Neuromechanics of Soft-bodied Locomotion

Tufts University, Medford MA

Investigators

Abstract

Animals are remarkably good at moving in complex and changing environments, far exceeding the capability of the most advanced robots. This adaptability is possible because the nervous system and the body interact with the environment as a single "neuromechanical system". In animals, this process is dominated by soft tissues such as muscles and skin but it is largely unknown how the movements of soft materials are controlled. This project will answer this question by studying how neural, mechanical and sensing mechanisms contribute to adaptable locomotion. The experiments will be carried out on caterpillars because they provide a unique opportunity to use electrophysiology (recording the electrical activity of neurons and muscles), motion capture and mechanical measurements during free behavior. In addition to addressing issues that are largely missing from our understanding of terrestrial soft animal locomotion, these studies have wide applications in robotics including the design and control of climbing machines, building assistive devices and manipulators and for controlling better prosthetics. Understanding the mechanical and sensory features that are important for caterpillar locomotion could also impact the development of pesticide-free pest-control strategies (e.g., plant surface engineering), managing sensitive ecological niches and modeling insect/host plant interactions. Three areas of research will determine how caterpillars adjust their locomotion in different environments. The first uses motion capture to analyze the stepping patterns (gaits) on different substrates and changes in orientation. The mechanical properties of the substrate (including stiffness, density, curvature, resilience, pseudo-elasticity, friction) will be altered systematically to establish what mechanical features of the environment are essential cues for caterpillars. The second will use implanted flexible electrode arrays to monitor the activity of neurons that underlie each gait. The Environmental Skeleton hypothesis predicts that motor patterns controlling the body wall muscles will change with orientation but that motor patterns controlling grip will determine the gait. Recently developed Softworm robots (entirely soft, 3D-printed crawling robots) will be used to further examine the mechanisms of gait switching and the evolution of inching behavior. Third, to begin identifying how interactions with the environment are detected by soft animals, recordings will be made from specific groups of touch and other mechanosensing neurons in the body wall to ascertain what movements they encode and how that relates to crawling and changes in gait.