Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
Locomotion is an essential evolutive innovation of living beings that allows them to colonize and dominate the planet. As diverse as animal morphologies are (living) and were (extinct), their locomotion modalities are also diverse. In particular, animal morphologies with elongated bodies have existed on the planet since hundreds of millions of years ago. Many of them have been so successful that we can still find them. These morphologies also come in different sizes, making slight changes in how they move. However, all are influenced by the same physics of interaction with the environment. The focus of this work is precisely this mechanical interaction between elongated animals and their locomotion media. By the use of comparative biomechanics, mathematical models, and robotics, this work aims to elucidate different aspects of these elongated animals, ranging from their body structure and viscoelasticity to their sensor modalities, the rules for motion control based on interaction with the surrounding environment, their natural inherently behaviors dictated by physics, up to transitions of media and adaptation of body plans to different media. Several scientific questions and potential solutions are presented, as well as the development of engineered machines and other tools for testing such interactions as a proxy for the real animals, as experimenting directly with animals is challenging. In this work, the mathematical representation of the viscoelastic properties of a muscle implemented in a robot's motors allows it to perform more naturally and similarly to animals. Providing an understanding of how muscle properties impact performance. Furthermore, the combination of this muscle model, and force sensing working as exteroceptive local feedback wired to a model of a spinal cord, allowed the spontaneous generation of traveling waves, even in the absence of coupling between the spinal cord oscillators, which ultimately demonstrated the robustness present in eels and lampreys, even after spinal cord transections. On the other hand, I present a simplified model and experimental proof that morphological changes like adding fins to an elongated body reduce the rolling instability produced while swimming. Similarly, a simplified model related to the conservation of angular momentum adds new insights into the locomotion of large (for its size) payload volumes towards the head of animals with elongated bodies like mosquito larvae. Finally, a tool is presented to analyze comprehensively the kinematics of the center of mass that extract rich data from simple videos and structural compositions of elongated body animals. With the contributions of this work, namely experimental proofs, mathematical models, robot software, and designs, I attempt to provide state of the art in robotics and comparative biomechanics with a fresh view and novel approaches to different problems.
Jonathan Patrick Arreguit O'Neill
Auke Ijspeert, Kamilo Andres Melo Becerra, Robin Thandiackal, Laura Isabel Paez Coy, Kyoichi Akiyama