Smart Robotic Surfaces - Multi-scale Actuation of Topographies for Adaptive Adhesion

当章鱼抓住一个物体时，它既通过包裹整个触手来抓住（宏观上），也通过吸盘环在局部粘附，然后再扭转这种粘附以释放。或者考虑一下苍蝇，它利用脚的可逆附着力着陆，抓住物体表面，然后迅速脱离物体表面。许多生物都有可逆改变粘附界面强度的机制，通过动态的局部结构变化（脚、皮肤、爪子）来进行攀爬、抓握/释放、摩擦控制和运动。很少有工程材料具有可切换附着力的例子。但是，任何使用接触粘附的技术系统，如机器人、传送带、手套和医疗设备，都可以通过可逆地改变其表面的界面粘附强度（或摩擦）的能力而从根本上改变。制造业中的输送机可以更好地操纵物体，尼龙搭扣可以打开或关闭，可切换的脚垫可以使爬行机器人，“智能”手套可以主动调整抓地力，或者新一代的医疗设备可以在手术中附着和释放软组织。该方案旨在通过驱动（移动和定位）微米和毫米尺度的表面形貌（微柱）来开发具有可切换控制和粘附强度动态优化的材料。我们开发了“动态微柱阵列”（DMA）来控制表面的机械粘附，使用气压（气动）或水压（液压）。压力的变化可以使微柱变形，使它们向外伸展或收缩，从而抓住物体表面。我们的方法使这种微柱的运动能够被数字控制，使它们成为“机器人表面”，因为它们的运动和属性可以被编程。我们还计划通过感官反馈来控制它们，“感觉”它们在表面上的粘附成功或失败。我们的目标是通过这种局部控制来更好地理解机械粘附的机制。传感将允许粘合剂DMA层优化其整体粘附强度，并积极响应粘合剂失效（滑动）。机器学习算法将被测试以响应压力传感信息。这种动态粘附的方法以前没有被证明过。这种方法的潜在应用可能对机器人技术和医学技术非常重要，我们将特别探索它在皮肤粘附方面的应用。

When an octopus holds an object, it both grips (macroscopically) by wrapping its whole tentacle, but also locally adheres with sucker rings, before reversing that adhesion for release. Or consider the fly that uses reversible adhesion of its feet to land, hold and then quickly detach from surfaces. A wide range of organisms have mechanisms to reversibly change interfacial strength of adhesion, through dynamic, local structural change (of feet, skin, claws), for climbing, gripping/release, frictional control and locomotion. There are few examples of engineered materials with switchable adhesion. But any technological system using contact adhesion, such as robots, conveyors, gloves and medical devices, could be radically changed by the ability to reversibly change the interfacial adhesion strength (or friction) of their surfaces. Conveyors in manufacturing could better manipulate objects, Velcro could be switched on/off, switchable foot pads could enable climbing robots, a `smart' glove could actively tune its grip, or a new generation of medical devices could attach and release from soft tissue in surgery. This proposal aims to develop materials with switchable control and dynamic optimization of adhesion strength through the actuation (movement and positioning) of micrometer and millimeter scale surface topographies (microposts). We have developed `dynamic micropost arrays' (DMA) to control mechanical adhesion to surfaces, using air pressure (pneumatics) or water pressure (hydraulics). The pressure change can deform the microposts to make them expand out or retract in and grip surfaces as a result. Our approach enables this micropost movement to be controlled digitally, to make them `robotic surfaces', as their movement and properties can be programmed. We also plan to control them through sensory feedback, to `feel' their successful or failed adhesion to a surface. We aim to better understand mechanisms of mechanical adhesion through this local control. Sensing will allow an adhesive DMA layer to optimize its overall adhesion strength, and actively respond to an adhesive failure (slipping). Machine learning algorithms will be tested to respond to pressure sensing information. This approach to dynamic adhesion has not been demonstrated before. Potential applications of this approach could be very significant for technologies in robotics and medicine, and we will explore its application to skin adhesion in particular.