A morphology-centered view towards describing bats dynamically versatile wing conformations
Eric Sihite,Alireza Ramezani
DOI: https://doi.org/10.1177/02783649241272132
2024-09-25
The International Journal of Robotics Research
Abstract:The International Journal of Robotics Research, Ahead of Print. Bats exhibit a remarkable ability to create periodic air jets, manifested as pulsating wake structures downstream of their flight path, as they navigate through their fluidic environment. This distinctive skill, driven by their fine-grained and dynamic wing motions, is integral to bats' unparalleled agility and flight efficiency. We refer to flight led by these dynamic wing adjustments as "dynamic morphing wing flight." The robotic emulation of bats' adept manipulation of their fluidic surroundings offers fresh perspectives for aerial robotics. Nevertheless, the design of micro aerial vehicles (MAVs) following the bat's style presents significant challenges, including constraints related to payload, actuation, and computation. While contemporary MAVs primarily consider body structures as mere hosts for components, our work explores a morphology-centric MAV design that aims to incorporate closed-loop motion control within the structural design itself, termed "computational structures." To fulfill this objective, we have developed Aerobat, an MAV equipped with dynamic morphing wings, onboard electronics, including actuators, an inertial measurement unit, an onboard camera, and a powerful computer. In the design of Aerobat, we meticulously addressed two specific needs: (1) the generation of gaits using a limited number of high-power actuators, and (2) the regulation of gaits employing an arbitrary number of low-power actuators. Our contributions in this work encompass (1) the introduction of a morphology optimization-based design framework, (2) the demonstration of stable, untethered dynamic morphing flights, and (3) achieving an aerodynamic force enhancement mechanism through wing surface maximization and wing rise-up time minimization.
robotics