Utter the word "robotics" and most people conjure up images of C-3PO from Star Wars, or the types of industrial robots used to manufacture automobiles or packaged cheese. Virtually all these robotic devices are composed of rigid mechanical systems, sometimes imbued with various forms of artificial intelligence (AI).
However, a new genre of robotic devices is taking shape. Incorporating exoskeletons and artificial skins, pneumatic logic circuits, soft actuators, flexible electronics, and entirely different engineering and design methods than their rigid cousins, soft robots are moving from the research lab to the real world.
The impact on society likely will be significant. "By mimicking the mechanical compliance and multi-functionality of soft-bodied natural organisms, soft robots can be useful for a wide array of tasks and purposes," says Carmel Majidi, Clarence H. Adamson Professor of Mechanical Engineering at Carnegie Mellon University.
Assistive and internal medicine, agriculture, food packaging, and search and rescue are just a few areas that soft robotics could touch. Yet despite enormous progress in the field, numerous challenges remain. Engineers must imbue soft robot systems with sophisticated sensory capabilities, miniaturized electronics, a power source and integrated AI.
Soft robotics represents an opportunity to design machines that bridge the human-machine divide, says Ryan Truby, an assistant professor of materials science and engineering at Northwestern University. "Building machines from soft and compliant materials opens up enormous possibilities. It can introduce new types of machines and improve many of the capabilities of today's machines."
A New Touch
Despite the fact that robots are common in manufacturing plants, surgical facilities, distribution centers, homes, and in the skies and underwater, they typically suffer from a severe limitation: they are rigid mechanical systems that repeat a single action or slight variation of the same task. Because they cannot shape-shift or easily adapt, they cannot reach places or do things that are fairly easy for living things to accomplish.
Soft robotics aims to boldly take the field where it has not gone before. Using nature as inspiration and drawing on expertise from biology, material sciences, psychology, and informatics, the idea is to build devices that conform to the contours of bodies and other objects. Instead of rigid mechanical arms and joints, soft systems incorporate pliable surfaces, skins, and other component systems, such as microfluidics, that allow them to interact in more natural ways.
The future of soft robotics is already on display inside the Pediatric Rehabilitation Technology Lab at the University of California, Riverside (UCR). Elena Kokkoni, an assistant professor in the university's Department of Bioengineering, is working to develop soft robotic systems that help infants and toddlers with limited motor function due to neuromuscular disorders learn to use their limbs.
At present, assistive devices can be bulky and somewhat intimidating. What's more, "They remain mostly passive and lack the ability to adapt in real-time. They don't provide the level of assistance that's needed," Kokkoni says.
The multi-disciplinary team at UCR already has developed different prototypes of wearable pediatric robotic exoskeletons that aid with arm mobility. The devices, which can be modified to work with shoulders, elbows, knees, and other joints, rely on pneumatics and microfluidic circuits, embedded soft actuators, and other sensors to respond to the specific needs of a child in real time, using AI and other feedback tools. "Our goal is to produce a low-power device that is safe on the body, lightweight, low profile, and makes almost no noise," she says.
While a fully functional exoskeleton remains a few years away, Kokkoni is optimistic assistive soft robotics will change healthcare. "These systems could help people with strokes or injuries rehabilitate faster and more effectively," she says.
The focus on young children is by design. Unlike adults, infants cannot provide verbal feedback—and the device fills the void by detecting non-verbal signals. Yet, "These devices aren't designed to create a dependency. They teach children how to use their bodies," Kokkoni explains.
Evolving from large, precise, motorized mechanisms to robotic systems that conform to a complex, dynamic world requires a mashup of diverse technologies. "Conventional robots are built around the idea of enabling intelligence computationally," Truby says. In order to execute tasks safely and efficiently, both the robot and the programming must be precise. "This has led us to constructing robots that are rigid."
New types of materials and components—artificial muscles and skins, soft electronics, even novel microfluidic systems—introduce opportunities to rethink designs. Yet Truby, whose research in the Robotic Matter Lab at Northwestern leans toward functional soft, polymeric, and nanoscale materials, says it also is important to reconsider the computational aspects of robotics. "When a person grabs an object like a smartphone, the process is neither precise nor ever exactly the same. We're utilizing the deformability of our skin and the mechanics of our own soft bodies to effortlessly perform this task."
For today's robots, however, handling a phone the same way is "complex and difficult," Truby points out. As a result, adapting algorithms for soft materials is essential. "We must revisit the way we algorithmically achieve autonomous behaviors to take advantage of the robot's body and optimize its performance."
At CMU's Soft Machines Lab, the focus is on designing untethered field robots that are more agile and equipped to change shape and crawl through confined spaces, even within the human body, or to combine tasks such as swimming and crawling. "We are now seeing tremendous advances in materials, architectures, and computational tools for both design and control," Majidi says. Once systems incorporate more advanced controllers and onboard electronics, the technology likely will "become a ubiquitous part of engineering."
In fact, he predicts future devices may stray even further from the idea of a standup robot, a mechanical machine in a factory, or even a Roomba that vacuums floors. Soft matter engineering methods could lead to devices that are worn on the body, and even interact with skin and internal organs. "The technology could also lead to minimally invasive surgical tools that are more fully autonomous and capable of performing physical rehabilitation, biomechanical assistance, or surgical procedures with minimal dependency on a human operator.
Concludes Truby, "We still have a ways to go to develop better actuators and appropriate power and control hardware for soft robots. But innovations in bioinspired materials will advance robotics and produce machines that perform like living organisms. Nature provides living proof about what's possible."
Samuel Greengard is an author and journalist based in West Linn, OR, USA.