Wang, Xi
(2024)
Multi-modal thin soft robots for narrow space locomotion.
PhD thesis, University of Nottingham.
Abstract
From everyday life to industrial applications, safety-critical built environments require periodic inspections to detect issues and prevent disruptions in the operation of high-value installations. Within these built environments, a specific class of obstacles presents itself as narrow access points, often at millimetre-scales. Examples include spaces beneath doors, heavy objects like containers, duct systems for air circulation, and tiny gaps within complex machinery, such as gas turbine engines. Numerous advanced robotic systems have been developed, ranging from humanoid to wheel/track-based and quadruped robots, all equipped with sophisticated intelligence and multi-modal locomotion capabilities for conducting inspection tasks. However, conventional robotic systems with bulky bodies face insurmountable challenges when it comes to navigating millimetre-scale narrow spaces. In contrast, soft robots that utilize smart materials for actuator manufacturing offer significant advantages in accessing small and confined spaces. These robots can be designed with volumes as small as centimetres or even millimetres, enabling them to navigate and operate in extremely tight and restricted environments.
Inspecting narrow spaces requires robots not only possess millimetre-scale thickness but, more importantly, multi-modal locomotion capabilities for navigating through narrow gaps and other obstacles commonly encountered in these built environments. To achieve this objective, extensive efforts have been made in the field of soft robotics to enable locomotion through techniques like in-plane shrinkage/extension or out-of-plane bending driven by pneumatics, dielectric elastomers, shape memory alloys, thermally sensitive polymers, or external magnetic fields. Among these intelligent materials, dielectric elastomer actuators (DEAs) have gained widespread use in constructing small mobile robots due to their exceptional attributes, including high power density, flexibility, and robustness. Recent studies have demonstrated various designs of DEA-driven soft robots, characterized by either miniaturized dimensions or the ability to perform multiple locomotion modes, suitable for environmental monitoring and deployment in challenging workspaces. However, to date, no DEA-based robot has been proposed with a millimetre-scale profile and capable of multimodal locomotion in narrow spaces.
This thesis project is centred on the dielectric elastomer actuator (DEA) and introduces a novel class of thin (1.7mm) soft robots, referred to as TS-Robots, capable of multi-modal locomotion. Achieving the thin profile of these robots is made possible by adopting a thin, soft DEA (TS-DEA) as the linear actuator and an electrostatic adhesive pad (EA-Pad) as the robot's foot. Specifically, a dual-actuation sandwich structure with a programmable Poisson’s ratio tensioning mechanism has been employed to develop this new family of thin-soft actuators powered by high-power-density dielectric elastomers. This empowers the TS-Robots with one or two-directional linear movement and impressive output force (approximately 41 times their own weight). As such, five TS-Robots systems have been proposed in this project for evaluating the performance of the thin-soft robots that are able for multimodal locomotion, e.g., Type-A TS-Robot, Type-B TS-Robot, Twin Type-A TS-Robot, Parallel Kinematics TS-Robot (PK-TS-Robot) and Serial Kinematics TS-Robot (SK-TS-Robot).
These TS-Robots have demonstrated the capability to access narrow spaces, crawl on horizontal surfaces, climb on walls, and navigate complex paths. Furthermore, by combining multiple TS-Robots passively or actively (as seen in Twin Type-A TS-Robot, PK-TS-Robot, and SK-TS-Robot), new configurations have been developed to transition between complex surfaces and even operate as multipurpose manipulators within confined spaces.
Actions (Archive Staff Only)
![Edit View Edit View](/style/images/action_view.png) |
Edit View |