Deformation analysis and planning in soft robotics for advanced sensorimotor functionality

Abstract

This thesis focuses on advancing deformation analysis and planning in soft robotics to achieve heightened sensorimotor functionality. Soft robotics, characterized by their adaptable and compliant structures, offer solutions to limitations in rigid robotics for complex and dynamic environments. However, challenges such as nonlinear deformation patterns, material heterogeneity, and insufficient sensing and actuation strategies hinder their real-world applications. To address these issues, this research proposes a comprehensive framework integrating multimodal sensing technologies, deformation-aware planning, and bio-inspired actuation systems. Specifically, the thesis highlights the development of multi-unit and multimodal sensing arrays, such as the Dynamics-Oriented Underwater Mechanoreception Interface (DOUMI), which enables general underwater mechanical perception by capturing spatial deformation data of receptor array for comprehensive systematic analyzing. Additionally, deformable structures inspired by biological systems, such as sensillum-inspired mechanoreceptors and soft forearm-mimicking manipulators, are introduced to improve adaptability and functionality in diverse contexts. The contributions of this work span novel sensing mechanisms, tailored algorithms for nonlinear deformation control, and engineering innovations for scalable soft robotic systems. These advancements enable soft robots to perform tasks like underwater manipulation, environmental monitoring, and even facial recognition, demonstrating their interdisciplinary potential and robustness in demanding scenarios. By establishing a foundational framework, this thesis broadens the applicability of soft robotics and sets a pathway for future research in adaptive, modular, and standardized robotic designs.