Extracting Causal Relationships Between Wing Kinematics and Flight Maneuvers in Bats for Transforming Bioinspired Flapping Flight

Bats provide a model system for small aerial vehicles by virtue of their versatility. They have evolved a myriad of capabilities ranging from traveling large distances, carrying loads as high as their own body weight, to chasing insects in flight with extraordinary speed and agility. All this is made possible by membrane-covered wings manipulated by fingers with about 20 joints which gives bat wings the unique capability to change shapes in support of the desired flight objective at any given time. Current drone technology mostly relies on rotary wings which do not provide much maneuverability and hence cannot navigate cluttered environments. The goal of the proposed work is to learn the relationship between changes in wing shape and the flight trajectory. These goals will be accomplished by measuring the wing motion and trajectory of bats as they fly through obstacles in a tunnel. The wing motions will then be used to compute the aerodynamic flow generated by the wings. By relating wing motion to the aerodynamic forces felt by the wings and to the trajectory of the bat, it is possible to identify the consequential wing motions. The design principles learned from bat flight will benefit society by allowing access to natural and man-made cluttered environments for agricultural, environmental surveillance, and other emerging humanitarian uses. It is proposed to make advances in measurements and geometry reconstruction using high-throughput techniques for recording flight and deep learning methods for reconstructing the 3D space-time bat wing kinematics. The experiments are to be conducted on the island of Borneo which is home to about one hundred bat species that include some of the most maneuverable flyers. High-fidelity computational fluid dynamics (CFD) on central processing unit (CPUs) and graphics processing unit (GPUs) is to be used to calculate the time-dependent turbulent flow field generated by the measured wing kinematics. This is to be combined with inertial forces and moments generated by the wings to predict the six degrees-of-freedom translational and rotational dynamics of the bat to be validated with the measured trajectory. The fluid dynamic events that lead to force and moment asymmetries to effectuate a maneuver will be investigated by isolating dominant events through advanced data analytics. The final objective is to use physics-guided deep-learning techniques to establish causal relationships between dominant wing kinematic traits and changes in trajectory during a maneuver for designing transformational bio-inspired bat-like drones. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.