The dragonfly’s superpowers: from predictive visual guidance and wing aerodynamics sensing to bio-inspired robotics

<Abstract>

The dragonfly is a successful aerial predator and is thought to have been the first animal to take to the air some 250M years ago. Its impressive repertoire of flight behaviors includes hovering, gliding, rapid translation, backward flight, and tandem flight. Such performance owes itself to the dragonfly’s two highly developed compound eyes and four independently controlled wings. My lab investigates key mechanisms that enable the dragonfly’s excellent visual guidance and flight control through methods from biomechanics, neuroscience, animal behavior, and bio-inspired robotics. In this seminar, I will cover two major research streams in the lab and illustrate the dragonfly’s unique superpowers.

The first stream focuses on visual guidance strategies the dragonfly uses for intercepting targets and implementing flight control. The dragonfly’s prey capture has been reported to have the highest success rate in the animal kingdom (>90%). We quantitatively scrutinized the visual behavior leading to an interception flight and found that the dragonfly carefully selects catchable targets and choreographs its pursuit initiation to optimize the success. During flight, it predictively steers its visual system to simplify the visual guidance tasks. These analyses reveal a simple yet reliable target interception strategy which can inspire robust visual control for small autonomous mobile robots.

The second research stream aims to reveal how many hundreds of wing mechanosensors encode complex unsteady aerodynamics and aeroelastic wing deformation in-flight. We performed a comprehensive characterisation of the dragonfly wing sensory system including sensor classification, neuronal mapping, sensor distribution, structural analysis, and electrophysiological recordings. In this talk, I will review our current understanding of the dragonfly wing mechanosensory system and share some recent progress in decoding the wing sensory signals. I will discuss the different approaches for complex flow sensing and wing strain sensing occurring on the dragonfly wing. The integration of wing structure and sensory system is an efficient way to monitor the aerodynamic contribution from each wing in-flight. It offers a guideline for the design and control of modern morphing wing applications both in air and in water.