Está en la página 1de 11

SCIENCE ROBOTICS | REVIEW

SOFT ROBOTICS 2016 © The Authors,


some rights reserved;

Soft robotics: Technologies and systems pushing the exclusive licensee


American Association
boundaries of robot abilities for the Advancement
of Science.

Cecilia Laschi,1* Barbara Mazzolai,2 Matteo Cianchetti1

The proliferation of soft robotics research worldwide has brought substantial achievements in terms of principles,
models, technologies, techniques, and prototypes of soft robots. Such achievements are reviewed here in terms of
the abilities that they provide robots that were not possible before. An analysis of the evolution of this field shows
how, after a few pioneering works in the years 2009 to 2012, breakthrough results were obtained by taking seminal tech-
nological and scientific challenges related to soft robotics from actuation and sensing to modeling and control. Further
progress in soft robotics research has produced achievements that are important in terms of robot abilities—that is, from
the viewpoint of what robots can do today thanks to the soft robotics approach. Abilities such as squeezing, stretching,
climbing, growing, and morphing would not be possible with an approach based only on rigid links. The challenge ahead
for soft robotics is to further develop the abilities for robots to grow, evolve, self-heal, develop, and biodegrade, which are
the ways that robots can adapt their morphology to the environment.

INTRODUCTION (14).” This definition refers to a seminal paper (14) in which con-
Robotics is the science of building machines with desired abilities of tinuum robots are defined as capable of bending via elastic deforma-
movement, perception, and cognition, and robots are machines tion. In Trivedi et al. (13), soft robots are distinct from traditional
intended to perform tasks in the service of human beings. They were robots, which have rigid links, and serpentine robots, which have
recently revealed to be excellent tools for research in science (1). Robot a large number of short rigid links and degrees of freedom: Thus,
abilities have progressed in terms of the complexity and the precision soft robotics has primarily involved construction from soft materials
of movements, as well as the accuracy of motion control and the level and intentionally compliant structures. The review in (9) simply re-
of intelligence (2), and such a progression of abilities is the ultimate fers to soft-bodied robots in an analogy with soft-bodied animals. In
goal in robotics. Although especially advanced in the industrial (15), soft robots are defined as robots built with soft materials. Re-
manufacturing context, robots are becoming ubiquitous, and so calling the definition given by the RoboSoft (16) community, in (17),
they are expected to soon be used in many tasks and scenarios. The soft robots are defined as “soft robots/devices that can actively inter-
use of soft matter for building robots (3) has been recognized as the act with the environment and can undergo “large” deformations rely-
current challenge (4) for pushing the boundaries of robotics tech- ing on inherent or structural compliance.” The review in (5) instead
nologies and building robotic systems for service tasks in natural en- defines soft robots in terms of the Young’s modulus of a material:
vironments (5). Softness can be intended in various ways: soft texture, “systems that are capable of autonomous behavior, and that are pri-
soft and deformable materials, soft movement, elastic materials and marily composed of materials with moduli in the range of that of soft
variable compliance actuators, and soft in friendly and natural inter- biological materials.” The definition of soft robotics given in (3) uses
actions with people (6). The compliance and the elasticity of soft body both the well-known concept of “soft matter” applied to materials and
parts allow reactions with interaction forces without control and the term “soft-matter robotics.” Here, we refer to the RoboSoft defini-
support the bioinspired approach that is increasingly being adopted tion, which focuses on compliance and deformability in the interac-
in robotics, as well as in many other disciplines (7). The study of living tion with the environment.
organisms can shed light on principles that can be fruitfully adopted One of the main benefits of the compliance possessed by soft robots
to develop additional robot abilities or to facilitate more efficient is that they can implement embodied intelligence principles (for exam-
accomplishment of tasks (8), because living organisms exploit soft ple, preflexes). They can also conform to surfaces or objects, absorb
tissues and compliant structures to move effectively in complex natural energy to maintain stability, and exhibit physical robustness and human-
environments (9). safe operation at potentially low cost. Figure 1 shows a pictorial spectrum
The term “soft robotics” was formerly used to indicate robots of soft robot approaches: from robots built mostly from rigid materials
with rigid links and mechanically (or passively) compliant joints (with rigid links and joints and few intentionally compliant structures) to
with variable stiffness (10), or compliance or impedance control completely soft robots. The spectrum passes through a variety of robots
(11), which is beyond the scope of this paper. Soft robotics was then built with soft materials or deformable structures, with few rigid parts for
used to underline the shift from robots with rigid links to bioinspired support or integration of components. Such soft robots compose the vast
continuum robots that are “inherently compliant and exhibit large majority today, and they are expected to be used most in diverse soft
strains in normal operations” (12). Then, the following definition robotics applications.
was given (13): “soft robotic manipulators are continuum robots made We review the current achievements of soft robotics, showing how
of soft materials that undergo continuous elastic deformation and pro- the use of soft materials or deformable structures stands as a gateway
duce motion through the generation of a smooth backbone curve toward more advanced or more efficient robot abilities that were not
possible before. We first present a review of how this field evolved,
1 reporting on the first examples and publications of relevant soft robot-
BioRobotics Institute, Scuola Superiore Sant’Anna, Pisa, Italy. 2Center for Micro-
BioRobotics, Istituto Italiano di Tecnologia, Pontedera, Italy. ics technologies and systems but excluding the soft robots ante litter-
*Corresponding author. Email: cecilia.laschi@sssup.it am, that is, the very first examples of soft robots, dating back to the

Laschi, Mazzolai, Cianchetti, Sci. Robot. 1, eaah3690 (2016) 6 December 2016 1 of 11


SCIENCE ROBOTICS | REVIEW

Fig. 1. Spectrum of soft robots approaches grading from mostly stiff with a few selectively compliant elements (left) to entirely soft (right). From left: the iSprawl robot
with flexible elements to accommodate external inputs and repositioning of the legs (97); X-RHex with variable compliant legs to optimize running speed (98); soft gripper
counting on four flexible fluidic fingers lodged on rigid handle (99); soft robotic fish with flexible tail actuated by FFAs (100); OCTOPUS is an underwater robot composed of a
rigid frame where eight soft arms are lodged (six tendon-driven arms for locomotion and two SMA-based arms for manipulation) (17); the PoseiDRONE robot swims using the
tendon-based soft thruster, and it has four arms for locomotion on the seafloor (101); universal gripper based on granular jamming phenomenon, thus granular material, an elastic
membrane and a rigid interface for the application vacuum (31); origami robot can change its shape through the activation of shape memory materials and composed of rigid
polymers or paper but flexible along bending lines (102); Tuft Softworm is a bioinspired soft robot that can replicate a caterpillar-like locomotion through the sequential activation
of SMA actuators (103); rehabilitation glove for human hands based on FFAs (PneuNets) (56); inflatable robots are based on multiple inflatable chambers whose shape determines
the robot motion (52); Octobot is completely made of soft elastomers and actuated by chemical reactions (104). [Credits (left to right): iSprawl, M. Cutkosky; X-RHex, K. C. Galloway;
soft robotic fish, D. Rus; OCTOPUS robot, photographer J. Hills; PoseiDRONE robot, photographer M. Brega; universal gripper, 2012 IEEE, reprinted with permission from (31);
Origami robot, J. K. Paik; Softworm, B. A. Trimmer; soft robotic glove, P. Polygerinos; inflatable robot, 2012 IEEE, reprinted with permission from (52); Octobot, J. A. Lewis].

’80s and ’90s, not named as such (18, 19). We then present an analysis ential SMA (26) springs, which contract and deform a mesh struc-
of the current state of the art in terms of robot abilities and the latest ture, alternatively increasing and decreasing in diameter and length
achievements, described from the viewpoint of what robots can do to- (27). The final soft robot showed peristaltic locomotion at a speed of
day thanks to the soft robotics approach. 3.47 mm/s (with a theoretical speed of 3.99 mm/s) and demonstrated
resistance to several impacts of a rubber hammer (28).

SOFT ROBOTICS EVOLUTION: SEMINAL The challenge of stiffening


TECHNOLOGICAL ACHIEVEMENTS A mass of granular material encased in an elastic membrane can pass
Soft robotics stems from robotics, especially in the quest for robots from being a very soft body to a completely stiff one when a vacuum is
operating in our natural environment, in the service of human beings. applied. The stiffness can increase up to 40-fold [for example, from 0.2
At the same time, the embodied intelligence paradigm (20) suggests to 7.8 N/mm (29)]. This principle can be used as a mechanism for
that a soft body partly controls movements by deforming, adapting, shape changing, mobility (30), and, most importantly, stiffness varia-
and reacting to interaction forces. The desired behavior can be ob- tion (31). It was the first solution proposed in soft robotics, although
tained efficiently through morphological computation by exploiting others are used today, as reviewed in (32) and discussed in the next
interaction forces and compliance instead of fighting them (21). section. A soft, granular jamming body can deform and adapt, for ex-
ample, to the shape of an object to grasp; when stiffened, it can keep
The first soft robots: The challenge of actuation its shape. This wide-range, simple to control, energetically efficient
Good examples of the first soft robots are the arms inspired by the mus- stiffening mechanism is now widely used in soft robotics.
cular hydrostats of the arms and tentacles of cephalopods and the
trunks of elephants, such as the OctArm, first presented in (22) and The birth of the Pneu-Net approach and the
then developed in increasingly refined versions up to version VI fabrication challenge
(12, 13). The OctArm is pneumatically actuated, and it has a backbone Although pneumatic actuation has long been used in robotics, the so-
without joints. The continuum yet strong structure allows it to lift and called Pneu-Net approach has been proposed recently, and it is widely
manipulate objects with different dimensions and weights. Trivedi used today in soft robotics: By appropriately designing chambers in
et al. (12) report a few other examples of soft robots based on EAP elastomeric polymers, pneumatic actuation can deform soft bodies
[electroactive polymers; (23)] and PAM [pneumatic artificial muscles; in a controlled way. The first example, presented in (33), was a pneu-
(24)] technologies that can bend, contract, and extend. Another exam- matically actuated robot composed exclusively of soft materials
ple is the caterpillar-like robot built with silicone rubber and actuated (elastomeric polymers) and capable of quadrupedal locomotion. It dem-
by shape memory alloy (SMA) springs: It was first presented in (25) onstrated some of the advantages of soft robots, such as the deform-
and then evolved toward more advanced versions. SMA springs have ability that is useful for passing through small apertures. This work
also been used to replicate an interesting mechanism for soft robot also proposed fabrication technologies for soft robots, which need un-
locomotion that exploits deformability based on peristalsis. The hydro- conventional approaches. For the most part, soft materials cannot ac-
stat skeleton structure of Oligochaeta (for example, earthworm) has been tively be shaped—that is, they cannot be machined, sculpted, ground,
implemented in a robot (Meshworm) with longitudinal and circumfer- or forged; instead, they have to be molded, as in (33), or extruded.

Laschi, Mazzolai, Cianchetti, Sci. Robot. 1, eaah3690 (2016) 6 December 2016 2 of 11


SCIENCE ROBOTICS | REVIEW

Multimaterial three-dimensional (3D) printers that can work with a mix eters related to the interaction with the environment. Interesting
of hard and soft materials are making soft robots easier to fabricate to- approaches to coping with this challenge are the use of evolutionary
day, although finishing of soft robot surfaces cannot achieve the accu- algorithms for exploring such a huge design parameter space and the
racy of machined surfaces, and the mechanical properties of the printable development of methods for the automatic design of soft robots (53).
materials are still limited (for example, in terms of elongation at break). Evolution is simulated with genetic algorithms, and evolutionary ro-
botics has achieved important results, encouraging the application of
The octopus-inspired challenges of actuation, sensing, this approach to the design and construction of new robots. Mor-
modeling, and control phological changes within the same individual robot are still under
Research inspired by the common octopus (34) faced a number of investigation. Recent results with simulated robots that evolve by
challenges in the development of soft components and integrated soft enlarging/shrinking their body units show how soft materials help
robotics systems. With a completely boneless body, an octopus can the robot morphology to change in response to a stimulus by taking
stretch and stiffen with enough strength for grasping, manipulation, into account morphological computation (54).
locomotion, and swimming, thus providing an ideal model for new
principles for the actuation and control of soft materials, such as EAP First applications of soft robots
muscles (35) and SMA springs (36) arranged in an octopus-like mus- While enabling technologies have been progressing, the soft robot-
cular hydrostat (37). The development of the integrated eight-arm ics approach has been used for building integrated systems in different
OCTOPUS robot (17) started a number of directions for the investi- application scenarios. For example, in the biomedical field, surgical
gation of soft robotics and produced the first results for fields that are instruments are an ideal application for soft robotics (55), and soft
now extensively investigated in soft robotics, such as modeling and robotics is finding applications in rehabilitation (56) and assistance
controlling of these continuum structures (38, 39) and their fluidody- (57–59). Other interesting areas for the application of soft robots
namics (40), exploiting their morphological computation (41), fabri- are explorations and rescues (60, 61), which are deeply related to un-
cating a stretchable sensitive skin (42), understanding and structured environments.
implementing the sucker adhesion mechanisms in water (43, 44),
and deriving principles for locomotion (45) and swimming (46).
Modeling and controlling soft robots have been approached by using WHERE SOFT ROBOTS DARE: ABILITIES THAT SOFT ROBOTS
the techniques for modeling of continuum robots (47, 48), because HAVE OBTAINED
the techniques for modeling the kinematics and the dynamics of tra- Starting from the pioneering works described in the previous sections,
ditional multilink robots become inapplicable. The modeling work in soft robotics has achieved important results that we analyze here in
(38) was based on a geometrically exact Cosserat approach, whereas terms of the abilities provided to robots. Such abilities attained by soft
the work in (39) used the model for applying a Jacobian method to robots are not just different ways to perform usual robot functions,
control positioning. Sensorizing soft robots also requires un- such as grasping or locomotion; they are capabilities for performing
conventional approaches for detecting deformations in soft robots actions, such as squeezing, stretching, climbing, and growing, that
without introducing rigid components into the system, such as flex- would not be possible with an approach to robot design based on rigid
ible and stretchable electronics and sensors, which have been inves- links only. In our definition, an ability is not related to a possible
tigated since 2004 (49). Among the first works in flexible sensors for application. Instead, applications of an ability can eventually material-
soft robots was the development of a contactless deflection sensor ize in diverse fields. The integration of soft-matter technologies with
with a light-emitting diode (LED) element and a photodiode placed morphological computation and embodied intelligence paradigms has
onto two connected substrates. The deflection angle between the two substantially contributed to an increase in the abilities of soft robots,
planes can be extracted from the LED light intensity detected at the which show improved performance in real environments compared
photodiode because of the bell-shaped angular intensity profile of the with their more traditional “relatives.” Table 1 summarizes the robot
emitted light, which has a range of deflection of 105° to 180° with a res- abilities obtained by exploiting soft materials and deformable
olution close to 1°. This sensor can be used to detect the deformations of structures, gives references to the robots described in the following
a soft robot body (50). Flexible electronics for stretchable circuitry is the and the related bibliography, and recalls the key principles and tech-
key for building “origami” robots (51) with actuators consisting of an an- nologies adopted.
nealed SMA actuator and a heater, with the capability of changing shape,
that is, folding themselves from a 2D to a 3D shape. Jumping
An interesting jumping ability (up to six body heights) is achieved in
Inflatable robots the combustion-powered robot presented in (62), whose body is based
A quite interesting approach is the use of textile materials for building on a stiffness gradient obtained by multimaterial 3D printing (Fig.
inflatable robots. After the seminal work from the ’80s reported in 2A). By using explosive chemical reactions, the actuation of Pneu-
(18), a manipulator was presented in (52), whose design was based Net was greatly improved (63) in terms of speed and power, facilitat-
on a strong fabric and a tight chamber that could be inflated. Actua- ing jumping abilities in soft robots.
tion was carried out with the aid of tendons and pulleys. GoQBot is a soft robot inspired by some species of caterpillars that
can show rolling locomotion as a jumping escape mechanism, which
Evolutionary approaches for soft robotics is among the fastest self-propelled wheeling behaviors in nature (Fig.
The soft approach to building robots soon enlarged the space of de- 2B) (64). Inspired by this behavior, GoQBot can reach 1g acceleration
sign parameters to a level that is difficult for a human designer to and an angular velocity of more than 200 rpm. This capability of
manage. They include morphological parameters, a number of adapting to perform different behaviors also makes GoQBot an exam-
mechanical properties, and environmental specifications and param- ple of a morphing robot.

Laschi, Mazzolai, Cianchetti, Sci. Robot. 1, eaah3690 (2016) 6 December 2016 3 of 11


SCIENCE ROBOTICS | REVIEW

Table 1. Summary of the robot abilities reviewed, the robots they are attributed to, and the key technologies or principles that allowed the
achievement of these abilities.
Ability Applied on Key technology/key principle Reference

GoQBot Fast-contracting SMAs (64)

Jumping Tripedal robot Pneu-Net powered by combustion (63)

Hemispheroid robot Functionally graded and powered by combustion (62)

Meshworm Flexible braided sleeve and SMAs (65)


Peristaltic locomotion
Softworm Flexible braided mesh and tendons (66)

Quadrupedal robot Pneu-Net (33)


Locomotion through
small crevices
Cockroach-like platform Origami-style structure (69)

Resilience Quadrupedal robot Pneu-Net (61)

Arm elongation and shortening OCTOPUS robot Flexible braided sleeve and SMAs (36)

OCTOPUS robot (45)


Underwater legged locomotion Octopus-inspired pushing by stiffening and elongation
PoseiDRONE robot (71)

Pulsed-jet swimming PoseiDRONE robot Vortex rings produced with body squeezing (46)

Sensing skin of (73)


Stretchability HLEC
crawling robot

Hierarchical structures, directional adhesion, and control of tangential (67)


Climbing Stickybot
contact forces

Growing PLANTOID Additive manufacturing (74)

Combined bending STIFF-FLOP (77)


FFAs and granular jamming
and stiffening manipulator

RBO Hand 2 PneuFlex actuators (83)


Adaptable grasping
Universal gripper Granular jamming phenomenon (31)

(78)
Self-deployment Mechanism Origami-style structure and shape memory materials
(79)

Morphing, self-reconfigurability Manipulator HMAs (86)

DA polymers (88)
Self-healing Soft pneumatic actuators
Polyaramid fibers in an elastomeric matrix (89)

Biodegradability and edibility Actuator Biodegradable hydrogels (90)

Peristaltic locomotion Climbing


Studies of how animals move using peristaltic locomotion have sug- A bioinspired approach coupled with soft-matter technologies led to
gested that a sinusoidal deformation of a soft continuous structure is the development of robots with the ability to climb walls and vertical
the most suitable way to implement it. Continuous wave peristaltic surfaces. This is the case of Stickybot (Fig. 2D) (67), a gecko-inspired
motion has been achieved by exploiting sleeves and tubular meshes com- robot that exploits some of the enabling features of its animal coun-
posed of thin flexible threads (such as the one used in McKibben terpart. The ability of the animal to climb with dry adhesion is as-
actuators). Such a structure is a key element for generating peristaltic- sured by a combination of functions, which include the use of van
like motion because it has the capability to translate a radial defor- der Waals forces generated by an intimate conformability of hierar-
mation into a longitudinal one. A reduction of the diameter is thus chical system of lamellae, setal stalks, and spatular tips to both rough
transformed into an elongation. By combining several actuation and smooth surfaces; directional and controllable position of the ad-
units lodged along the structure and activating them in the right hesive structures; a high cycle rate of attachment release (steps per
sequence, it is possible to generate a continuous wave of deforma- second); and self-cleaning of the gecko’s feet, aided by the smaller
tion. So far, this mechanism has been implemented using SMAs in dimensions of the terminal spatula with respect to the size of dirt
the Meshworm robot (65) and motor-driven tendons in the Soft- particles (68). Stickybot uses several design principles translated from
worm robot (Fig. 2C) (66). the animal, such as the hierarchical structures, directional adhesion,

Laschi, Mazzolai, Cianchetti, Sci. Robot. 1, eaah3690 (2016) 6 December 2016 4 of 11


SCIENCE ROBOTICS | REVIEW

Fig. 2. (A) 3D-printed stiffness-graded robot jumping by combustion (62). (B) GoQBot morphing into a Q shape and jumping (64). (C) Softworm peristaltic robot (66).
(D) Stickybot climbing on a glass surface by exploiting the hierarchical microstructure of its soft feet (67). [Credits (A to D): N. W. Bartlett; GoQbot, B. A. Trimmer;
Softworm peristaltic robot, R. D. Quinn; Stickybot, M. Cutkosky].

and control of tangential contact forces, to achieve adhesion control. Underwater legged locomotion
Bioinspired force control strategies, innovative fabrication methods, Although legged locomotion has been extensively investigated in ro-
and soft-matter technologies supply the robot with compliant struc- botics as an appropriate way to negotiate rough terrain, underwater
tures on the toes (based on arrays of small, angled polymer stalks) legged locomotion, which could allow exploration of the benthic zone,
and on other parts of the body, which balance forces among the feet is still an open challenge. With inspiration from the octopus, models for
and promote smooth attachment and detachment of the toes. underwater legged locomotion have been developed on the basis of ap-
propriate compliance of the legs. The OCTOPUS and PoseiDRONE soft
Squeezability: Locomotion through small crevices robots, with silicone-based legs, show adaptive walking behavior (45, 71).
and resilience
Quadrupedal locomotion has been implemented in a soft tetrapod ro- Pulsed-jet swimming
bot that can control each leg independently by using a network of Cephalopods also inspired pulsed-jet swimming, which exploits the
pneumatic channels for each limb and another channel dedicated to deformability of their soft bodies. Water expelled through a funnel
the spine of the robot (33). The completely soft robot, fabricated using generates propulsion, which is optimized by body and funnel size
soft lithography, shows the ability to locomote through small crevices and by the frequency of the contractions that expel water. A model
(Fig. 3A). On a larger scale, this approach was implemented in a soft of the deformations of a soft body combined with a model of the flu-
robot made of 0.65-m-long silicone elastomer, which is innately resil- idodynamics of pulsed-jet propulsion allowed for the design of an op-
ient to a variety of adverse environmental conditions, including snow, timized propelling body in the PoseiDRONE soft robot with pulsed-jet
puddles of water, direct (albeit limited) exposure to flames, and the swimming abilities (Fig. 3D) (72).
crushing force of being run over by an automobile (Fig. 3B) (61).
The same ability of squeezing to fit small apertures is exhibited by Stretchability
a robot inspired by cockroaches (69). Their jointed exoskeletons per- Softness, stretchability, and conformability in mollusks have recently
mit rapid appendage-driven locomotion but retain their soft-bodied, led to the development of a hyperelastic light-emitting capacitor
shape-changing ability. Like the animal, the robot successfully loco- (HLEC) that enables both light emission and touch sensing in a thin
motes in vertically confined spaces by compressing its body by half rubber sheet, which stretches to >480% strain (73). These HLECs are
(54%; 75 to 35 mm). The robot shell can withstand compressive composed of layers of transparent hydrogel electrodes sandwiching a
forces (a 1-kg weight) that are 20 times its body mass through its ZnS phosphor–doped dielectric elastomer layer, creating thin rubber
overlapping abdominal plates, and it can dissipate impacts and colli- sheets that change illuminance and capacitance under deformation
sions, demonstrating the capabilities of robots constructed using soft (Fig. 4A). These sheets were integrated into the skin of a soft robotic
materials (Fig. 3C). worm made of three pressurized chambers, providing it with highly
stretchable deformability, dynamic coloration, and sensory feedback
Elongation and shortening from external and internal stimuli.
Softness, stretchability, and conformability are some of the capabilities
that make cephalopods such as octopuses as models for excellence in Growing
soft robotics. The octopus arm is an ideal model for robot arms that Recently, plants have also been considered models in bioinspired and
can bend in any direction, elongate, and shorten. Such abilities have soft robotics. The abilities of these organisms to grow for their entire
been reached in the OCTOPUS robot with the longitudinal and trans- life by incremental addition of new material have been exploited in a
versal arrangement of SMA springs in a braided conical sheath, which new generation of robots inspired by roots (that is, plantoids) (Fig. 4B)
takes inspiration from the muscle arrangement of the mollusk muscu- that can grow at the tip region of the robotic root using layer-by-layer
lar hydrostat (37). The OCTOPUS arm can reduce its diameter by deposition achieved with an additive manufacturing technique similar
25%, consequently elongating by 41.3% (70). to fusion deposition modeling (Fig. 4C) (74).

Laschi, Mazzolai, Cianchetti, Sci. Robot. 1, eaah3690 (2016) 6 December 2016 5 of 11


SCIENCE ROBOTICS | REVIEW

Fig. 3. Examples of squeezability, resilience, and underwater propulsion. (A) Steps of locomotion under a small crevice by a soft robot based on Pneu-Net (33). (B) Demonstration
of resilience of the robot built from a silicone elastomer (61). (C) Cockroach-inspired robot passing through small crevices (69). (D) Steps of pulsed-jet propulsion with a soft deformable body
(72). [Credits (A to D): multigait soft robot, R. F. Shepherd; resilient-untethered soft robot, M. T. Tolley; vockroach-inspired soft legged robot, K. Jayaram and R. J. Full; PoseiDRONE robot, C. Laschi].

Combined bending and stiffening blocks (Fig. 6A) (78) to entire robots (Fig. 6B) (79). A very recent review
The design of soft manipulators can offer revolutionary approaches paper on soft origami can be found in (80).
beyond the current limitations in medical sectors, such as minimally
invasive surgery. Rigid instruments can exert remarkable forces, but Adaptable grasping
they usually lack dexterity: Clashing often occurs in the control of The compliance of soft-bodied robots and devices has also been
these instruments. A solution to overcome these limits is based on exploited for its high adaptation capabilities. Although grasping has
soft actuators [flexible fluidic actuators (FFAs)], which guarantee been extensively investigated in robotics and effective grippers exist in
high dexterity together with an intrinsic capability of safely interact- manufacturing, prosthetics, and many other fields (81), soft materials
ing with tissues and organs. At the same time, surgical manipulators can be exploited as an excellent interface when grasping tasks have to
also have to guarantee the possibility of varying their stiffness, to be accomplished. If the entire body of the robot is made of soft
generate relatively high forces when needed. Technologies for materials, then intrinsic safety can also be enhanced. This concept
stiffening soft bodies have greatly progressed recently, as reviewed has been used to develop a number of soft grippers based on different
in (32). In general, two approaches are possible: the use of antago- technologies. We highlight two cases exemplifying two different
nistic arrangement of soft actuators (so that a simultaneous activa- approaches. The first, the RBO Hand 2, starts from the five-finger
tion produces an isometric stress without net deformations) and the configuration and uses soft materials to exploit constraints present in
use of semiactive actuators with variable stiffness ability. The STIFF- the environment to obtain robust grasping (82). The RBO Hand 2 is
FLOP (75) manipulator is an example of how to match all these re- built with PneuFlex (silicone pneumatic-based actuators). Because of
quirements (76). The key aspect is to be able to manipulate objects its deformability, the hand can exploit environmental constraints, such
while controlling the stiffness of selected body parts and being inher- as the support of the objects to grasp, and it can robustly grasp objects
ently compliant when interacting with objects. It is based on a of different shapes (83). The second example adopts a completely dif-
modular architecture: It is composed of three identical modules, ferent morphology, that is, a ball-like shape, and uses the granular
and each module is composed of three chambers used as FFA and jamming mechanism (described in a previous section) to build a uni-
one central channel that lodges a granular jamming system for versal gripper that can grasp objects of any shape by adapting its shape
stiffening the arm in the desired position. By activating the corre- to theirs and then changing its stiffness to hold them (84). Tendon-
sponding combination of actuators, omnidirectional bending, elon- driven approaches are very effective for obtaining adaptable grasping, even
gation, and tunable stiffening are thus obtained in the STIFF-FLOP with rigid links (85). The combination of soft structures provides further
manipulator (Fig. 5) (77). abilities and allows the building of wearable devices. A tendon-driven ap-
proach that exploits a jointless silicone structure driven by cables in the
Self-deployment and folding Exo-Glove wearable glove to restore hand motion capabilities of impaired
As mentioned in a previous section, some soft self-deployable structures, people has been proposed (58). Another interesting case in assistance and
inspired by the ancient Japanese technique of paper folding, have been rehabilitation has its basis in the different technological principles of flex-
proposed. Many different devices based on origami structures have been ible fluidic actuation (56). It consists of a completely soft glove based on
demonstrated to change their shape into another predetermined form elastomeric bending actuators aligned with the hand fingers and thus can
through external stimuli (mostly heat application), from basic building assist grasping when activated.

Laschi, Mazzolai, Cianchetti, Sci. Robot. 1, eaah3690 (2016) 6 December 2016 6 of 11


SCIENCE ROBOTICS | REVIEW

Fig. 4. Example of highly extensible sensor and growing capability. (A) Images of the hyperelastic stretchable skin with examples of light emission (73). (B) Image
of a plantoid robot showing robotic roots with bending and growing capabilities. (C) Steps of growth in a plant root–inspired robot. By rotating the deposition head
(DH), the filament (green) passes through the nozzle and is deposited around the outer surface of the tubular body. The mature zone (MZ) of the robotic body remains
strongly anchored to the soil. L indicates the increased length after each deposition step at that specific time t (74). [Credits: highly stretchable electroluminescent skin,
R. F. Shepherd; Plantoid robot, B. Mazzolai].

to literally detach robot body parts and at-


tach them elsewhere, resulting in the ability
to completely reconfigure and/or extend
the robot body. Currently, this has been
demonstrated using HMAs (hot-melting
adhesives) to reversibly connect/disconnect
building blocks to the robot, but in (87),
the possibility of building the entire body
with such a material by combining the
benefit of the mechanical properties of
the material and the shape-changing abil-
ity was also envisaged.

Self-healing
Thermoreversibility is also at the base of a
very new possibility that has been investi-
gated for robots based on soft materials:
self-healing. Self-healing has been intro-
duced as a restoring mechanism for dam-
aged actuators (88, 89), but it could be
theoretically extended to all robot body
Fig. 5. Example of multibending combined with stiffening capabilities. STIFF-FLOP manipulator interacting with parts. The idea has its basis in the molec-
water-filled balloons: passing below and grasping (A), passing in between and grasping (B), and applying a force while
ular mobility of Diels-Alder (DA) poly-
retracting (C) (77). [Credits: STIFF-FLOP manipulator, M. Cianchetti].
mers, which tend to form an elastic
network at room temperature but behave
Morphing as a viscous gel if the temperature exceeds 90°C. When perforated by
Recent soft robotics research has also explored morphological change. sharp objects (thus without implying removal of material), the poly-
The use of a thermoplastic material was demonstrated to enable the pos- mer demonstrates the ability to cure the macroscopic damage
sibility of hardware modifications (morphing) (86). The peculiar (through quite a long thermal treatment) and to completely restore
characteristic of the materials to pass from solid to liquid can be exploited its previous mechanical properties.

Laschi, Mazzolai, Cianchetti, Sci. Robot. 1, eaah3690 (2016) 6 December 2016 7 of 11


SCIENCE ROBOTICS | REVIEW

Biodegradability The abilities of soft-body robots to stretch, squeeze, and morph open
Soft materials often imply the use of polymers. Rarely does the devel- new scenarios of applications for robotics.
opment of robots or devices based on these materials take into con- From different viewpoints, impacts of soft robotics will be large.
sideration environmental implications. However, increasing interest in Robotics is going to change dramatically, and a radically new generation
this subject has pushed some researchers to work on biodegradable of machines is being developed, with better performance in the real
and even edible materials (90). Applied to soft actuators, the materials world and greater adaptability in a variety of tasks (6). The intrinsic
demonstrated good mechanical performance and represented a step multidisciplinary nature of soft robotics favors new and nonconven-
toward environmentally friendly robotics. tional scientific approaches, given that most of the well-known robot-
ics techniques are not applicable. An example is the coupling between
soft robotics and morphological computation (21), which stands as a
CONCLUSIONS new approach to robot design. Supported by a dynamic, engaged, risk-
We have seen how the use of soft materials and variable stiffness tech- taking research community spanning from engineering to materials
nologies in robotics represents an emerging way to build new classes science, biology, mathematics and modeling, medicine, and many oth-
of robotic systems with new abilities that are expected to interact more er disciplines, soft robotics has great potential for promoting scientific
effectively with natural, unstructured environments and with humans. and technological progress. Recent achievements in important
enabling techniques and their theoretical
frameworks are reinforcing this progress.
The current soft robotics literature num-
bers at almost 1000 publications, steadily
increasing from a value close to zero in
2004 (Fig. 7).
Although the advantages of being soft
are many for robots, in terms of abilities,
as outlined in this paper, being complete-
ly soft has important drawbacks, too.
Looking at living organisms, one may ob-
serve that soft animals tend to be small,
whereas larger animals typically need a
skeleton to support their own body
weight (9). Large soft animals without a
skeleton are found in water (for example,
squid and jellyfish) or underground (for
example, giant earthworms), where their
Fig. 6. Origami robots, with the ability of folding and unfolding, changing their shape in response to external bodies are supported by the surrounding
stimuli. (A) Simple cubic building blocks based on SMA that can unfold in 3 s and fold again in 2 s (78); (B) a planar sheet of medium. This biological evidence sug-
paper self-folding into a complex 3D shape thanks to embedded shape memory composites (79). [Credits: self-folding gests a similar approach in soft robotics,
origami robot, K.-J. Cho; self-folding machines, Wyss Institute at Harvard University]. that is, to merge rigid structures with

Fig. 7. Data on soft robotics literature from Scopus in July 2016. Query: (TITLE-ABS-KEY("soft robot*")OR TITLE-ABS-KEY("soft bodied robot*")OR (TITLE-ABS-KEY
("soft material*")AND TITLE-ABS-KEY(robot*))).

Laschi, Mazzolai, Cianchetti, Sci. Robot. 1, eaah3690 (2016) 6 December 2016 8 of 11


SCIENCE ROBOTICS | REVIEW

soft-matter technologies following new design rules for an integrated 7. E. Snell-Rood, Interdisciplinarity: Bring biologists into biomimetics. Nature 529, 277–278
(2016).
approach. This will also enhance the forces that soft robots can exert
8. R. Pfeifer, M. Lungarella, F. Iida, Self-organization, embodiment, and biologically inspired
on the environment, potentially increasing their applications. robotics. Science 318, 1088–1093 (2007).
The challenge to realize autonomous soft robots definitely requires 9. S. Kim, C. Laschi, B. Trimmer, Soft robotics: A bioinspired evolution in robotics. Trends
the integration of different materials and functionalities, such as actua- Biotechnol. 31, 287–294 (2013).
tion, powering, and logic. To achieve the required integration, including 10. A. Albu-Schaffer, M. Fischer, G. Schreiber, F. Schoeppe, G. Hirzinger, Soft robotics: What
Cartesian stiffness can obtain with passively compliant, uncoupled joints?, in
rigid and soft structures, a combination of advanced fabrication techni- Proceedings of the 2004 IEEE/RSJ International Conference on Intelligent Robots and
ques represents a promising solution for the next generation of soft ro- Systems, Sendai, Japan, 28 September to 2 October 2004 [Institute of Electrical &
bots. Among many, such techniques include micromolding (91), soft Electronics Engineers (IEEE), 2004], pp. 3295–3301.
lithography (92), and multimaterial embedded 3D printing (93). 11. A. Albu-Schäffer, O. Eiberger, M. Grebenstein, S. Haddadin, C. Ott, T. Wimböck, S. Wolf,
G. Hirzinger, Soft robotics. IEEE Robot. Automat. Mag. 15, 20–30 (2008).
A greater understanding of the properties of soft materials and how
12. D. Trivedi, C. D. Rahn, W. M. Kier, I. D. Walker, Soft robotics: Biological inspiration, state of
they interact with control systems and the environment is also the art, and future research. Appl. Bionics Biomech. 5, 99–117 (2008).
needed to produce desired robotic behavior in real contexts (94). 13. D. Trivedi, A. Lotfi, C. D. Rahn, Geometrically exact models for soft robotic manipulators.
Among the open challenges of soft robotics, a complete theoretical IEEE Trans. Robot. 24, 773–780 (2008).
framework for designing morphological computation in a soft robot 14. G. Robinson, J. B. C. Davies, Continuum robots—A state of the art, in Proceedings
of the 1999 IEEE International Conference on Robotics and Automation, Detroit, MI, 10
is still missing, but methods have been defined for measuring the to 15 May 1999 (IEEE, 1999), pp. 2849–2854.
role of a soft body in processing information. These methods are 15. C. Laschi, M. Cianchetti, Soft robotics: New perspectives for robot bodyware and control.
important contributions in this sense. The method proposed in Front. Bioeng. Biotechnol. 2 3 (2014).
(41) quantitatively characterizes the spatiotemporal dynamics of a 16. RoboSoft is a Coordination Action on Soft Robotics funded by the European Commission.
17. M. Cianchetti, M. Calisti, L. Margheri, M. Kuba, C. Laschi, Bioinspired locomotion and
soft body through an information-theoretic approach based on
grasping in water: The soft eight-arm OCTOPUS robot. Bioinspir. Biomim. 10, 035003 (2015).
Shannon entropy. 18. R. Baldur, W. Blach, Inflatable Manipulator (Society of Manufacturing Engineers, 1985).
Methods and techniques for modeling soft robots (95) for 19. K. Suzumori, S. Iikura, H. Tanaka, Development of flexible microactuator and its
simulation and control (96) are becoming available, and they represent applications to robotic mechanisms, in Proceedings of IEEE International Conference
an enabling body of knowledge and techniques for further progress of on Robotics and Automation, Sacramento, CA, 9 to 11 April 1991 (IEEE, 1991),
pp. 1622–1627.
soft robotics. The challenges given by the complex nonlinear behaviors 20. R. Pfeifer, J. C. Bongard, How the Body Shapes the Way We Think: A New View of
of soft bodies can be adequately approached with learning-based con- Intelligence (MIT Press, 2007).
trol methods, in which the controllers can learn the robot embodi- 21. C. Laschi, B. Mazzolai, Lessons from nature: The symbiosis of morphological
ment through motor babbling (6). computation and soft robotics. IEEE Robot. Automat. Mag. 23, 107–114 (2016).
22. M. D. Grissom, V. Chitrakaran, D. Dienno, M. Csencsits, M. Pritts, B. Jones, W. McMahan,
Despite the large achievements in this field, soft robotics technol-
D. Dawson, C. D. Rahn, I. D. Walker, Design and experimental testing of the OctArm
ogies can still be improved by learning from the biological system ca- soft robot manipulator. Proc. SPIE 6230, 10.1117/12.665321 (2006).
pabilities of morphological changes and the adaptation of the bodies 23. Y. Bar-Cohen, Electroactive polymers as artificial muscles: Capabilities, potentials and
and functionality during their lifetimes by growing, or even repairing, challenges, in Handbook on Biomimetics, Y. Osada, Ed. (NTS Inc., 2000), pp. 188–196.
damaged body parts. The scientific challenges ahead for soft robotics 24. F. Daerden, D. Lefeber, Pneumatic artificial muscles: Actuators for robotics and
automation. Eur. J. Mech. Environ. Eng. 47, 10–21 (2002).
are further developing these lifelike abilities to change shape, stiffen, 25. B. Trimmer, A. Takesian, B. Sweet, C. B. Rogers, D. C. Hake, D. J. Rogers, Caterpillar
grow, self-heal, develop, and evolve. We envisage abilities for robots to locomotion: A new model for soft-bodied climbing and burrowing robots, in
adapt, not only their behavior but also their morphology, and technol- Proceedings of the 2006 7th International Symposium on Technology and the Mine
ogies and techniques for task-adapted morphologies. Going beyond Problem, Monterey, CA, 2 to 5 May 2006 (Naval Postgraduate School, 2006).
26. M. Cianchetti, Fundamentals on the use of shape memory alloys in soft robotics, in
soft robotics, their bodyware can change shape to adapt to physical
Interdisciplinary Mechatronics: Engineering Science and Research Development,
or task constraints. Going beyond robot control, their behavior will M. K. Habib, J. Paulo Davim, Eds. (Wiley-ISTE, 2013), pp. 227–254.
develop in accordance with their soft body. Going beyond bioinspired 27. S. Kim, E. Hawkes, K. Cho, M. Joldaz, J. Foleyz, R. Wood, Micro artificial muscle fiber using
robotics, their morphology can be designed according to the intended NiTi spring for soft robotics, in Proceedings of the 2009 IEEE/RSJ International Conference
environment. on Intelligent Robots and Systems, St. Louis, MO, 10 to 15 October 2009 (IEEE, 2009)
pp. 2228–2234.
28. S. Seok, C. Onal, R. Wood, D. Rus, S. Kim, Peristaltic locomotion with antagonistic
actuators in soft robotics, in Proceedings of the 2010 IEEE International Conference
SUPPLEMENTARY MATERIALS on Robotics and Automation, Anchorage, AK, 3 to 8 May 2010 (IEEE, 2010),
robotics.sciencemag.org/cgi/content/full/1/1/eaah3690/DC1 pp. 1228–1233.
Movie S1. Soft robots with abilities not found in rigid robots. The compilation uses parts of
29. T. Kaufhold, V. Bohm, K. Zimmermann, Design of a miniaturized locomotion system with
videos from (73, 74, 77).
variable mechanical compliance based on amoeboid movement, in Proceedings of the
IEEE RAS EMBS International Conference on Biomedical Robotics and Biomechatronics, Rome,
Italy, 24 to 27 June 2012 (IEEE, 2012) pp. 1060–1065.
REFERENCES AND NOTES 30. E. Steltz, A. Mozeika, N. Rodenberg, E. Brown, H. M. Jaeger, JSEL: Jamming Skin
1. B. Webb, What does robotics offer animal behavior? Anim. Behav. 60, 545–558 Enabled Locomotion, in Proceedings of the 2009 IEEE/RSJ International Conference on
(2000). Intelligent Robots and Systems, St. Louis, MO, 11 to 15 October 2009 (IEEE, 2009)
2. B. Siciliano, O. Khatib, Eds. in Springer Handbook of Robotics (Springer International pp. 5672–5677.
Publishing, ed. 2, 2016). 31. J. R. Amend, E. Brown, N. Rodenberg, H. M. Jaeger, H. Lipson, A positive pressure
3. L. Wang, F. Iida, Deformation in soft-matter robotics: A categorization and quantitative universal gripper based on the jamming of granular material. IEEE Trans. Robot. 28, 341–350
characterization. IEEE Robot. Automat. Mag. 22, 125–139 (2015). (2012).
4. H. Shen, Meet the soft, cuddly robots of the future. Nature 530, 24–26 (2016). 32. M. Manti, V. Cacucciolo, M. Cianchetti, Stiffening in soft robotics: A review of the state of
5. D. Rus, M. T. Tolley, Design, fabrication and control of soft robots. Nature 521, 467–475 the art. IEEE Robot. Automat. Mag. 23, 93–106 (2016).
(2015). 33. R. F. Shepherd, F. Ilievski, W. Choi, S. A. Morin, A. A. Stokes, A. D. Mazzeo, X. Chen,
6. R. Pfeifer, M. Lungarella, F. Iida, The challenges ahead for bio-inspired ‘soft’ robotics. M. Wang, G. M. Whitesides, Multigait soft robot. Proc. Natl. Acad. Sci. U.S.A. 108,
Commun. ACM 55, 76–87 (2012). 20400–20403 (2011).

Laschi, Mazzolai, Cianchetti, Sci. Robot. 1, eaah3690 (2016) 6 December 2016 9 of 11


SCIENCE ROBOTICS | REVIEW

34. C. Laschi, B. Mazzolai, V. Mattoli, M. Cianchetti, P. Dario, Design of a biomimetic robotic manipulators with precision positioning, in Proceedings of the 2016 IEEE Conference on
octopus arm. Bioinspir. Biomim. 4, 015006 (2009). Biomedical Robotics and Biomechatronics, Singapore, 26 to 29 June 2016 (IEEE, 2016),
35. M. Cianchetti, V. Mattoli, B. Mazzolai, C. Laschi, P. Dario, A new design methodology of pp. 457–463.
electrostrictive actuators for bio-inspired robotics. Sens. Actuators B 142, 288–297 60. C. Jianfei, W. Xingsong, Conception of a tendon-sheath and pneumatic system driven soft
(2009). rescue robot, in Robot Intelligence Technology and Applications 3: Results from the 3rd
36. C. Laschi, M. Cianchetti, B. Mazzolai, L. Margheri, M. Follador, P. Dario, Soft robot arm International Conference on Robot Intelligence Technology and Applications, J.-H. Kim,
inspired by the octopus. Adv. Robotics 26, 709–727 (2012). W. Yang, J. Jo, P. Sincak, H. Myung, Eds. (Springer International Publishing, 2015).
37. K. K. Smith, W. M. Kier, Trunks, tongues, and tentacles: Moving with skeletons of muscle. 61. M. T. Tolley, R. F. Shepherd, B. Mosadegh, K. C. Galloway, M. Wehner, M. Karpelson,
Am. Sci. 77, 28–35 (1989). R. J. Wood, G. M. Whitesides, A resilient, untethered soft robot. Soft Robot. 1, 213–223 (2014).
38. F. Renda, M. Cianchetti, M. Giorelli, A. Arienti, C. Laschi, A 3D steady-state model of a 62. N. W. Bartlett, M. T. Tolley, J. T. B. Overvelde, J. C. Weaver, B. Mosadegh, K. Bertoldi,
tendon-driven continuum soft manipulator inspired by the octopus arm. Bioinspir. G. M. Whitesides, R. J. Wood, A 3D-printed, functionally graded soft robot powered by
Biomim. 7, pp. 5033–5039 (2012). combustion. Science 349, 161–165 (2015).
39. M. Giorelli, F. Renda, G. Ferri, C. Laschi, A feed-forward neural network learning 63. R. F. Shepherd, A. A. Stokes, J. Freake, J. Barber, P. W. Snyder, A. D. Mazzeo,
the inverse kinetics of a soft cable-driven manipulator moving in three-dimensional L. Cademartiri, S. A. Morin, G. M. Whitesides, Using explosions to power a soft robot.
space, in Proceedings on the 2013 IEEE/RSJ International Conference on Intelligent Angew. Chem. Int. Ed. Engl. 52, 2892–2896 (2013).
Robots and Systems, Tokyo, Japan, 3 to 7 November 2013 (IEEE, 2013), pp. 5033–5039. 64. H.-T. Lin, G. G. Leisk, B. Trimmer, GoQBot: A caterpillar-inspired soft-bodied rolling robot.
40. A. Kazakidi, V. Vavourakis, N. Pateromichelakis, J. A. Ekaterinaris, D. P. Tsakiris, Bioinspir. Biomim. 6, 026007 (2011).
Hydrodynamic analysis of octopus-like robotic arms, in Proceedings of the 2012 IEEE 65. S. Seok, C. D. Onal, K.-J. Cho, R. J. Wood, D. Rus, S. Kim, Meshworm: A peristaltic soft
Conference on Robotics and Automation, St. Paul, MN, 14 to 18 May 2012 (IEEE, 2012) robot with antagonistic nickel titanium coil actuators. IEEE/ASME Trans. Mechatron. 18,
pp. 5290–5300. 1485–1497 (2013).
41. K. Nakajima, H. Hauser, R. Kang, E. Guglielmino, D. G. Caldwell, R. Pfeifer, A soft body 66. A. S. Boxerbaum, K. M. Shaw, H. J. Chiel, R. D. Quinn, Continuous wave peristaltic motion
as a reservoir: Case studies in a dynamic model of octopus-inspired soft robotic in a robot. Int. J. Robot. Res. 31, 302–318 (2012).
arm. Front. Comput. Neurosci. 7, 529–5300 (2013). 67. S. Kim, M. Spenko, S. Trujillo, B. Heyneman, D. Santos, M. Cutkosky, Smooth vertical
42. J. Hou, R. H. C. Bonser, G. Jeronimidis, Design of a biomimetic skin for an octopus- surface climbing with directional adhesion. IEEE Trans. Robot. 24, 65–74 (2008).
inspired robot part I: Characterising octopus skin. J. Bionic Eng. 8, 288–296 (2011). 68. M. R. Cutkosky, Climbing with adhesion: From bioinspiration to biounderstanding.
43. F. Tramacere, L. Beccai, E. Sinibaldi, C. Laschi, B. Mazzolai, Adhesion mechanisms Interface Focus 5, 20150015 (2015).
inspired by octopus suckers. Procedia Comput. Sci. 7, 192–193 (2011). 69. K. Jayaram, R. J. Full, Cockroaches traverse crevices, crawl rapidly in confined
44. M. Follador, F. Tramacere, B. Mazzolai, Dielectric elastomer actuators for octopus spaces, and inspire a soft, legged robot. Proc. Natl. Acad. Sci. U.S.A. 113, 640–645 (2016).
inspired suction cups. Bioinspir. Biomim. 9, 046002 (2014). 70. M. Follador, M. Cianchetti, C. Laschi, Development of the functional unit of a completely
45. M. Calisti, M. Giorelli, G. Levy, B. Mazzolai, B. Hochner, C. Laschi, P. Dario, An soft octopus-like robotic arm, in Proceedings of the IEEE RAS/EMBS International
octopus-bioinspired solution to movement and manipulation for soft robots. Bioinspir. Conference on Biomedical Robotics and Biomechatronics, Rome, Italy, 24 to 27 June 2012
Biomim. 6, 036002 (2011). (IEEE, 2012) pp. 640–645.
46. F. Giorgio-Serchi, A. Arienti, C. Laschi, Biomimetic vortex propulsion: Toward the new 71. M. Calisti, F. Corucci, A. Arienti, C. Laschi, Dynamics of underwater legged
paradigm of soft unmanned underwater vehicles. IEEE/ASME Trans. Mechatron. 18, locomotion: Modeling and experiments on an octopus-inspired robot. Bioinspir.
484–493 (2013). Biomim. 10, 046012 (2015).
47. D. B. Camarillo, C. R. Carlson, J. K. Salisbury, Configuration tracking for continuum 72. F. Renda, F. Giorgio-Serchi, F. Boyer, C. Laschi, Modelling cephalopod-inspired pulsed-jet
manipulators with coupled tendon drive. IEEE Trans. Robot. 25, 798–808 (2009). locomotion for underwater soft robots. Bioinspir. Biomim. 10, 055005 (2015).
48. G. S. Chirikjian, J. W. Burdick, A modal approach to hyper-redundant manipulator 73. C. Larson, B. Peele, S. Li, S. Robinson, M. Totaro, L. Beccai, B. Mazzolai, R. Shepherd,
kinematics. IEEE Trans. Robot. Autom. 10, 343–354 (1994). Highly stretchable electroluminescent skin for optical signaling and tactile sensing.
49. D.-H. Kim, J. A. Rogers, Stretchable electronics: Materials strategies and devices. Adv. Science 351, 1071–1074 (2016).
Mater. 20, 4887–4892 (2008). 74. A. Sadeghi, A. Tonazzini, L. Popova, B. Mazzolai, A novel growing device inspired by
50. M. K. Dobrzynski, R. Pericet-Camara, D. Floreano, Contactless deflection sensor plant root soil penetration behaviors. PLOS ONE 9, e90139 (2014).
for soft robots, in Proceedings of the 2011 IEEE/RSJ International Conference on 75. STIFF-FLOP: “STIFFness controllable Flexible and Learn-able Manipulator for surgical
Intelligent Robots and Systems, San Francisco, CA, 25 to 30 September (IEEE, 2011) OPerations.”
pp. 1913–1918. 76. M. Cianchetti, T. Ranzani, G. Gerboni, T. Nanayakkara, K. Althoefer, P. Dasgupta,
51. J. K. Paik, R. K. Kramer, R. J. Wood, Stretchable circuits and sensors for robotic A. Menciassi, Soft robotics technologies to address shortcomings in today’s minimally
origami, in Proceedings of the 2011 IEEE/RSJ International Conference on Intelligent invasive surgery: The STIFF-FLOP approach. Soft Robot. 1, 122–131 (2014).
Robots and Systems, San Francisco, CA, 25 to 30 September 2011 (IEEE, 2011) 77. T. Ranzani, G. Gerboni, M. Cianchetti, A. Menciassi, A bioinspired soft manipulator for
pp. 414–420. minimally invasive surgery, Bioinspir. Biomim. 10, 035008 (2015).
52. S. Voisembert, A. Riwan, N. Mechbal, Numerical evaluation of a new robotic 78. S.-R. Kim, D.-Y. Lee, J.-S. Koh, K.-J. Cho, Fast, compact, and lightweight shape-shifting
manipulator based on inflatable joints, in Proceedings of the 2012 IEEE International system composed of distributed self-folding origami modules, in Proceedings of the IEEE
Conference on Automation Science and Engineering, Seoul, 20 to 24 August 2012 International Conference on Robotics and Automation, Stockholm, Sweden, 16 to 21 May
(Institute of Electrical and Electronics Engineers, 2012), pp. 544–549. 2016 (IEEE, 2016) pp. 4969–4974.
53. J. Hiller, H. Lipson, Automatic design and manufacture of soft robots. IEEE Trans Robot. 79. S. Felton, M. Tolley, E. Demaine, D. Rus, R. J. Wood, A method for building self-folding
28, 457–466 (2012). machines. Science 345, 644–646 (2014).
54. F. Corucci, N. Cheney, H. Lipson, C. Laschi, J. Bongard, Material properties affect 80. C. M. Wheeler, M. L. Culpepper, Soft origami: Classification, constraint, and actuation of
evolution’s ability to exploit morphological computation in growing soft-bodied highly compliant origami structures. J. Mech. Robot. 8, 051012 (2016).
creatures, in Proceedings of the 15th International Conference on the Synthesis and 81. C. Melchiorri, M. Kaneko, Robot hands, in Springer Handbook of Robotics (Springer-
Simulation of Living Systems–ALIFE XV, Cancun, Mexico, 4 to 8 July 2016 (MIT Press, Verlag, 2008), pp 345–360.
2016). 82. C. Eppner, R. Deimel, J. Álvarez-Ruiz, M. Maertens, O. Brock, Exploitation of
55. J. Burgner-Kahrs, D. C. Rucker, H. Choset, Continuum robots for medical applications: A environmental constraints in human and robotic grasping. Int. J. Robot. Res. 34,
survey. IEEE Trans. Robot. 31, 1261–1280 (2015). 1021–1038 (2015).
56. P. Polygerinos, Z. Wang, K. C. Galloway, R. J. Wood, C. J. Walsh, Soft robotic glove for combined 83. R. Deimel, O. Brock, A novel type of compliant and underactuated robotic hand for
assistance and at-home rehabilitation. Robot. Auton. Syst. 73, 135–143 (2015). dexterous grasping. Int. J. Robot. Res. 35, 161–185 (2016).
57. E. C. Goldfield, Y.-L. Park, B.-R. Chen, W.-H. Hsu, D. Young, M. Wehner, 84. J. Amend Jr., E. Brown, N. Rodenberg, H. Jaeger, H. M. Lipson, A positive pressure
D. G. Kelty-Stephen, L. Stirling, M. Weinberg, D. Newman, R. Nagpal, E. Saltzman, universal gripper based on the jamming of granular material. IEEE Trans. Robot. 28,
K. G. Holt, C. Walsh, R. J. Wood, Bio-inspired design of soft robotic assistive devices: The 341–350 (2012).
interface of physics, biology, and behavior. Ecol. Psychol. 24, 300–327 (2012). 85. C. Cipriani, M. Controzzi, M. C. Carrozza, Objectives, criteria and methods for the design
58. H. In, B. B. Kang, M. Sin, K. J. Cho, Exo-glove: A wearable robot for the hand with a soft of the SmartHand transradial prosthesis. Robotica 28, 919–927 (2010).
tendon routing system. IEEE Robot. Automat. Mag. 22, 97–105 (2015). 86. L. Wang, F. Iida, Towards “soft” self-reconfigurable robots, in Proceedings of the IEEE
59. Y. Ansari, E. Falotico, Y. Mollard, B. Busch, M. Cianchetti, C. Laschi, A multi-agent RAS/EMBS International Conference on Biomedical Robotics and Biomechatronics,
reinforcement learning approach for inverse kinematics of high dimensional Rome, Italy, 24 to 27 June 2012 (IEEE, 2012), pp. 593–598.

Laschi, Mazzolai, Cianchetti, Sci. Robot. 1, eaah3690 (2016) 6 December 2016 10 of 11


SCIENCE ROBOTICS | REVIEW

87. L. Brodbeck, F. Iida, An extendible reconfigurable robot based on hot melt adhesives. International Conference on Robotics and Automation, 2011 IEEE International,
Auton. Robot. 39, 87–100 (2015). Shanghai, China, 9 to 13 May 2011 (IEEE, 2011) pp. 1243–1249.
88. S. Terryn, G. Mathijssen, J. Brancart, D. Lefeber, G. V. Assche, B. Vanderborght, 99. www.softroboticsinc.com/
Development of a self-healing soft pneumatic actuator: A first concept. Bioinspir. 100. A. D. Marchese, C. D. Onal, D. Rus, Autonomous soft robotic fish capable of escape
Biomim. 10, 046007 (2015). maneuvers using fluidic elastomer actuators, Soft Robot. 1, 75–87 (2014).
89. R. F. Shepherd, A. A. Stokes, R. M. D. Nunes, G. M. Whitesides, Soft machines that are 101. A. Arienti, M. Calisti, F. Giorgio Serchi, C. Laschi, PoseiDRONE: Design of a soft-bodied
resistant to puncture and that self seal. Adv. Mater. 25, 6709–6713 (2013). ROV with crawling, swimming and manipulation ability, in Proceedings of the MTS/IEEE
90. L. D. Chambers, J. Winfield, I. Ieropoulos, J. Rossiter, Biodegradable and edible gelatine OCEANS Conference, San Diego, CA, 21 to 27 September 2013 (IEEE, 2013) pp. 1–7.
actuators for use as artificial muscles. Proc. SPIE 9056, 90560B (2014). 102. J. K. Paik, R. K. Kramer, R. J. Wood, Stretchable circuits and sensors for robotic origami,
91. M. Heckele, W. K. J. Schomburg, Review on micro molding of thermoplastic polymers. in Proceedings of the 2011 IEEE/RSJ International Conference on Intelligent Robots and
J. Micromech. Microeng. 14, R1–R14 (2004). Systems, San Francisco, CA, 25 to 30 September 2011 (IEEE, 2011), pp. 414–420.
92. D. Qin, Y. Xia, G. M. Whitesides, Soft lithography for micro- and nanoscale patterning. 103. T. Umedachi, V. Vikas, B. A. Trimmer, Softworms: The design and control of non-
Nat. Protoc. 5, 491–502 (2010). pneumatic, 3D-printed, deformable robots. Bioinspir. Biomim. 11, 025001 (2016).
93. W. Wu, A. DeConinck, J. A. Lewis, Omnidirectional printing of 3D microvascular 104. M. Wehner, R. L. Truby, D. J. Fitzgerald, B. Mosadegh, G. M. Whitesides, J. A. Lewis,
networks. Adv. Mater. 23, H178–H183 (2011). R. J. Wood, An integrated design and fabrication strategy for entirely soft, autonomous
94. C. Paul, Morphological computation: A basis for the analysis of morphology and control robots. Nature 536, 451–455 (2016).
requirements. Robot. Auton. Syst. 54, 619–630 (2006).
95. F. Renda, M. Giorelli, M. Calisti, M. Cianchetti, C. Laschi, Dynamic model of a Acknowledgments
multibending soft robot arm driven by cables, IEEE Trans. Robot. 30, 1109–1122 Funding: This work was supported by the European Commission through the RoboSoft
(2014). Coordination Action (grant number 619319) and the I-Support project (grant number 643666).
96. F. Largilliere, V. Verona, E. Coevoet, M. Sanz-Lopez, J. Dequidt, C. Duriez, Real-time
control of soft-robots using asynchronous finite element modeling, in Proceedings of Submitted 4 August 2016
the IEEE International Conference on Robotics and Automation, Stockholm, Sweden, Accepted 4 October 2016
16 to 21 May 2016 (IEEE, 2016) pp. 2550–2555. Published 6 December 2016
97. S. Kim, J. E. Clark, M. R. Cutkosky, iSprawl: Design and tuning for high-speed 10.1126/scirobotics.aah3690
autonomous open-loop running. Int. J. Robot. Res. 25, 903–912 (2006).
98. K. C. Galloway, J. E. Clark, M. Yim, D. E. Koditschek, Experimental investigations into Citation: C. Laschi, B. Mazzolai, M. Cianchetti, Soft robotics: Technologies and systems pushing
the role of passive variable compliant legs for dynamic robotic locomotion, the boundaries of robot abilities. Sci. Robot. 1, eaah3690 (2016).

Laschi, Mazzolai, Cianchetti, Sci. Robot. 1, eaah3690 (2016) 6 December 2016 11 of 11

También podría gustarte