This effort seeks to understand and develop strategies for effective movement in biological and synthetic locomoting systems. Gaits are a fundamental aspect of animal locomotion; examples include a horse's walking, a fish's strokes, and a snake's slithering. In these motions, the animals undergo cyclic motions which interact with the surrounding environment to gain a net displacement over each cycle. The efficacy of such gaits suggests they form a core capability in locomotion of mechanical systems. Understanding the principles of gait-based locomotion offers two opportunities: to gain deep insight into biological processes and to create sophisticated synthetic locomotors to send mechanical systems into dangerous and dirty environments. To gain this insight, questions arise: how to model locomotion, and with this model, how to both evaluate and design gaits to achieve desired locomotive capabilities? In this project, the focus will be on limbless locomotors, including snakes, slender lizards, bacteria, spermatozoa and nematode worms. Limbless locomotor controllers for confined space applications, such as search and rescue in collapsed buildings and landslide debris, will be developed.
The investigators' preliminary work reveals that geometric mechanics allows intuitive understanding of how and why gaits, produce successful locomotion. Much of the prior work with geometric tools, however, provided computationally burdensome approaches to design gaits: choose parameterized basis functions for gaits, simulate the motion of the system and then optimize the input parameters to find gaits that meet the design requirements. Such optimization with forward simulation is computationally expensive. Moreover, existing geometric approaches ignore real world considerations such as body-shape and granular (e.g., dirt) interaction between the mechanism and the environment. Therefore, the intellectual merit of this work is to advance the design and evaluation of gaits for complex systems by representing complex shapes as a basis of curvature functions, while all along empirically deriving from biological observation linear relationships between these parameters and the resulting displacement in granular media. Calculations will then take minutes rather than the days needed for multi-particle discrete element method (DEM) simulation, mitigating the challenges inherent in performing many experiments on real mechanical systems. This work will contribute to a new understanding of biological locomotors as well as help create life-life locomotion in mechanical systems.
