comes in, for remote control, data gathering and recharging automatically in the field. These devices are likely to be deployed at scale and will require autonomous operation if they are to deliver cost effective benefits.