Development of a quadruped robot with continuum inflatable legs

连续充气腿四足机器人的研制

• 批准号: 2598257
• 金额: --
• 依托单位: Queen Mary University of London
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2021
• 资助国家: 英国
• 起止时间: 2021 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2598257
• 关键词: Development; quadruped; robot; continuum; inflatable

With the increasing dependence on automation in virtually all parts of industry, the need for intelligent machines that can safely navigate through different terrains and environments is now more pronounced than ever. At present, one of the major difficulties surrounding soft robotics is the lack of controlled mobility. Getting soft robots to walk in a controlled way in different environments without depending on rigid limbs is a difficult challenge to overcome and requires further research.Most walking robots such as Spot by Boston Dynamics are made from rigid material. Their success in being able to walk as well as they do is dependent on heavy computer resources which are responsible for controlling how they move. Even while being armed with such computer power, Spot is still quite limited in the variety of places it can walk in. For example, according to the user's manual: it is not able to walk in wet conditions or in any situation which requires it needs to climb or descend more than 30o inclines. To put this into perspective: this is far less than the maximum angle of 42o allowed by the British Building Regulations for domestic staircases.Nature, on the other hand, has had this problem solved for millions of years. Soft robots which have attempted to capture some of nature's wisdom in this regard have already shown to be useful in a variety of highly delicate situations including surgery, rehabilitation, underwater exploration and even disaster scenarios. This project will develop on work done previously with the ambition of making a robot that can walk on four inflatable legs, similar to, for example: an octopus' limbs. Analysing relevant literature and also shows that there is great support of the hypothesis that inflatable limbs similar to biological legs will be more capable of adapting to their environment without needing massive computational power. As of this writing, no such robot exists; which highlights the importance of the proposed research.AIMS/OBJECTIVESAims:1. To manufacture four soft legs that are made of inflatable material2. To build a soft (or rigid) frame that can hold all four legs, as well as the hardware needed to power and control the way the legs move3. To use pneumatic technology for controlling the pressure in the legs, and getting them to walk using changes in pressureObjectives:The first step is to create a single inflatable leg design will be informed by previous work in the literature. Once the this is made, it will be tested based on how well it can bend using air pressure. The next step will then be to make a frame (the robot's body) which will be able to hold four of these legs and control how they move based on increasing/decreasing air pressure. Finally, the robot's performance will be tested based on how well it can walk in different environments without relying on rigid limbs or massive amounts of computer power.METHODOLOGYThe main line of research will be based on designing a fully inflatable leg that can bend based on the amount of pressure inside. This will require knowledge to be drawn from a variety of different fields in the physical sciences. To successfully make the inflatable limb, new research will need to be conducted in the field of controlled air pressurization technology so that the robot does not depend on an external air compressor. This will all need to be based on a robust network of sensors. The application of such sensors to the proposed robot will also require new research in order to be successfully implemented.EPSRC ALIGNMENTThis is a broad subject and is expected to align with the following EPSRC themes, according to https://epsrc.ukri.org/research/ourportfolio/researchareas/:Artificial intelligence technologiesControl engineeringElectrical motors and drives/electromagneticsEngineering designImage and vision computingRoboticsSensors and instrumentationSoftware engineeringSynthetic biology

随着几乎所有行业对自动化的依赖日益增加，对能够安全地穿越不同地形和环境的智能机器的需求比以往任何时候都更加明显。目前，软机器人技术的主要困难之一是缺乏受控的移动性。让软机器人在不同环境中以受控方式行走而不依赖刚性肢体是一个难以克服的挑战，需要进一步研究。大多数行走机器人，如波士顿动力公司的 Spot，都是由刚性材料制成的。他们能否成功行走取决于负责控制他们移动方式的大量计算机资源。即使配备了如此强大的计算机能力，Spot 在可行走的场所方面仍然相当有限。例如，根据用户手册：它无法在潮湿的条件下或任何需要它行走的情况下行走。攀爬或下降超过 30 度的斜坡。客观地说，这远远小于英国建筑法规所允许的家用楼梯的最大角度42°。而另一方面，大自然已经解决了这个问题数百万年。试图在这方面汲取自然智慧的软机器人已经被证明在各种高度微妙的情况下非常有用，包括手术、康复、水下探索甚至灾难场景。该项目将在之前完成的工作基础上进一步发展，目标是制造一个可以用四个充气腿行走的机器人，例如类似于章鱼的四肢。分析相关文献还表明，类似于生物腿的充气肢体将更有能力适应环境，而不需要大量的计算能力，这一假设得到了大力支持。截至撰写本文时，尚不存在这样的机器人。这强调了拟议研究的重要性。目的/目标：1。制造四个由充气材料制成的软腿2。构建一个可以容纳所有四条腿的软（或刚性）框架，以及驱动和控制腿移动方式所需的硬件3。使用气动技术控制腿部压力，并利用压力变化让腿部行走目标：第一步是创建单个充气腿部设计，这将参考文献中的先前工作。制作完成后，将根据其使用气压弯曲的程度进行测试。下一步将是制作一个框架（机器人的身体），该框架将能够容纳其中的四个腿，并根据气压的增加/减少来控制它们的移动方式。最后，机器人的性能将根据其在不同环境中行走的能力进行测试，而无需依赖僵硬的四肢或大量的计算机能力。方法论研究的主线将基于设计一条完全充气的腿，该腿可以基于内部的压力大小。这需要从物理科学的各个不同领域汲取知识。为了成功制造充气肢体，需要在受控气压技术领域进行新的研究，以便机器人不依赖外部空气压缩机。这一切都需要基于强大的传感器网络。将此类传感器应用于拟议的机器人还需要新的研究才能成功实施。EPSRC 对齐这是一个广泛的主题，预计与以下 EPSRC 主题保持一致，根据 https://epsrc.ukri.org/研究/我们的投资组合/研究领域/：人工智能技术控制工程电机和驱动器/电磁工程设计图像和视觉计算机器人传感器和仪器软件工程合成生物学