The unassuming term ?soft robotics? has recently made its way from the rarefied environs of academic journals and scholarly symposia and into the mainstream.
Recent articles from publications as diverse as such as IEEE Spectrum, Forbes magazine and US News and World Report, along with wide reportage in the science blogosphere, speak to the level of interest in the subject.
Clearly, we are on the verge of a soft robotics revolution. But exactly what type of revolution? Core academic research and their accompanying videos are fine, but a revolution, at least for the readers of Robotics Business Review, requires that robotics technologies make for applications and some support for commercialization. In that light, we will review this fascinating subject.
As with most formal examinations, we can begin by defining terms. More to the point? ?What is meant by soft robotics?? ?How is the subject defined?? In broad strokes, soft robotics can take three forms, and as a consequence, three different definitions:
- Soft Materials–This class of soft robots are entirely, or largely, composed of soft, compliant materials. Many systems are modeled after non-chordate biologic forms, especially soft bodied aquatic animals (cephalopods mostly), larval insects and worms.
- Soft Body, Hard Armature–These robotic systems incorporate a rigid form internally, which is then enclosed within a soft material covering. Examples include smart toys and certain classes of service robots (here, here and here).
- Hard Body, Flexible Joints–These robotic systems are composed of ridged structures that are jointed in a manner that provides for a great deal of flexibility. Examples include snakelike robots (here and here) and highly jointed manipulators (here, here and here, for example).
An analysis of all three soft robotics classes is beyond the scope of this article, and, therefore, let us focus on the newest, and potentially most promising soft robotics sector? Soft Materials Robots.
Soft Materials Robots
Soft material robots exist at the intersection of robotics technology and materials chemistry. Soft robots are differentiated from their hard body counterparts in that the former can undergo some type of elastic deformation. ?Deformation?, a term borrowed from materials science, describes the change in shape of an object when some type of force or action is applied ? mechanical, electrical, chemical, temperature and so on. ?Elastic deformation? implies that the object will return to its original shape once the force is removed.
Bioinspiration and More
For many years scientists and engineers have turned to nature as a source of inspiration for their designs. This method has been formalized as biomimetics, the study of the structure and function of biological systems as models for the engineering of machines and materials.
Biomimetics is relatively young as a formal scientific technique, although the approach has been applied informally and haphazardly in the past (Perhaps the best example of a commercial success using a biomimetic approach was the invention of Velcro in 1948 by Swiss engineer George de Mestra).
Biomimetics is now a global movement, with biologically inspired models applied to all manner of engineered products and materials.
Roboticists, too, have been inspired by biological forms, especially those that have evolved over time to provide movement such as walking, running, swimming, and flying. As a result, many systems are modeled on the wings, fins and legs of animals and insects. To date, however, most biomimetically derived robotics systems have been comprised of hard structures – external, internal or both.
But many animals, such as worms, octopuses and jellyfish, are composed largely, or exclusively, of soft materials. By employing soft bodied animals as biomimetic models, researchers have developed soft materials robotic systems that provide functionality that is not possible with more traditional robots. Extreme flexibility, deformability and stretchability make for novel capabilities and great task versatility.
It is important to note that developers of robotics systems based on biomimetic models, both rigid and soft bodied machines, are not limiting their designs only to that found in nature. Both groups are extending the designs to incorporate functionality found outside the natural realm. That is, biological principals inform design, but developers are not limiting themselves to sheer mimicry.
Inspired by DARPA
Only with the rarest of exceptions, soft materials robotics development is confined to research projects. Development at the university level is occurring throughout the world, with most funding coming from governmental bodies (see Representative Research Labs Specializing in Soft Robotics, below). Examples include the French Ministère de la Recherche et de la Technologie and the Swiss National Science Foundation. Corporate research sponsors include Boeing and Microsoft Research.
In the United States, the funding centers for soft materials robotics research is the National Science Foundation and, particularly, the Defense Advanced Research Projects Agency (DARPA). DARPA has funded a number of projects that focuses on, or intersect with, soft materials robotics. DARPA?s Maximum Mobility and Manipulation (M3), Biologically Inspired Multifunctional Dynamic Robots (BIODYNOTICS) and Chemical Communications programs provide examples.
A key funder was DARPA?s Chemical Robots program (ChemBots). The ChemBots program funded a variety of groups and resulted in multiple classes of products. The overall emphasis for the ChemBots program was the development of soft, flexible systems that move, maneuver through openings smaller than their original dimensions, and then reconstitute its size once through.
The primary impetus for soft robotics development efforts is to provide systems that offer functionally not found in traditional hard-bodied robots. Examples include:
- Fine, Dexterous Manipulation ? Many types of robotic systems require highly dexterous, precise grasping and fine manipulation of objects in an accurate, delicate, yet firm manner. Due to their extreme flexibility, deformability and pliancy, soft materials robots can grip, manipulate and flexibly control objects that are fragile, deformable or otherwise ill-suited for traditional grippers and grasping mechanisms.
- Safe Physical Human-Robot Interaction ? Increasingly, robotics systems require that they operate in close approximation with humans. Collaborative human-robot interaction necessarily demands robots that are inherently safe to work with. Lacking the unyielding componentry of traditional robotics systems, soft material robots are inherently safer for humans to work with.
- Adaptable Morphology ? Compared to rigid robots, soft materials systems are inherently more flexible and can be engineered to change their form in response to the environments in which they are operating and the tasks at hand.
- Novel Forms of Locomotion and Movement ? The flexibility of soft materials robots allow them to locomote in unstructured, irregular environments that would be prohibitive to traditional robotics forms. Soft materials robotic systems can support traditional forms of robotics movement ? wheeled, walking, rolling, hopping and slithering. Additionally soft robots can be designed to move in novel ways formally unavailable to robotic systems including undulation, oscillation, peristalsis and brachiation.
- Inexpensive Development ? Traditional robotics systems, requiring complex mechanical components and complicated, multi-jointed limb structures, are difficult and expensive to develop. Many classes of soft materials robotic systems can be engineered relatively easily and at low cost.
Currently, the field of soft robotics is in the research phase (the early research phase). Commercial products do not exist, although companies like iRobot and Boston Dynamics are working on research projects. However, the promise for breakthrough soft materials robotics capabilities are so compelling that research will continue and commercial products will eventually come to market in some form.
At this stage in the development of the soft materials robotics market, it is the lack of additional research and critical enabling technology that are the primary challenges to commercial development. Of course, it is exactly these gating factors that are the source of opportunities for researchers, businesses and the investment community. Primary among them are:
- Materials ?As their name implies, soft materials robotics will require the development of all manner of materials, some of which will incorporate actuation and sensing technology, and others will include information and programming within the materials themselves. Much more work is required to develop materials technology that is tough, yet highly flexible. Advanced polymers, shape memory alloys, nano and biomimetic materials are just a few areas where innovative solutions are required.
- Sensors ? Soft materials robotics will require the development of new classes of sensors. These technologies must necessarily be deformable ,yet robust, small in size, with little or few moving parts (and possibly with processing occurring on the sensors themselves). In addition, their power consumption must be very low, while providing excellent signal quality.
- Actuation ? A key challenge for soft materials robotics is the exertion of sufficient force to support effective actuation. Movement for soft robotics is actuated by a wide range of forces and methods including chemical, pneumatic, electromagnetic, fluidic and other ways. Regardless of the specific type of actuator, movement in soft materials robots can be improved in a number of general ways including the development of actuators whose physical design subsumes functions that are normally under direct control, integrates sensing and control directly into the actuators themselves, as well as increases the power efficiency of actuators while concurrently reducing their size.
- Control Technology ? What traditional rigid robots lack in flexibility, they make up for in relatively simple controllability. Control of soft materials robots is another matter. The difficulty arises when control systems themselves are soft. Researchers require new classes of compact, lightweight and deformable high powered actuators (see above) and new forms of controlling them.
- Simulation and Modeling of Soft Bodies ? Developers of any class of advanced robotics system require technologies to model and simulate their designs both during development and at run time. Many classes of simulation products and technologies exist for the development of traditional robots (but not enough). No such technology exists for the development of soft materials robots.
No Market, But Many Opportunities
The intersection of biomimetic principals, materials chemistry and robotics has resulted in the development of a new class of highly flexible soft-body robots composed entirely of deformable, compliant materials. These soft materials robotic systems provide a number of novel capabilities which cannot be supported in ridged-bodied systems.
Soft materials robotics is still in the very earliest of research phases. The development of commercial class soft materials robotics technologies will require advancements along a number of fronts, primarily core research and new technologies (applications will come later).
For every challenge in this exciting new field, there are multiple opportunities for researchers, technologists and businesses (and investors) to provide innovative solutions.
Representative Research Labs Specializing in Soft Materials Robotics
- Bio – Inspired Materials and Devices Laboratory (BMDL), Virginia Polytechnic Institute
- Bio Robotics Design Lab, Chuo University
- Biomimetics Laboratory, Auckland Bioengineering Institute
- BioRobotics Institute, Scuola Superiore Sant?Anna (SSSA)
- Biorobotics Laboratory (BioRob), École Polytechnique Fédérale de Lausanne (EPFL)
- Bristol Robotics Laboratory, University of Bristol
- Center for Micro-BioRobotics, Istituto Italiano Tecnologia (IIT)
- Correll Lab, Colorado University
- Creative Machines Lab, Cornell University
- Hatsopoulos Microfluids Lab, Massachusetts Institute of Technology
- Institute of Robotics and Mechatronics, German Aerospace Center (DLR)
- Jaeger Lab, University of Chicago
- James Franck Institute, University of Chicago
- Laboratory of Intelligent Systems, École Polytechnique Fédérale de Lausanne (EPFL)
- Neuromechanics and Biomimetic Devices Laboratory (BDL), Tufts University
- Robotics and Mechanisms Laboratory, Virginia Polytechnic Institute
- Soft Machines Lab, Carnegie Mellon University
- Whitesides Group Research, Harvard University