This manuscript describes a unique class of locomotive robot: A soft robot, composed exclusively of soft materials (elastomeric polymers), which is inspired by animals (e.g., squid, starfish, worms) that do not have hard internal skeletons. Soft lithography was used to fabricate a pneumatically actuated robot capable of sophisticated locomotion (e.g., fluid movement of limbs and multiple gaits). This robot is quadrupedal; it uses no sensors, only five actuators, and a simple pneumatic valving system that operates at low pressures (< 10 psi). A combination of crawling and undulation gaits allowed this robot to navigate a difficult obstacle. This demonstration illustrates an advantage of soft robotics: They are systems in which simple types of actuation produce complex motion.

Multigait soft robot

In areas from assembly of machines to surgery, and from deactivation of improvised explosive devices (IEDs) to unmanned flight, robotics is an important and rapidly growing field of science and technology. It is currently dominated by robots having hard body plans—constructions largely of metal structural elements and conventional joints—and actuated by electrical motors, or pneumatic or hydraulic systems. Handling fragile objects—from the ordinary (fruit) to the important (internal organs)—is a frequent task whose importance is often overlooked and is difficult for conventional hard robots; moving across unknown, irregular, and shifting terrain is also. Soft robots may provide solutions to both of these classes of problems, and to others. Methods of designing and fabricating soft robots are, however, much less developed than those for hard robots. We wish to expand the methods and materials of chemistry and soft-materials science into applications in fully soft robots. A robot is an automatically controlled, programmable machine. The limbs of animals or insects—structures typically based on rigid segments connected by joints with constrained ranges of motion—often serve as models for mobile elements of robots. Although mobile hard robots sometimes have limb-like structures similar to those of animals (an example is “Big Dog” by Boston Robotics), more often, robots use structures not found in organisms—for example, wheels and treads. The robotics community defines “soft robots” as: 1) machines made of soft—often elastomeric—materials, or 2) machines composed of multiple hard-robotic actuators that operate in concert, and demonstrate soft-robot-like properties; here, we consider only the former. Soft animals offer new models for manipulation and mobility not found, or generated only with difficulty and expense, using hard robots. Because materials from which this class of devices will be fabricated will usually be polymers (especially elastomers), they fall into the realm of organic materials science. The use of soft materials allows for continuous deformation. This type of deformation, in turn, enables structures with ranges of motion limited only by the properties of the materials. Soft robots have the potential to exploit types of structures found, for example, in marine organisms, and in non-skeletal parts of land animals. The tentacles of squid, trunks of elephants, and tongues of lizards and mammals are such examples; their structures are muscular hydrostats. Squid and starfish 14] are highly adept locomotors; their modes of movement have not been productively used, and permit solutions of problems in manipulation, locomotion, and navigation, that are different from those used in conventional hard robotics. The prototypical soft actuator—muscle—developed through the course of evolution. There is currently no technology that can replicate the balanced performance of muscle: it is simultaneously strong and fast, and enables a remarkable range of movements (such as those of a tongue). Muscle-like contraction and dilation occur in ionic polymeric gels on changes in the acidity or salinity of a surrounding ionic solution, but actuation in macroscopic structures is masstransport limited, and typically slow. Other electroactive polymers (EAPs) include dielectric elastomers, electrolytically active polymers, polyelectrolyte gels, and gel-metal composites. Pneumatically-driven McKibben-type actuators are among the most highly developed soft actuators, and have existed for more than fifty years; they consist of a bladder covered in a shell of braided, strong, inextensible fibers. These actuators can be fast, and have a length-load dependence similar to that of muscle but possess only one actuation mode—contraction and extension when pressurization changes. They are, in a sense, an analogue to a single muscle fibril ; using them for complex movements requires multiple actuators acting in series or parallel. Pneumaticallydriven flexible microactuators (FMAs) have been shown to be capable of bending, gripping, and manipulating objects. Roboticists have explored scalable methods for gripping and manipulating objects at the micro and nano scales. The use of compliant materials allows grippers to manipulate objects such as fruit with varied geometry. The field of robotics has not yet caught the attention of soft-materials scientists and chemists. Developing new materials, techniques for fabrication, and principles of design will create new types of soft robots. The objective of this work is to demonstrate a type of design that provides a range of behaviors, and that offers chemists a test bed for new materials and methods of fabrication for soft robots. Our designs use embedded pneumatic networks (PneuNets) of channels in elastomers [*] Prof. G. M. Whitesides Wyss Institute for Biologically Inspired Engineering Harvard University, 3 Blackfan Circle, Boston, MA 02115 (USA) Fax: (+ 1)617-495-9857 and Kavli Institute for Bionano Science & Technology 29 Oxford Street, Cambridge MA (USA) E-mail: gwhitesides@gmwgroup.harvard.edu Homepage: http://gmwgroup.harvard.edu/

/pdf/soft-robotics-for-chemists-2q91i30o62.pdf

Soft robotics for chemists

Legged robots pose one of the greatest challenges in robotics. Dynamic and agile maneuvers of animals cannot be imitated by existing methods that are crafted by humans. A compelling alternative is reinforcement learning, which requires minimal craftsmanship and promotes the natural evolution of a control policy. However, so far, reinforcement learning research for legged robots is mainly limited to simulation, and only few and comparably simple examples have been deployed on real systems. The primary reason is that training with real robots, particularly with dynamically balancing systems, is complicated and expensive. In the present work, we introduce a method for training a neural network policy in simulation and transferring it to a state-of-the-art legged system, thereby leveraging fast, automated, and cost-effective data generation schemes. The approach is applied to the ANYmal robot, a sophisticated medium-dog-sized quadrupedal system. Using policies trained in simulation, the quadrupedal machine achieves locomotion skills that go beyond what had been achieved with prior methods: ANYmal is capable of precisely and energy-efficiently following high-level body velocity commands, running faster than before, and recovering from falling even in complex configurations.

/pdf/learning-agile-and-dynamic-motor-skills-for-legged-robots-41hdxn2m1m.pdf

Learning agile and dynamic motor skills for legged robots

A new versatile hydraulically powered quadruped robot (HyQ) has been developed to serve as a platform to study not only highly dynamic motions, such as running and jumping, but also careful navigat...

Design of HyQ – a hydraulically and electrically actuated quadruped robot:

Soft robots are capable of mimicking the complex motion of animals. Soft robotic systems are defined by their compliance, which allows for continuous and often responsive localized deformation. These features make soft robots especially interesting for integration with human tissues, for example, the implementation of biomedical devices, and for robotic performance in harsh or uncertain environments, for example, exploration in confined spaces or locomotion on uneven terrain. Advances in soft materials and additive manufacturing technologies have enabled the design of soft robots with sophisticated capabilities, such as jumping, complex 3D movements, gripping and releasing. In this Review, we examine the essential soft material properties for different elements of soft robots, highlighting the most relevant polymer systems. Advantages and limitations of different additive manufacturing processes, including 3D printing, fused deposition modelling, direct ink writing, selective laser sintering, inkjet printing and stereolithography, are discussed, and the different techniques are investigated for their application in soft robotic fabrication. Finally, we explore integrated robotic systems and give an outlook for the future of the field and remaining challenges. 3D printing can be used to directly fabricate soft robots. This Review discusses advances in 3D printing technologies and soft materials for the fabrication of soft robotic systems with sophisticated capabilities, such as 3D movement and responsiveness to the environment.

3D printing of soft robotic systems

https://www.cs.swarthmore.edu/~meeden/DevelopmentalRobotics/bigdog.pdf

BigDog, the Rough-Terrain Quadruped Robot

The center of mass (CoM) has been widely used in planning and control for humanoid locomotion, because it carries key information about the position of a robot. In contrast, an ''angular center of mass'' (ACoM), which provides an ''average'' orientation of a robot, is less well-known in the community, although the concept has been in the literature for about a decade. In this paper, we introduce the ACoM from a CoM perspective. We optimize for an ACoM on the humanoid robot Nadia, and demonstrate its application in walking with natural upper body motion on hardware.

Angular Center of Mass for Humanoid Robots

Complex multibody legged robots can have complex rotational control challenges. In this paper, we propose a concise way to understand and formulate a whole-body orientation that (i) depends on system configuration only and not a history of motion, (ii) can be representative of the orientation of the entire system while not being attached to any specific link, and (iii) has a rate of change that approximates total system angular momentum. We relate this orientation coordinate to past work, and discuss and demonstrate, including on hardware, several different uses for it.

Gabriel Nelson

Papers

BigDog, the Rough-Terrain Quadruped Robot

Angular Center of Mass for Humanoid Robots

Integrable Whole-body Orientation Coordinates for Legged Robots