The Emergent Agility of Insect Flight

The study of how agile animal locomotion emerges from an integration of neural processing and biomechanics is not just facilitating the discovery of new physical and biological principles but also pushing the boundaries of engineering knowledge. Insects serve as ideal models for this exploration due to their relatively simple musculature and neural circuits. Their neurological and biomechanical features have co-evolved – driven by the critical role of flight performance in their survival. The dynamics of their flight place functional demands on the nervous system, which adeptly processes sensory feedback from various sources and orchestrates the coordination of numerous muscles to actuate the wings. Despite the inherent challenges posed by dynamics of flight and rapid wing flapping with limited degrees of freedom in wing movement, insects exhibit remarkably agile flight – outperforming their engineered counterparts. This research aims to study how the insect sensorimotor system and flight mechanics integrate to elicit emergent agility in flight. In Chapter 1, I introduce the major aspects and significance of the integration question I tackle in this thesis. In Chapters 2 and 3, I refine a quasi-steady aerodynamic model to discover that the evolution of divergent wing morphology and movement in sister clades hawkmoths and wild silkmoths have led to distinct flight strategies in terms of their kinematic, dynamical and energetic demands of emergent agile flight. In Chapter 4, I leverage the power of machine learning to create a visuomotor system model of the hawkmoth that encodes visual data to predict precisely-timed motor responses. This will help in understanding the emergence of agile flight once integrated with models of insect biomechanics in a bottom-up integration approach. The final chapter, Chapter 5 is dedicated to a top-down approach employing system identification experiments to examine the emergent linearity of flower-tracking behavior in hawkmoths. Analyses of frequency responses of sensorimotor control and flight mechanics show that the linearity emerges because the two subsystems operate linearly throughout the entire dynamic range of flower-tracking. Thus, in addition, top-down approaches can also be used to inform bottom-up approaches by specifying the operating ranges of inputs and outputs of component models during an agile locomotion behavior. Overall, exploring the integration principles across timescales and modeling approaches presented in this thesis can help design bioinspired robots with well-integrated components and thus capable of agile locomotion in complex and uncertain environments.

Date

2024-04-28

Resource Type

Text

Resource Subtype

Dissertation

Full item page