A soft robotic approach to underwater maneuvering and manipulation

Abstract

The earth is a planet covered by water. Over the centuries researchers tirelessly pursue new ways of exploring deeper and wider to the 72\% surface of our world. Thanks to the daily improving technologies, now we are able to reach over 10,000 m down to the underwater realm with the help of underwater robots. However, underwater robotics systems are complex and expensive, which severely limit wider applications of underwater robots. Soft robotics is a new field compared with traditional rigid-bodied robots. The excellent mechanical properties of soft matters have shown great potential in this field. Specifically, the natural waterproofing and compliance of soft robots make them ideal for underwater applications. In this case, this thesis is going to address some of the existing challenges of underwater robots using a soft robotic approach. We try to use the compliance of soft robots to tackle the challenges of underwater robots in three aspects: actuation, sensing, and control. For actuation, we have developed five different actuators serving different purposes. The Fiber-reinforced Soft Muscle (FRSM) and Ring-shaped Bracing Actuator (RBA) can supply much higher force output compared with those made out of pure soft matters. The Smart Bending Actuator (SBA) and Black Origami Fluidic Elastomer Actuator (BOFEA) are combined with soft strain sensors and can give accurate position feedback. And the Compensating Soft Bladder (CSB) can be used in the hydraulic control framework to achieve multiple purposes. For sensing, we separate the sensing layers and substrates with the help of the compliance of soft matters. For both sensors developed in the thesis (a soft stretchable bending sensor and a optical sensor with soft waveguide), we achieve decoupling bending \& stretching and cost-effectiveness. For underwater hydraulic control, we propose a new hydraulic control framework, which can achieve compactness, fast dynamic performance, and ambient pressure independence at the same time. Finally, we explore the potential of underwater maneuvering and manipulation based on the above contributions. The astonishing aquatic animals are firstly studied and we get inspiration from cephalopods and fishes that use body/caudal fin propulsion. For the cephalopod-inspired robot, we try to mimic the overall compact biological design, which results in a compact biomimetic underwater robot. The robot uses a vortex-based propulsion mechanism to generate thrust, and have a SBA-based turning mechanism. For body/caudal propulsion, we try to mimic the locomotion of Ostraciiform and Thunniform. We successfully achieved a steady swim pattern with oscillation frequency up to 70 Hz by applying a compliant two-segment fin design. And we built a robotic fish that has the swimming pattern of Thunniform and can reach a maximum speed of 1.87m/s. We also investigate the potential application of underwater manipulation. In total we built two manipulation systems. The BOFEA-based manipulation can give accurate position feedback from each actuator using the soft waveguide-based optical sensors. The FRSM-based manipulation system is a complete underwater manipulation system. The features of compactness, fast response, and ambient water pressure independence are all validated through experiments. Finally, both manipulators are tested feasible for underwater operations.