A pigeon-inspired flying robot is solving avian mysteries that may help create more stable aircrafts.
While birds seem to seamlessly maintain stability during turbulence, airplanes need rudders and vertical tails to avoid being rocked from side to side. Scientists have suspected that birds maintain balance by reflexively adjusting their wings and tails but this hypothesis has been hard to prove with real birds in field studies.
To overcome this challenge, scientists at the American Association for the Advancement of Science (AAAS), developed PigeonBot II, a robot equipped with morphing wings, pigeon-like wingtips, and even 52 real pigeon feathers. Just like a real bird, it can elevate, spread its wings, and control its tail.
The study, led by Eric Chang, tested PigeonBot II in both indoor and outdoor conditions, and, according to the researchers, confirms the theory and could help inspire more efficient, rudderless aircrafts.
PigeonBot II morphs its wing and tail reflexively to fly rudderless autonomously.
(A) Biomimetic skeleton and connective elastic ligaments underactuate the wing’s 40 remiges with four servomotors (two wrists and two second digits) and the tail’s 12 feathers with five servomotors (movie S3). Scale bars, 100 mm.
(B) Wing and tail can be arbitrarily morphed; shown here are three typical pigeon postures during gliding (green) with the wing and tail approximately proportionally morphed from spread to middle to tucked (25). Scale bar, 100 mm for PigeonBot II photos.
(C) Avatars illustrate the six controlled tail and wing morphing DOFs used in this study.
(D) Reflexive control loop for autonomous controlled flight in the atmosphere based on sensor fusion–based body angle and angular rate feedback: Pitch is controlled with tail elevation; roll is controlled with wing asymmetry, tail tilt, and lateral deviation; and the yaw rate is damped with tail tilt (movie S4).
To navigate the robot, the teleoperator inputs an airspeed setpoint and either high-level roll and pitch setpoints via radio or switches to autonomous circling setpoints.
The autopilot then computes conventional roll, pitch, and yaw commands with PID control. Next, a microcontroller adapts the output with fitted gains tuned for all morphing wing and tail permutations (spread, middle, and tucked). Last, the roll command is mixed [Kasymm, Kdev, Ktilt] = [1, 2, ¼] and mapped to the servomotors to reflexively morph the wing and tail.
n this article, we developed a biomimetic morphing wing and tail robot model (PigeonBot II) to test the biological hypothesis that birds can automatically control and stabilize their flight with morphing wing and tail reflexes and to create a flight-demonstrated engineering model that shows how robots can accomplish rudderless flight by harnessing autonomous reflexive morphing. We first established that pigeon-like rudderless planforms are Dutch roll unstable (Fig. 2 and movie S2).
By tuning our bird-inspired reflexive controller (Fig. 1D) in a virtual flight-testing setup in a wind tunnel, we found that appropriate mixing of wing asymmetry, tail tilt, and tail lateral deviation in response to roll perturbations maximizes roll tracking (Figs. 2 and 3) and stability in turbulence (Fig. 4) across wing and tail spread permutations. Using the reflexive controller tuned in the wind tunnel, we tested PigeonBot II during autonomous atmospheric flights to demonstrate robust flight stabilization and navigation at near-proportional wing and tail spread morphing combinations (Fig. 5 and movie S7).
This work confirms how birds can accomplish rudderless flight via reflex functions, and it can inspire rudderless aircraft with reduced radar signature and increased efficacy.
