Soft actuators provide an attractive means for locomotion, gripping, and deployment of those machines and robots used in biomedicine, wearable electronics, automated manufacturing, etc. In this study, we focus on the shape-morphing ability of soft actuators made of pneumatic networks (pneu-nets), which are easy to fabricate with inexpensive elastomers and to drive with air pressure. As a conventional pneumatic network system morphs into a single designated state, achieving multimodal morphing has required multiple air inputs, channels, and chambers, making the system highly complex and hard to control. In this study, we develop a pneu-net system that can change its shape into multiple forms as a single input pressure increases. We achieve this single-input and multimorphing by combining pneu-net modules of different materials and geometry, while harnessing the strain-hardening characteristics of elastomers to prevent overinflation. Using theoretical models, we not only predict the shape evolution of pneu-nets with pressure change but also design pneu-nets to sequentially bend, stretch, and twist at distinct pressure points. We show that our design strategy enables a single device to carry out multiple functions, such as grabbing-turning a light bulb and holding-lifting a jar.

中文翻译：

---

单输入神经网络的顺序多模式变形。

软致动器为用于生物医学、可穿戴电子产品、自动化制造等的机器和机器人的运动、抓取和部署提供了一种有吸引力的手段。在这项研究中，我们关注由气动网络制成的软致动器的形状变形能力（ pneu-nets），它们很容易用廉价的弹性体制造并用气压驱动。当传统的气动网络系统变形为单一指定状态时，实现多模式变形需要多个空气输入、通道和腔室，使系统高度复杂且难以控制。在这项研究中，我们开发了一个 pneu-net 系统，它可以随着单个输入压力的增加而改变其形状为多种形式。我们通过组合不同材料和几何形状的 pneu-net 模块来实现这种单输入和多态，同时利用弹性体的应变硬化特性来防止过度充气。使用理论模型，我们不仅可以预测气动网的形状随压力变化的演变，还可以设计气动网以在不同的压力点依次弯曲、拉伸和扭曲。我们展示了我们的设计策略使单个设备能够执行多种功能，例如抓住 - 转动灯泡和握住 - 提起罐子。