It’s a well-established and indeniable scientific incontrovertible fact that squirrels are probably the most attention-grabbing animal within the recognized universe, and anybody that disagrees is fallacious. I imply, have you ever ever seen them chasing one another up bushes and zooming between branches quicker than an IndyCar driver staring down the checkered flag? These high-wire antics could seem reckless, however squirrels are filled with surprises. They will (in concept, no less than) safely fall from any top by stretching out and giving themselves a really giant size-to-mass ratio, which vastly reduces their terminal velocity.
Suffice it to say, when any squirrel-related tech information arises we clamor to cowl it right here at Hackster Information. The newest instance involves us from the Pohang College of Science and Know-how, the place a staff of researchers is constructing a highly-maneuverable drone modeled after the habits of the flying squirrel. It incorporates foldable wings into its design that permit it to carry out tighter, extra responsive maneuvers than standard drones. Contemplating how agile the furry number of flying squirrel is within the air, that appears like a recipe for superb.
A better take a look at the {hardware} design (📷: D. Lee et al.)
Your complete system weighs simply 548 grams, with the foldable, silicone wings themselves including solely 24 grams. It’s powered by 4 high-speed motors, two servos for wing deployment, and is guided by an Arduino Portenta H7 with a dual-core STM32 microcontroller. Sensors embrace twin IMUs, a barometer, and GNSS — all of which assist the drone keep oriented and on-course mid-flight.
What makes this drone so particular is not only the addition of wings, however how these wings are used. The staff designed a Thrust-Wing Coordination Management (TWCC) framework to intelligently handle when and how one can deploy the foldable wings. You possibly can consider it because the drone’s built-in instinctual reflexes, like how an actual squirrel would instinctively alter its glide to keep away from an owl or make a daring leap. The TWCC algorithm coordinates the wing deployment with propeller thrust to develop the drone’s set of controllable actions, notably the pesky vertical ones which can be usually difficult for normal quadcopters.
An outline of the Thrust-Wing Coordination Management framework (📷: D. Lee et al.)
To precisely predict how air will behave across the wings in real-world flight, the staff developed a physics-assisted recurrent neural community. This machine studying mannequin was skilled on actual flight knowledge to dynamically alter the angle of assault — that’s, how the wing slices by way of the air — primarily based on precise aerodynamic habits. By doing this, the drone could make smarter in-flight choices and higher anticipate how folding or unfolding its wings will have an effect on efficiency.
In outside checks involving dynamic trajectories and digital obstacles, the flying squirrel drone demonstrated a 13.1% enchancment in monitoring accuracy over standard drones.
By mixing a bio-inspired design with a cutting-edge management system and machine studying, the researchers have created a drone that does extra than simply fly — it adapts, reacts, and performs with a stage of agility that brings it one step nearer to what’s seen within the pure world. If the way forward for drones seems to be something just like the flying squirrel, then it might appear that we’re on the fitting path.
