A review on robotic fish enabled by ionic polymer–metal composite artificial muscles
Chen Robot. Biomim.
A review on robotic fish enabled by ionic polymer-metal composite artificial muscles
Zheng Chen 0
0 Department of Mechanical Engineering, University of Houston , 4726 Calhoun Road, Room N207, Houston, TX 77204-4006 , USA
A novel actuating material, which is lightweight, soft, and capable of generating large flapping motion under electrical stimuli, is highly desirable to build energy-efficient and maneuverable bio-inspired underwater robots. Ionic polymer-metal composites are important category of electroactive polymers, since they can generate large bending motions under low actuation voltages. IPMCs are ideal artificial muscles for small-scale and bio-inspired robots. This paper takes a system perspective to review the recent work on IPMC-enabled underwater robots, from modeling, fabrication, and bio-inspired design perspectives. First, a physics-based and control-oriented model of IPMC actuator will be reviewed. Second, a bio-inspired robotic fish propelled by IPMC caudal fin will be presented and a steady-state speed model of the fish will be demonstrated. Third, a novel fabrication process for 3D actuating membrane will be introduced and a bio-inspired robotic manta ray propelled by two IPMC pectoral fins will be demonstrated. Fourth, a 2D maneuverable robotic fish propelled by multiple IPMC fin will be presented. Last, advantages and challenges of using IPMC artificial muscles in bio-inspired robots will be concluded.
Ionic polymer-metal composite; Bio-inspired robotic fish; Dynamic modeling; Fabrication
Species invasions, such as Asian carps invasion recently
found in the Illinois River, have caused ecological
problems for local species [
]. To control the quantity of
invasive species, habitat study plays an important role in
figuring out an ecological effective way [
]. To enable the
study, autonomous, stealthy, and highly maneuverable
underwater vehicles are highly desirable in monitoring
of the invasive species. Traditional underwater
vehicles, such as submarines, are driven by electric motors,
which rely on a rotated propeller to generate propulsion.
Rotation-based propulsion creates unfavorable acoustic
noise, which draws attentions from underwater creatures
and thus leads to unfaithful data for their habitat study.
More stealthy and environmentally friendly propulsive
approaches need to be investigated and adopted for the
underwater vehicles in such applications.
After thousand years of evolution, underwater
creatures, such as fish and rays, are extremely best swimmers
which man-made underwater vehicles cannot compete
with. In order to mimic the swimming behavior of
biological fish, much effort has been spent on how
propulsion is generated by the fish locomotion. For example,
] studied large-amplitude elongated-body
theory of fish locomotion. Lauder studied kinematics
and dynamics of fish fin [
]. Through those studies, it was
found that most of underwater creatures adopt
flappingbased propulsion for fast and energy-efficient moving
and highly maneuvering through water. Flapping-based
propulsion systems have been studied for many years
]. However, in most of cases, the propulsion
systems for robotic fish are still driven by electrical motors,
which need a power transmission to convert rotation to
flapping. Most of power transmission systems are bulky,
energy inefficient, and noisy, which are unsuitable for
small-size and bio-inspired robots. To avoid using power
transmission, a novel actuating material that can
naturally generate flapping is greatly needed. It will enable us
to design bio-inspired and stealthy robots for ecological
underwater monitoring applications.
Electroactive polymers are emerging actuating
materials that can generate large deformation under electrical
]. EAPs win their nickname, artificial
muscles, due to their similarities to the biological muscles in
terms of achievable stress and strain. EAPs have
different configurations and basically they can be divided into
two categories: ionic EAPs and dielectric EAPs.
Dielectric EAPs are driven by the electrostatic force applied to
dielectric polymers, which can generate large contraction
]. Dielectric EAPs require high actuation voltage
(typically higher than 1 kV), which limits their applications
in underwater bio-inspired robots. Ionic EAPs are driven
by the ionic transportation-induced swelling effect, which
typically only needs small actuation voltage (1 or 2 V) and
can naturally generate bending motion. Ionic polymer–
metal composites (IPMCs) are an important category of
ionic EAPs due to their chemical stability under wet
condition and built-in actuation and sensing capability [
An IPMC has a sandwiched structure that consists
of an ion exchange membrane coated with two noble
metal electrodes, such as gold or platinum, on its surface
(Fig. 1) [
]. Application of a small voltage (less than 2 V)
to the IPMC creates an electric field that drives the
cations (positive ions) to transport to the cathode side while
anions (negative ions) are fixed on the carbon polymer
]. The unbalanced cation density distribution
along the thickness direction introduces a swelling effect
on the cathode side and a shrinking effect on the anode
side. Eventually, the IPMC bends to the anode side and
thus leads to an actuation effect. Due to their naturally
flapping capabilities under wet condition, IPMCs are an
ideal engineering actuation material for small-scale and
bio-inspired underwater robots.
To achieve a desired actuation performance of IPMC
for underwater applications, many researchers have been
working on modeling and control of IPMCs. Chen et al.
] developed a physics-based and control-oriented
model for IPMC and then validated the model through
designing and implementing a model-based H-infinity
control. To accommodate the large system’s
uncertainties, such as hydration level, many researchers have
developed adaptive robust controls for IPMCs. For example,
Anh et al. [
] developed a robust control using
quantitative feedback technique (QFT) which identified the
system’s characteristics using a pseudorandom binary
signal (PRBS) and then a QFT controller was designed and
implemented online based on the identified model. Kang
et al. [
] developed H-infinity controls with and without
loop shaping or μ-synthesis. Their results showed that the
robust control techniques can significantly improve the
IPMC performance against non-repeatability or
parametric uncertainties in terms of the faster response and lower
overshoot than the PID control, using lower actuation
voltage. Moreover, Chen et al. [
] developed an adaptive
control for IPMCs to compensate the hysteresis in IPMC.
To avoid using bulky external sensors, many
researchers have been focusing on developing a compact sensing
scheme for IPMC. For example, Chen et al. developed an
IPMC/PVDF sensory actuator and implemented a
feedback control using integrated sensing feedback. Leang
et al. [
] developed an integrated sensing scheme for
IPMCs using strain gauges and then developed a tracking
control of an IPMC in an underwater environment.
IPMC-enabled underwater robots have been
investigated by many researchers. Guo et al. [
ionic conductive polymer film (ICPF)-enabled robotic
fish which can achieve 0.137 body length per second
(BL/s) swimming speed. Laurent et al. [
] studied the
efficiency of microrobot propelled by IPMC, which can
achieve about 1.4% efficiency. Then researchers
developed different types of underwater robots, such as
robotic fish [
], robotic ray [
], robotic jellyfish [
and robotic worm [
], for various applications. In this
paper, a systems perspective will be taken to review
the recent work on IPMC-enabled bio-inspired
underwater robots, including (1) a physics-based and
control-oriented modeling approach that can capture the
intrinsic actuation dynamics of IPMC and the
hydrodynamics of robotic fish; (2) a fabrication technology for
creating IPMC actuating membranes capable of
generating 3D kinematic motions; and (3) bio-inspired design of
robotic fish and ray. Finally, discussions and conclusions
will be presented at the end.
Physics‑based control‑oriented modeling
of robotic fish propelled by IPMC caudal fin
Although control of robotic fish powered by electrical
motors has been well developed by many research groups
], control of the robotic fish enabled by IPMC has
rarely been studied based on our best knowledge. The
possible reason might be lacking of a faithful and
practical dynamic model of the robotic fish enabled by IPMC.
Due to the complex actuation dynamics of IPMC and
hydrodynamics of fish, it was a great challenge to get a
physics-based control-oriented model. Two types of
models have been developed, including a steady-state
speed model [
] developed by Tan’s group at Michigan
State University and a dynamic model [
] developed by
Porfiri’s group at New York University. In this section, we
will review the steady-state speed model [
by Tan’s group. In this review, we will discuss Lighthill’s
theory on elongated-body propulsion first. Then the
IPMC beam dynamics in fluid will be discussed next,
considering general force and moment inputs. This will
be followed by the actuation model of IPMC caudal fin.
Last, the model for computing the speed of
IPMC-propelled robotic fish will be obtained by merging Lighthill’s
theory and the hybrid tail dynamics. Most of the
modeling work presented in this section was published in [
A body is considered elongated if its cross-sectional area
changes slowly along its length. Suppose that the tail is
bending periodically with the bending displacement at
z denoted by w(z, t). See Fig. 2 for notation [
]. At the
steady state, the fish will achieve a periodic, forward
motion with some mean speed U. In the discussion here,
the word “mean” refers to the average over one period.
The mean thrust T produced by the tail can be calculated
∂w(z, t) 2
− U 2 ·
∂w(z, t) 2
where z = L1 denotes the end of tail, denotes the
mean value, and m is the virtual mass density at z = L1,
balanced by the drag FD, from which one can solve the
cruising speed U as
CDρwS + m ·
Since the speed of the fish is related to the lateral
velocity and the slope of the trailing edge, one needs to fully
understand the actuation dynamics of the tail.
Model of IPMC hybrid tail
The model combines the seemingly incompatible
advantages of both the white-box models (capturing key
physics) and the black-box models (amenable to control
design). The proposed modeling approach provides an
interpretation of the sophisticated physical processes
involved in IPMC actuation from a systems perspective.
The model development starts from the governing PDE
] that describes the charge redistribution dynamics
under external electrical field, electrostatic interactions,
ionic diffusion, and ionic migration along the thickness
direction. The model incorporates the effect of
distributed surface resistance, which is known to influence the
actuation behavior of IPMCs . Moreover, by
converting the original PDE into the Laplace domain, an exact
solution is obtained, leading to a compact, analytical
model in the form of infinite-dimensional transfer
function. The model can be further reduced to low-order
models, which again carry physical interpretations and
are geometrically scalable.
m = π4 Sc2ρwβ,
where Sc is the width of the tail at the end z = L1, ρw is the
fluid density, and β is a non-dimensional parameter close
to 1. Equation (1) indicates that the mean thrust depends
only on the lateral velocity (∂w/∂t) and the slope (∂w/∂z)
at the tail end. A cruising fish, under inviscid flow
conditions, will experience a drag force FD as
where S is the wetted surface area and CD is the drag
coefficient. At the steady state, the mean thrust T is
MIPMC(z, s) =
α0WKke(γ (s) − tanh (γ (s)))V (s) cosh
(sγ (s) + K tanh (γ (s))) ·
Moment generated by IPMC
Geometric definitions of IPMC cantilever beam are shown
in Fig. 3. Let D, E, φ, and ρ denote the electric
displacement, the electric field, the electric potential, and the
charge density, respectively. The following equations hold:
∇ · D = ρ = F (C+ − C−),
= −∇φ = κe ,
where κe is the effective dielectric constant of the polymer,
F is Faraday’s constant, and C+ and C− are the cation and
anion concentrations, respectively.The ion transportation
can be captured by a second-order linear PDE in terms of
charge density [
∂t − d ∂x2 + κeRT
1 − C− V ρ = 0,
Nemat-Nasser and Li assumed that the induced stress is
proportional to the charge density [
where α0 is the coupling constant. To ease the equation,
σ = α0ρ,
∂3φ K ∂φ
∂x3 − d ∂x
1 − C− V .
Farinholt investigated the current response of a
cantilevered IPMC beam when the base is subject to step and
harmonic actuation voltages [
]. A key assumption is that
the ion flux at the polymer/metal interface is zero, which
serves as a boundary condition for (7), and leads to
With distributed surface resistance, we can relate the
actuation-induced bending moment MIPMC(z, s) at point z to
the actuation voltage V(s) by an infinite-dimensional
transfer function [
√B(s)z − sinh
1 + r2θ (s)
B(s) = r1
γ (s) =
s + K ,
θ (s) 2
(1 + r2θ (s)) + Rp
1 − C− V ,
Wkesγ (s)(s + K )
θ (s) = h(sγ (s) + K tanh (γ (s))) ,
Beam dynamics in fluid
In order to obtain the full actuation model of IPMC,
Chen et al. started with a fourth-order PDE for the
dynamic deflection function w(z, t) [
= f (z, t), (12)
where Y, I, C, ρm, and A denote the effective Young’s
modulus, the area moment of inertia, the internal
damping ratio, the density, and the cross-sectional area of the
IPMC beam, respectively, and f(z, t) is the distributed
force density acting on the beam.
The force on the beam consists of two components, the
hydrodynamic force Fhydro from water and the driving
force Fdrive due to the actuation of IPMC
F (z, s) = Fhydro(z, s) + Fdrive(z, s).
Assuming that amplitude of flapping is small, the
hydrodynamic force acting on the IPMC beam can be
expressed as [
Fhydro(z, s) = −ρw
W 2s2Γ1(ω)w(z, s),
0 ≤ z ≤ L, (14)
Mode summation method to solve beam equation
Mode summation method is used to solve the beam
dynamics equation. According to the mode analysis
method, we can express the solution to (12) as the sum of
different modes [
and the natural frequency ωi and the damping ratio ξi for
the ith mode are
w(z, s) =
where φi(z) is the beam shape for the ith mode and qi(s)
is the corresponding generalized coordinate. The mode
shape φi(z) takes the form
ϕi(z) = cosh ( iz) − cos ( iz) − βi(sinh ( iz) − sin ( iz)),
where λi can be obtained by solving
1 + cos ( iL) cosh ( iL) = 0,
sinh ( iL) − sin ( iL)
βi = cosh ( iL) + cos ( iL)
qi(s) = fi(s)Qi(s),
where fi(s) is the generalized force
Qi(s) = s2 + 2ξiωis + ωi2
The generalized coordinate qi(s) can be represented as
where W is the width of the IPMC beam, Γ1(ω) is the
hydrodynamic function for the IPMC beam subject to an
oscillation with radial frequency ω, and ρw is the density
of fluid. The hydrodynamic function for a rectangular
beam can be represented as [
Γ1(ω) = Ω( Re) 1 + √iReK0 −i√iRe
where the Reynolds number Re of a vibrated beam in
water is given by
K0 and K1 are modified Bessel functions of the third type,
Ω(Re) is the correction function associated with the
rectangular beam cross section [
], and η is the viscosity of
Hydrodynamic force on passive fin
The hydrodynamic force acting on the passive fin can be
written as [
ftail(z, s) = − 4 ρws2b(z)2Γ2(ω)w(z, s),
L0 ≤ z ≤ L1,
where Γ2(ω) is the hydrodynamic function of the passive
fin. Note that the hydrodynamic force acting on the active
IPMC beam has been incorporated in Eq. (12), and
therefore, only the hydrodynamic force on the passive fin needs
to be considered here. Since the passive fin used is very light,
its inertial mass is negligible compared to the propelled
virtual fluid mass and is thus ignored in the analysis here.
Considering that the passive fin is rigid compared to IPMC, its
width b(z) and deflection w(z, s) can be expressed as
b1 − b0 (z − L0) + b0,
b(z) = L1 − L0
w(z, s) = w(L0, s) +
(z − L0),
where b0, b1, L, L0, and L1 are defined in Fig. 2. Then, one
can calculate the moment introduced by the passive fin:
for L0 ≤ z ≤ L1.
If we define
ξi = 2µ v(ωi)ωi
and Ci = λiL. Noting that Γ1 (ω) is almost a constant
value in the frequency region around ωi, one can consider
μv(ωi) as a constant in (18). Therefore, ωi can be obtained
approximately. Then, with ωi, ξi can be obtained from
The moment MIPMC(z, s) can be replaced by actuation
by three components: a distributed force density Fd(z, s)
acting along the length, a concentrated force Fc(L, s), and
a moment M(L, s) acting at the IPMC tip z = L, where
Fc(L, s) = −
Fd(z, s) =
M(L, s) = MIPMC(L, s).
where Mtail an Ftail are defined in (19) and (20),
respectively, Fd(z, s) and M(L0, s) are defined in (31) and (32),
Then with the generalized force (33), one can solve the
beam equation using the mode summation method (22).
Finally, the transfer functions relating w(L0, s) to V(s) and
that relating to w′(L0, s)(s) = ∂w(z, s)/∂z|z=L0 to V(s) can
be found as
H2(L0, s) = (1 +(1 C+s)F(s1)A+s F−s)B−sEBssJs ,
H2d (L0, s) = (1 +(1C+s)C( 1s)+EsF−s)A−sJBssJs ,
where As, Bs, Cs, Fs, Js, and Es are transfer functions
related to the dimensions of the caudal fin. See [
the detailed derivation. From (18), (34), and (35), one can
obtain the transfer functions relating the bending
displacement and the slope at z = L1 to the voltage input
V(s) as follows:
H3(L1, s) =
H3d (L1, s) =
= H2(L0, s) + H2d (L0, s)D, (36)
= H2d (L0, s).
Fig. 4 Forces and moments acting on the hybrid tail [
Speed model of robotic fish
Given a voltage input V(t) = Amsin(ωt) to the IPMC
actuator, the bending displacement and the slope of the tail at
the tip z = L1 can be written as
w(L, t) = Am H (jω) sin(ωt + ∠H (jω)),
|z=L = Am Hd (jω) sin(ωt + ∠Hd (jω)), (39)
where ∠(·) denotes the phase angle, and H(s) and Hd(s)
represent H3(L1,s) and H3d (L1, s), respectively. From
(4), one can then obtain the steady-state speed U of the
robotic fish under the square wave actuation voltage as
m · π2
CDρwS + m · π2
n∞=1,3,5,... H jnω
n∞=1,3,5,... |Hd(njn2ω)|2 . (40)
Fabrication of IPMC actuating membrane capable
of 3D deformation
3D kinematic motions have been observed from many
types of biological fins, including pectoral fin and caudal
]. To mimic the swimming behavior of fish, flapping
only motion is not sufficient enough to generate high
efficient propulsion and high maneuverability. Since IPMC
can only generate bending motion, in this section, we
present two different fabrication technologies that
enable us to fabricate IPMC actuation membrane capable
of generating 3D deformation. Comparison of these two
approaches will be given based on the characterization
results. Most of the work presented in this section was
published in [
Lithography‑based fabrication process
The first fabrication process is lithography-based,
monolithic fabrication process for creating multiple IPMC
regions that are mechanically coupled through
compliant, passive membrane. Both the IPMCs and the
passive regions are to be formed from a same Nafion film.
There are two major challenges in fabricating such
actuators. First, the passive areas can substantially
constrain the motion of the active areas. An effective,
precise approach is needed for tailoring the stiffness of the
passive areas. Second, Nafion films are highly swellable
in a solvent. Large volume change results in poor
adhesion of photoresist to Nafion and creates problems in
photolithography and other fabrication steps. To
overcome these challenges, two novel fabrication techniques
have been introduced: (1) selectively thinning down
Nafion with plasma etch, to make the passive areas thin
and compliant; (2) impregnating Nafion film with
platinum ions, which significantly reduces the film
swellability and allows subsequent lithography and other steps.
Fabrication of an artificial pectoral fin is taken as an
example. As illustrated in Fig. 5, the major process steps
1. Create an aluminum mask on Nafion with e-beam
deposition, which covers the intended IPMC regions.
2. Etch Nafion with argon and oxygen plasmas to thin
down the passive regions.
3. Remove the aluminum mask and place the sample in
platinum salt solution to perform ion exchange. This
will stiffen the sample and make the following steps
4. Pattern with photoresist (PR), where the targeted
IPMC regions are exposed while the passive regions
5. Perform the second ion exchange and reduction to
form platinum electrodes in active regions. To
further improve the conductivity of the electrodes,
100nm gold is sputtered on the sample surface.
6. Remove PR and lift off the gold on the passive areas.
Soften the passive regions with HCl treatment (to
undo the effect of step 3).
7. Cut the sample into a desired shape.
Based on the above process, a pectoral fin has been
fabricated, which is shown in Fig. 6. The fin was able to
generate complex deformations, including bending, twisting,
and cupping, by controlling the phase angle among the
signals applied to the active areas. The fin was
characterized in terms of twisting angle and deflection. The fin was
able to achieve 15 degree peak-to-peak twisting angle
with about 2-mm bending displacement.
Assembly‑based PDMS bonding process
The second fabrication process is assembly-based PDMS
molding fabrication process to create an IPMC-based
actuating membrane, capable of complex 3D
deformations. The first step in the process is to synthesize the
IPMC actuator. Many groups have developed different
IPMC fabrication processes to accommodate various
]. In [
], Chen et al. followed most of
the fabrication procedure outlined in [
] but add a
multiple platinum plating process that reduces the surface
resistance of the electrodes to improve the actuation
The assembly process to produce an integrated IPMC/
PDMS actuating membrane is shown in Fig. 7. A mold is
fabricated from Delrin using a CNC rapid mill machine
(MDX-650, Roland). The mold is designed to house the
IPMC beams (280 μm thick), which is then surrounded
with uncured PDMS gel (Ecoflex 0030, Smooth-on Inc.).
The mold, containing the IPMCs and uncured PDMS,
Fig. 5 Lithography-based fabrication process [
]. a Deposit
aluminum mask on both sides of Nafion film. b Thin down passive
area with plasma etch. c Remove aluminum mask and perform
ionexchange to make Nafion stiffer. d Deposit PR and then pattern PR
through lithography. e Perform another ion-exchange and electroless
palting of platinum to create IPMC electrodes. f Remove PR and
perform final treatment and g Cut the patterned IPMC into a fin shape
is clamped and the PDMS is allowed to cure at room
temperature for 3 h. The mold is removed leaving the
IPMC/PDMS membrane actuator (Fig. 8). The PDMS
has a final thickness of 190 μm that is measured using
a caliper (CD-S6”CT, Mitutoyo). The characterization of
the actuating membrane has shown that the maximum
twist angle can reach up to 15°, the flapping deflection
can reach up to 25% of span-wise length, the tip force
can reach up to 0.5 g force, and the power consumption
is below 0.5 W.
Bio‑inspired design of underwater robots
Three types of bio-inspired underwater robots have been
developed in this research, including robotic fish
propelled by a caudal fin, robotic manta ray propelled by
two pectoral fins, and robotic fish propelled by
multiple IPMC fins. The robotic fish was fabricated to verify
the speed model described in “Physics-based
controloriented modeling of robotic fish propelled by IPMC
caudal fin” section while the manta ray was built to
validate the fabrication process for pectoral fin described in
“Bio-inspired design of underwater robots” section. The
robotic fish propelled by multiple IPMC fins was
developed to validate both forward swimming and turning
capabilities. Most of the work presented in this section
was published in [
33, 42, 48, 49
Robotic fish propelled by caudal fin
A robotic fish propelled by an IPMC caudal fin was
developed in Tan’s group at Michigan State University, shown
in Fig. 6 [
]. Inspired by biological fish fins, where
passive, collagenous membranes are driven by
muscle-controlled fin rays [
], a passive, plastic fin was attached to
the tip of IPMC to enhance propulsion. It consists of a
rigid body and an IPMC caudal fin (Fig. 9).
The speed model has been validated through
experiments. Four different types of caudal fins with different
dimensions have been used in the robotic fish to verify
the geometrical scalability of the speed model. A series
of square wave signals with the frequency ranging from
0.2 to 2 Hz were applied to the tail to propel the fish.
Figure 10 shows one set of data for the robotic fish with tail
A. It shows that the experimental data can be captured by
model prediction well. There was an optimal frequency
at which the fish swam at its fast speed (0.02 m/s), which
was 0.125 BL/s [
Robotic manta propelled by pectoral fin
Based on the assembly-based fabrication process, Chen
et al. [
] developed a robotic manta ray using two
pectoral fins. Two acrylic frames with gold electrodes were
made to clamp the artificial wings to the body support.
Gold electrodes were used to minimize corrosion. A
polymer foam was put into the middle of the frame to make
the robot slightly positively buoyant. The fully assembled
robot is 8 cm long (not including the length of the tail),
18 cm wide, 2.5 cm high, and weights 55.3 g. The
freeswimming robot is shown in Fig. 11. The total cost of the
robot is about $200.
The robot was tested in a water tank (1.5 m wide, 4.7 m
long, and 0.9 m deep). As the first attempt, the
operating frequency of the square wave actuation voltage is
tuned at 0.4 Hz and the amplitude is set at 3.3 V. A digital
video camera (VIXIA HG21, Canon) is used to capture
the movies of the swimming robot. Figure 12 shows six
snapshots of the swimming robot from top view. Each
snapshot is taken every 5 s. One can extract the speed
of the robot from the movie through the Edge Detection
program in the Labview. The swimming speed shown in
Fig. 12 is 0.42 cm/s. Since the body length is 8 cm, one
can calculate the speed is 0.053 BL/s. This was believed
to be the first demonstration of an IPMC-propelled
freeswimming robotic batoid ray. It also validated the
proposed fabrication process for making IPMC actuator
capable of 3D kinematic motions.
To simplify the control strategy, Chen et al. [
developed a robotic manta ray with two pectoral fins where
only one IPMC was placing at the leading edge. The fully
assembled robot was 11 cm long, 21 cm wide, and 2.5 cm
high with a mass of 55 grams. The free-swimming robot
with the control unit is shown in Fig. 13.
Figure 14 presents six snapshots of the swimming robot
from top view. Each snapshot was taken every 5 s. A
swim speed of 0.74 cm/s was calculated from the movie
using the Edge Detection program in the Labview. Since
the body length was 11 cm, the speed in body length per
second (BL/s) was 0.067.
Robotic fish propelled by multiple IPMC fins
In the experiments of two robotic manta ray developed
by Chen et al, it has been observed that if two pectoral
fins were controlled differently, the turning performance
of robotic manta ray is better than that of the robotic fish
propelled by a caudal fin only. It would be good to
utilize two pectoral fins for maneuvering and one caudal
fin for main propulsion. Followed by this idea, Ye et al.
] developed a robotic fish propelled by multiple IPMC
fins. Figure 15 shows the assembled robotic fish, which
was 18 cm long and 8 cm wide. The total weight of the
robot was 290 g. Overall, the fish had slightly positive
The fish’s forward swimming speed was controlled
by changing the flapping frequency of the caudal fin. A
square wave signal with 7.3 V magnitude and 0.55 Hz
frequency was applied to the caudal fin. The pectoral
fins were also actuated. The forward swimming speed
reached about 0.067 BL/s. Also, there was a threshold
whereby the frequency was neither too high nor too low
for the fish to swim. Figure 16 shows the snapshots of a
forward swimming test.
Turning tests were conducted to verify the steering
capability of the pectoral fins. To make a left turn, the left
pectoral fin was actuated with the same actuation signal
applied to the caudal fin, while the right pectoral fin was
kept inactive. The caudal fin provided the forward
swimming direction, while the force generated by the left
pectoral fin made the fish tail turn to the left.
To make a right turn, the right pectoral fin was
actuated with the same actuation signal applied to the caudal
fin while the left pectoral fin was kept inactive. Actuation
of the right pectoral fin made the fish turn to the right.
Figure 17 shows the snapshots of a left-turning
swimming test. The robot reached up to 2.5 degree/s turning
Discussion and conclusion
The physics-based and control-oriented model of the
robotic fish will offer a great help in optimal design of
the caudal fin and real-time control of the fish. It
incorporates the hydrodynamics of the fish and actuation
dynamics of IPMC. However, it can only capture the
propulsion generated by bending only motion. If a 3D
kinematic motion is generated by a caudal fin or pectoral fin,
the model needs to be modified to capture the thrust
created by the fin which will be also three dimensional. The
thrust can be calculated by integrating the hydrodynamic
force acting on the fin. It will be a great challenge to
capture the fluid-to-soft-membrane interaction since the
boundary conditions of the fluid dynamics PDE equation
are more complicated since the shape of the membrane
changes with time. An approximation method must be
found to simplify the 3D modeling of complex pectoral
fin or caudal fin, which would be a future direction in this
The fabrication process for creating an IPMC that is
capable of 3D kinematic motion follows two different
approaches. The lithography-based approach is able to
create meso- or microsize pectoral fin and it is suitable
for batch production. However, the passive area of the
fin is still Nafion membrane, which is not as stretchable
as PDMS material. The twisting angle of 3D kinematic
motion is constrained by the passive area even the active
areas are controlled differently. The assembly-based
approach solves the problem in the passive areas since a
soft and stretchable PDMS can be selected in those areas.
However, the process is non-monolithic and
unsuitable for batch production. The process is also unable to
create meso- or microsize pectoral fin. As a conclusion,
each process has unique advantages and disadvantages.
Choosing which process for making pectoral fin depends
on the size of the robot and its application. The future
direction of fabricating 3D deformable membrane would
be 3D printing other soft materials and Nafion film into
a seamless and arbitrary-shaped membrane which will
consist of active areas and passive areas. This printing
Fig. 16 Snapshots of forward swimming test [
process would be either scaled up or scaled down, which
could print mesoscale or microscale fish fins. The
challenges would be 3D printing of two different soft
materials in one platform.
Three types of bio-inspired underwater robot were
reviewed in this paper. The robotic fish propelled by a
caudal fin shows a reasonable good swimming forward
performance (0.125 BL/s), while the robotic manta ray
shows a slow swimming forward performance (0.053
BL/s). The 2D maneuverable robotic fish propelled by
multiple IPMC fins showed some maneuvering
capabilities (forward speed 0.067 BL/s and 2.5 degree/s), which
are not very promising. The possible reasons might be
that the pectoral fin and caudal fin were not optimally
designed and the body was not optimally designed.
However, using IPMC only in robotic fish or robotic rays
might not be a good idea since IPMC cannot generate
high-frequency flapping which is really needed for
highspeed swimming or quick turning. The future direction
of bio-inspired robots design using IPMC would be
combining both IPMC and other fast responsive actuators,
such as electrical motors, to achieve both high speed
and high maneuvering capabilities. The challenges would
be bio-inspired design of a hybrid fish tail and dynamic
modeling and control of the robot propelled by such
The author wants to thank NSF’s support for this review.
The author declares that he has no competing interests.
Availability of data and materials
This is a review paper. For all the data and material, please contact the original
authors of the papers that are cited in this paper.
Ethics approval and consent to participate
Ethic approval was not sought for this paper as all the data extracted were
from publicly available datasets; which were scrutinized by the data owner
before release and publication. This review does not individual level data.
This review is funded by NSF under the grant CNS # 1446557.
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
1. Irons KS , Sass G , McClelland M , Stafford J . Reduced condition factor of two native fish species coincident with invasion of non-native Asian carps in the Illinois River, USA Is this evidence for competition and reduced fitness ? J Fish Biol . 2007 ; 71 : 258 - 73 .
2. DeGrandchamp KL , Garvey JE , Colombo RE . Movement and habitat selection by invasive Asian carps in a large river . Trans Am Fish Soc . 2008 ; 137 : 45 - 56 .
3. Lighthill M. Large-amplitude elongated-body theory of fish locomotion . Proc R Soc Lond B Biol Sci . 1971 ; 179 : 125 - 38 .
4. Lauder GV , Anderson EJ , Tangorra J , Madden PG . Fish biorobotics: kinematics and hydrodynamics of self-propulsion . J Exp Biol . 2007 ; 210 : 2767 - 80 .
5. Moored KW , Smith W , Hester J , Chang W , Bart-Smith H . Investigating the thrust production of a myliobatoid-inspired oscillating wing . Adv Sci Technol . 2008 ; 58 : 25 - 30 .
6. Moored K , Bart-Smith H . Investigation of clustered actuation in tensegrity structures . Int J Solids Struct . 2009 ; 46 : 3272 - 81 .
7. Morgansen KA , Triplett BI , Klein DJ . Geometric methods for modeling and control of free-swimming fin-actuated underwater vehicles . IEEE Trans Rob . 2007 ; 23 : 1184 - 99 .
8. Bar-Cohen Y . Electroactive polymers as artificial muscles: capabilities, potentials and challenges . Robotics . 2000 ; 2000 : 188 - 96 .
9. Shahinpoor M , Bar-Cohen Y , Simpson J , Smith J . Ionic polymer-metal composites (IPMCs) as biomimetic sensors, actuators and artificial muscles-a review . Smart Mater Struct . 1998 ; 7 : R15 .
10. Carpi F , De Rossi D , Kornbluh R , Pelrine RE , Sommer-Larsen P . Dielectric elastomers as electromechanical transducers: fundamentals, materials, devices, models and applications of an emerging electroactive polymer technology . Amsterdam: Elsevier; 2011 .
11. Pelrine R , Kornbluh R , Pei Q , Joseph J . High-speed electrically actuated elastomers with strain greater than 100% . Science. 2000 ; 287 : 836 - 9 .
12. Suo Z. Theory of dielectric elastomers . Acta Mech Solida Sin . 2010 ; 23 : 549 - 78 .
13. Shahinpoor M , Kim KJ . Ionic polymer-metal composites: I. Fundamentals . Smart Mater Struct. 2001 ; 10 : 819 .
14. Chen Z , Tan X. A control-oriented and physics-based model for ionic polymer-metal composite actuators . IEEE/ASME Trans Mechatron . 2008 ; 13 : 519 - 29 .
15. Nemat-Nasser S . Micromechanics of actuation of ionic polymer-metal composites . J Appl Phys . 2002 ; 92 : 2899 - 915 .
16. Ahn KK , Truong DQ , Nam DNC , Yoon JI , Yokota S. Position control of ionic polymer metal composite actuator using quantitative feedback theory . Sens Actuators A . 2010 ; 159 : 204 - 12 .
17. Kang S , Shin J , Kim SJ , Kim HJ , Kim YH . Robust control of ionic polymermetal composites . Smart Mater Struct . 2007 ; 16 : 2457 .
18. Chen X , Su C -Y. Adaptive control for ionic polymer-metal composite actuators . IEEE Trans Syst Man Cybern Syst . 2016 ; 46 : 1468 - 77 .
19. Leang KK , Shan Y , Song S , Kim KJ . Integrated sensing for IPMC actuators using strain gages for underwater applications . IEEE/ASME Trans Mechatron . 2012 ; 17 : 345 - 55 .
20. Guo S , Fukuda T , Kato N , Oguro K. Development of underwater microrobot using ICPF actuator . In: Proceedings of IEEE international conference on robotics and automation , 1998 , p. 1829 - 34 .
21. Laurent G , Piat D. Efficiency of swimming microrobots using ionic polymer metal composite actuators . In: Proceedings of 2001 ICRA IEEE international conference on robotics and automation , 2001 , p. 3914 - 3919 .
22. Tan X , Kim D , Usher N , Laboy D , Jackson J , Kapetanovic A , et al. An autonomous robotic fish for mobile sensing . In: 2006 IEEE/RSJ international conference on intelligent robots and systems , 2006 , p. 5424 - 29 .
23. Punning A , Anton M , Kruusmaa M , Aabloo A. A biologically inspired ray-like underwater robot with electroactive polymer pectoral fins . In: International IEEE conference on mechatronics and robotics , 2004 , p. 241 - 5 .
24. Yeom S-W , Oh I-K . A biomimetic jellyfish robot based on ionic polymer metal composite actuators . Smart Mater Struct . 2009 ; 18 : 085002 .
25. Pak NN , Scapellato S , La Spina G , Pernorio G , Menciassi A , Dario P . Biomimetic design of a polychaete robot using IPMC actuator . In: The first IEEE/RAS-EMBS international conference on biomedical robotics and biomechatronics , BioRob 2006 , p. 666 - 71 .
26. Liu J , Hu H , Gu D . A hybrid control architecture for autonomous robotic fish . In: 2006 IEEE/RSJ international conference on intelligent robots and systems , 2006 , p. 312 - 17 .
27. Hu H , Liu J , Dukes I , Francis G . Design of 3D swim patterns for autonomous robotic fish . In: 2006 IEEE/RSJ international conference on intelligent robots and systems , 2006 , p. 2406 - 11 .
28. Kopman V , Laut J , Polverino G , Porfiri M . Closed-loop control of zebrafish response using a bioinspired robotic-fish in a preference test . J R Soc Interface . 2013 ; 10 : 20120540 .
29. Yu J , Liu L , Wang L , Tan M , Xu D. Turning control of a multilink biomimetic robotic fish . IEEE Trans Rob . 2008 ; 24 : 201 - 6 .
30. Low KH . Locomotion and depth control of robotic fish with modular undulating fins . Int J Autom Comput . 2006 ; 3 : 348 - 57 .
31. Zhao W , Hu Y , Zhang L , Wang L . Design and CPG-based control of biomimetic robotic fish . IET Control Theory Appl . 2009 ; 3 : 281 - 93 .
32. Yu J , Tan M , Wang S , Chen E. “ Development of a biomimetic robotic fish and its control algorithm . IEEE Trans Syst Man Cybern Part B (Cybern) . 2004 ; 34 : 1798 - 810 .
33. Chen Z , Shatara S , Tan X . Modeling of biomimetic robotic fish propelled by an ionic polymer-metal composite caudal fin . IEEE/ASME Trans Mechatron . 2010 ; 15 : 448 - 59 .
34. Aureli M , Kopman V , Porfiri M . Free-locomotion of underwater vehicles actuated by ionic polymer metal composites . IEEE/ASME Trans Mechatron . 2010 ; 15 : 603 - 14 .
35. Farinholt KM . Modeling and characterization of ionic polymer transducers for sensing and actuation . 2005 .
36. Nemat-Nasser S , Li JY . Electromechanical response of ionic polymer-metal composites . J Appl Phys . 2000 ; 87 : 3321 - 31 .
37. Shahinpoor M , Kim KJ . The effect of surface-electrode resistance on the performance of ionic polymer-metal composite (IPMC) artificial muscles . Smart Mater Struct . 2000 ; 9 : 543 .
38. Clough RW , Penzien J . Dynamics of structures. Walnut Creek: Computers & Structures Inc; 2003 .
39. Sader JE . Frequency response of cantilever beams immersed in viscous fluids with applications to the atomic force microscope . J Appl Phys . 1998 ; 84 : 64 - 76 .
40. Lu P , Lee K. An alternative derivation of dynamic admittance matrix of piezoelectric cantilever bimorph . J Sound Vib . 2003 ; 266 : 723 - 35 .
41. Chen Z , Tan X . Monolithic fabrication of ionic polymer-metal composite actuators capable of complex deformation . Sens Actuators A . 2010 ; 157 : 246 - 57 .
42. Chen Z , Um TI , Bart-Smith H . A novel fabrication of ionic polymer-metal composite membrane actuator capable of 3-dimensional kinematic motions . Sens Actuators A . 2011 ; 168 : 131 - 9 .
43. Kim KJ , Shahinpoor M. Ionic polymer-metal composites: II. Manufacturing techniques . Smart Mater Struct . 2003 ; 12 : 65 .
44. Kim KJ , Shahinpoor M. A novel method of manufacturing three-dimensional ionic polymer-metal composites (IPMCs) biomimetic sensors, actuators and artificial muscles . Polymer . 2002 ; 43 : 797 - 802 .
45. Chung C-K , Fung P , Hong Y , Ju M-S , Lin C-CK , Wu T. A novel fabrication of ionic polymer-metal composites (IPMC) actuator with silver nano-powders . Sens Actuators B Chem . 2006 ; 117 : 367 - 75 .
46. Kim SJ , Lee IT , Kim YH . Performance enhancement of IPMC actuator by plasma surface treatment . Smart Mater Struct . 2007 ; 16 : N6 .
47. Lee SJ , Han MJ , Kim SJ , Jho JY , Lee HY , Kim YH . A new fabrication method for IPMC actuators and application to artificial fingers . Smart Mater Struct . 2006 ; 15 : 1217 .
48. Chen Z , Um TI , Bart-Smith H . Bio-inspired robotic manta ray powered by ionic polymer-metal composite artificial muscles . Int J Smart Nano Mater . 2012 ; 3 : 296 - 308 .
49. Ye Z , Hou P , Abbaspour A , Chen Z. 2D maneuverable robotic fish propelled by multiple ionic polymer-metal composite artificial fins . Int J Intell Robot Appl . 2017 ; 1 : 195 - 208 .