Content uploaded by Hsieh-Yu Li
Author content
All content in this area was uploaded by Hsieh-Yu Li on Jan 21, 2019
Content may be subject to copyright.
Abstract— Micro-scale operation is a challenging and time-
consuming procedure for humans. With the involuntary
motions like physiological hand tremor and jerks, operator’s
performance is limited and the inaccuracy increases the risk of
damage to nearby tissues or products. A cooperative
micromanipulator whose end-effector with the instrument is
stiff in response to human’s intended force is a potential
solution to reduce the tremors and increase the accuracy of the
operation. However, operating such a stiff micromanipulator is
energy consuming in some particular operational path. For
instance, it requires the operator to move the instrument with
large force if a large operational path is desired. During tasks
with large movements, compliance is often needed rather than
accuracy. Hence, in this paper, we design a collaborative
control that ensures compliance or accuracy of the operation
based on human’s intended motion. Particularly, a variable
admittance controller is employed to achieve the best trade-off
between compliance and accuracy. We use micro-suturing as
an illustration and conduct the experiments. The experimental
results verify the effectiveness of the control design and the
improvement for performance of the micromanipulator.
I. INTRODUCTION
Micro-scale operations, such as microsurgery [1]–[5] and
micro-devices assembly [6], have been gaining increases in
demand among different industries and have brought
revolution in many aspects. However, it is challenging to
deal with the micro-scale tasks because the accuracy
requirements of the operation are close to the limit of the
human hand accuracy. An example is in micro-assembly.
Automation is not possible during the design phase of micro-
devices as the parts are not fixed and micro-assembly is
required [6]. Another example is in microsurgery, where the
procedure is challenging and time-consuming for human
operation as some of the anatomical structures of sizes (e.g.,
veins on the human retina ranges 40 to 350 μm [3]) are small
and fragile [5]. Both mentioned operations are examples of
tasks that are limited by physiological hand tremors.
To improve the performance of micro-scale operation in
terms of accuracy, one of the potential approaches is the
cooperative micromanipulators whose end-effector is
response to the force exerted by the human operator. A
force/torque (F/T) is normally mounted on the end-effector
with the operational instrument to sense the human forces.
The force information is used as the input for robot
1
H.-Y. Li, and U-X. Tan are with Pillar of Engineering Product
Development, Singapore University of Technology and Design, Singapore
(email: hsiehyu_li@sutd.edu.sg, uxuan_tan@sutd.edu.sg).
2
T. Nuradha, S. A. Xavier is with Department of Electronic and
Telecommunication, Faculty of Engineering, University of Moratuwa, Sri
Lanka (email:theshani.nuradha@gmail.com, alex26xavier@gmail.com).
controllers to provide tremor-free, precise positional control
for the operator [3].
A well-established cooperative micromanipulator, Steady-
Hand, is developed to aid microsurgical operations by
providing precise movements on small scale tasks [1], [3],
[7]–[10]. Steady-Hand is designed with three Cartesian joints
in global motions and two degree-of-freedom (DOF) for tilt
mechanism along with proportional controllers to convert
human force to robot desired velocity [1], [3], [7], [8]. The
general motion is stiff since the robot is controlled for precise
and tremor-free cooperation. The other robotic manipulators
are also developed to assist vitreoretinal surgeons in [9], [10].
However, the cooperative controller usually provides stiff
co-manipulation to enhance the accuracy by tremor
suppression. Precise operator’s hand motion is achieved but
the compliance is compromised. Compliance is also essential
in some of the operational paths that involve large
movements. A stiff operation causes inconvenience for the
operators to execute the cooperative tasks with these robots
as it is time-consuming to maneuver the instrument.
Therefore, this paper is aimed towards developing a
cooperative micromanipulator with both accuracy and
compliance based on different operator’s intended motion.
To ensure the best trade-off between compliance and
accuracy, variable admittance/impedance control is widely
used in physical human-robot interaction (pHRI) [11]–[16].
Researchers in [11], [12] used variable admittance control to
conduct the cooperative task that requires compliance and
accuracy such as calligraphy. In [13]–[15], the robot is
controlled by human to follow or draw a narrow path. A tank-
based approach from [17], [18] was employed with variable
admittance in [16] to minimize the effort for the operator. In
[19], the variable admittance control is employed with an
adaptive controller to provide the interaction stably and
accurately between robot, human and environment. However,
the micro-scale level tasks have less been conducted with
variable admittance control to assist the operation with
respect to operator’s intention.
Variable admittance control is used in [20], [21] to
improve the force scaling in micromanipulation. The
admittance is varied, however, based on the insertion depth
of the tool and it is applied only on 2-DOF to switch between
force scaling mode to virtual RCM which has different
intentions with this paper. Another application for variable
admittance control in cooperative tasks is developed in [22]
to achieve positional accuracy of neurosurgery based on prior
knowledge of surgical sites. Despite the promising
development, most of the cooperative micromanipulator are
aimed for high precision. The compliance is still lacked in
large operational motion and it appears that the source of the
gap stems from the challenges to design a controller with
both the compliance and accuracy.
Towards A Compliant and Accurate Cooperative
Micromanipulator using Variable Admittance Control
Hsieh-Yu Li
1
, Member, IEEE, Theshani Nuradha
2
, Student Member, IEEE, Sebaratnam Alex Xavier
2
,
Student Member, IEEE, and U-Xuan Tan
1
, Member, IEEE
2018 3rd International Conference on Advanced Robotics an
d
Mechatronics (ICARM)
978-1-5386-7066-8/18/$31.00 ©2018 IEEE 230
Hence, to address the need, variable admittance control is
employed in this paper to achieve the best trade-off between
accuracy and compliance. The intuitive human motion is
achieved by varying the admittance based on their
operational force. With such a controller and developed
robot, we hope to enhance the performance of the micro-scale
tasks in terms of accuracy with minimum human’s efforts.
Fig 1 illustrates an overview of our objective. To assist
the operators, in the existing cooperative micromanipulator
(A), the operation is stiff for precision and tremor
suppression. Our target (B) is to build a micro-scale
cooperative robot whose precision is achieved during fine
motion and at the same time more compliance is obtained for
large motion.
Stiff
Instrument
Cooperative
Micro-
manipulator
Large motion fine motion
Stiff
Compliance Stiff
Instrument
Cooperative
Micro-
manipulator
Large motion fine motion
Fig 1. (A) Existing cooperative micromanipulators are stiff (accuracy) for
all motion and (B) our proposed robot is compliant for large movement and
stiff for accuracy during fine motion.
In this paper, we use microsurgical suturing with the
cooperative robot as illustration to demonstrate our idea due
to the complexity of this task. Microsurgical suturing is
composed of knot tying, thread operation and dexterous
needle manipulation [23]. In general, needle insertion and
extraction are two main targeted points where the accuracy is
of the utmost importance. However, the operational paths
before and after these two points normally involve large
movements (e.g., pull the needle out of the extraction point)
where the compliance is required. Therefore, we focus on the
operation around these two main points in this paper as the
evaluation of the performance for the proposed controller in
terms of accuracy and compliance. To achieve compliance or
accuracy with respect to different paths, the controller varies
the cooperative behavior of the robot based on the operator’s
intention, which also serve as the main contribution.
The paper is structured as follows. The problem statement
of the control design in existing research for the
micromanipulator is first addressed in Section II. The
proposed variable admittance control is then designed and
discussed in Section III. Section IV describes the proposed
robotic system and experiments are conducted in Section V.
Finally, conclusion is presented in Section VI.
II. PROBLEM STATEMENT
In pHRI using touch as the sense of interaction on the
end-effector, the direction and the magnitude of the force
measured by the F/T sensor is typically interpreted as the
human operator’s intention [24]. This force/torque
information is employed as input of the controller to
engender the desired velocity input for a velocity-controlled
robot [1], [11], [14], [16], [19], [24]. More specifically in
micro-scale tasks, a proportional controller is often employed
to deal with the human cooperative forces [1], [3]. The
controller is defined as
dpext
=
xkF
(1)
where
n
d
R
∈
x
is the desired velocity, n is the number of
DOF,
n
p
R
∈k
is the proportional gain and
n
ext
R
∈F
is the
external human force applied on the F/T sensor.
Another type of control in pHRI is admittance control [25]
dd dd ext
+=
Mx Bx F
(2)
where
n
d
R
∈
x
is the desired acceleration, and
,nxn
dd
∈MB R
are the desired inertia and damping. However,
for both controllers, the adjustments of the parameters are
not flexible. In other words, the stiffness of the robot co-
manipulation with human interaction is fixed. If the robot is
compliant in response to human force, the accuracy is
compromised, and vice versa. Therefore, to address this
issue, we apply the variable admittance control to enhance
the performance by varying the inferred human force and
robot velocity.
III. VARIABLE ADMITTANCE CONTROL IN MICROSURGICAL
SUTURING
This section illustrates how our proposed controller
contributes to microsurgical suturing in terms of compliance
and accuracy. In addition, the conditions of passivity for our
proposed on-line adjustments of admittance parameters are
also discussed.
A.
Controller Design Concept for Microsurgical Suturing.
In microsurgical suturing, needle insertion and extraction
are two key moments where the accuracy is most needed.
Targeting one precise insertion or extraction point and
holding the needle until insertion is energy consuming
causing hand-tremor. The robot with stiff co-manipulation
provides precise control with tremor suppression resulting in
better performance [1], [3]. On the contrary, when the
operator extracts the needle and makes the loop to tie knots,
the hand motion involves large movements where fast
acceleration is typically desired. The fast acceleration with
manual operational force is described as compliance. Using a
robot with compliant co-manipulation, the operator can move
the instrument with minimum efforts [13], [14], [24].
The objective of variable admittance control is to provide
compliant or accurate co-manipulation based on the human
intentions with respect to different operational paths [13]–
[15]. To ensure the best trade-off between accuracy and
compliance, the admittance is adjusted with the force exerted
by the operator. Low admittance parameters are required
when fine and accurate movements in low accelerations are
performed by the operator; and high admittance parameters
are recommended for movements involving large
accelerations [13], [14], [19], [24]. In microsurgery, to
2018 3rd International Conference on Advanced Robotics an
d
Mechatronics (ICARM)
978-1-5386-7066-8/18/$31.00 ©2018 IEEE 231
attenuate the physiological hand tremor for fine movement
(where the insertion and extraction are normally carried out),
low admittance should be designed because it provides the
stiff co-manipulation. On the other hand, compliance is
desired when the movement involves large acceleration,
such as knotting or dexterous hand motion, and high
admittance should be used. With the high admittance control,
the robot provides compliant co-manipulation for the
operator.
Hence, it seems to be a practical and feasible solution to
apply variable admittance control on the microsurgical robot
for suturing. With the variable admittance, the robot not only
maintains the precise control with tremor suppression for
needle insertion and extraction but yields the compliant co-
manipulation when the operator intends to execute large
motion.
B.
Variable Admittance Control.
1)
Control Design
The variable admittance control is defined as
() ()
d d d d ext
tt+=
MxBxF
(3)
where
,n
dd
R
∈xx
are the desired velocity and acceleration in
the task space, and
(), () nxn
dd
ttR∈MB
are the desired
variable inertia and damping matrices, which are symmetric,
diagonal, positive definite for all
n
d
R
∈
x
.
d
x
ext
F
x
Fig 2. Control scheme of the variable admittance control
The control scheme is illustrated in Fig 2 where the human
force is the external force and the admittance parameters,
()
d
tM
and
()
d
tB
, are adjusted according to it. Based on the
concepts mentioned in the previous sub-section, the
admittance is reduced to improve accuracy at small
operational force profiles when fine movements are required;
vice versa, the admittance should be increased for compliant
motion when operational force exerted by the human is
larger. Therefore, for every axis, the parameter adjustment of
the admittance is therefore used as
,0,ext,
() 0
di i i i
Bt B F
α
=− >
(4)
,
,0,
0,
()
() 0
di
di i
i
Bt
Mt M B
=>
(5)
where
M
0,i
and
B
0,i
are the initial values of the desired inertia
and damping in
i
th
axis when no external force is applied,
F
ext,i
is the external force with respect to the corresponding
i
th
axis for the admittance,
i
α
is the updated gain for damping
and
i
= 1,…,
n
. With this approach, the admittance is
maintained in the smallest value (largest
B
d
(
t
) and
M
d
(
t
))
when the external force is zero. Moreover, for low
acceleration, the admittance will be near the initial values that
benefit the operator to execute fine movement as the robot is
stiff. Conversely, given
B
d
(
t
) and
M
d
(
t
) that are reduced if the
operators wants to create large hand motion, the admittance is
increased for compliant co-manipulation. This shows our
proposed controller varies the admittance of the robot with
respect to the variation of
F
ext
to provide the corresponding
cooperative interaction based on human’s intention.
2)
Proof of Passivity
Passivity condition of the proposed controller is derived in
this section. In the application of pHRI, the tracking error is
assumed to be negligible [13], [14], [16], namely,
d
≅
x
x
,
where
x
is the robot velocity output. In addition, because
ext
F
varies with time, equation (3) is taken as a closed-loop
time-varying system
() ()
ddext
tt+=
MxBxF
(6)
with respect to the input-output pair
()
,
ext
Fx
and with a
storage function
1
() () .
2
T
d
Vt t=
x
Mx
(7)
The derivative of
()Vt
is
1
() () () .
2
TT
dd
Vt t t=+
x
Mx xMx
(8)
Taking
x
from (6) and replace it in (8), we obtain
[]
1
() () () .
2
TT
ext d d
Vt t t=−+
x
FBx xMx
(9)
By rearranging some terms, equation (9) becomes
1
() () 2 () .
2
TT
ext d d
Vt t t
ªº
=+ −
¬¼
x
FxM Bx
(10)
Therefore
0
0
0
Stored Energy Dissipated Energy
() (0)
1
( ) 2 ( )
2
( ) (0) ( )
tT
ext
tT
dd
t
dVtV
ttd
Vt V W d
τ
τ
ττ
=−
ªº
−−
¬¼
=−−
³
³
³
xF
xM B x
(11)
If the dissipated energy
0
() 0,
t
Wd
ττ
≤
³
(12)
we can obtain
0
() (0) (0).
tT
ext
dVtV V
τ
≥− ≥−
³
xF
(13)
Hence, the passivity is guaranteed if
() 2 () 0.
dd
tt−≤MB
(14)
The condition (14) guarantees the closed-loop system (6) is
always dissipative, which also ensures the stable interaction
2018 3rd International Conference on Advanced Robotics an
d
Mechatronics (ICARM)
978-1-5386-7066-8/18/$31.00 ©2018 IEEE 232
between variable-admittance-controlled robot and the human
operator
.
IV.
M
ICROSURGICAL
R
OBOTIC
S
YSTEM
In this section, the developed micromanipulator and the
mechatronic units are described. The 5-DOF robotic system
for micro-suturing is developed as shown in Fig 3. Motion
control is implemented with a 3-axis translational motor
actuated stage (8MT173-30; Standa Ltd., Lithuania), 1
rotational stepper (8MR191-28; Standa Ltd., Lithuania) for
yaw axis at the base and another rotational stepper (8MR174-
11-28; Standa Ltd., Lithuania) for roll axis at the end-
effector. The control units are 3 multi-axis controller
(8SMC4; Standa Ltd., Lithuania). A 6-DOF F/T sensor
(HEX-70-XE-200N, OptoForce Ltd, Hungary) is mounted on
the last joint, roll axis, with a needle holder. A customized
portable USB digital camera is used to carry out microscope
imaging. A customized platform that consists a thin and
flexible natural rubber latex is used to simulate the artificial
tissue surface. The shape of the microsurgical needle is 3/8
and the size is 7-0.
5-DOF
Micromanipulator
6-DOF
F/T Sensor
Portable
Microscope
Artificial Tissue
Surface
Operator’s Hand
Needle Holder
(instrument) xy
z
Needle
Fig 3. The appearance of our microsurgical robotics system
V.
E
XPERIMENTATION
In this section, the experimental setup is described first.
Experiments are then conducted to verify the effectiveness of
the proposed variable admittance controller. (A video
demonstration is in https://youtu.be/j8EocAyWW94 )
A.
Experimental Setup
A general experimental setup is depicted in this sub-
section. Three needle paths under microscope imaging of the
artificial tissue surface that consists of the cut and two
targeted points, namely insertion and extraction, are shown
in Fig 4. The operator guides the needle with the instrument
using the hand to the first targeted point for insertion (path
1). After crossing beneath the first point to the second one
for extraction (path 2), the needle is pulled out by non-
dominant hand (path 3).
In addition, to evaluate our proposed variable admittance
control that achieves the best trade-off between compliance
and accuracy of the robot, each experiment compares with
constant low and high admittance. High admittance is
designed for compliance and low admittance is desired for
accuracy. Parameters of the variable admittance controller
are M
0
= diag[10 10 50 0.01 0.01] kg and kgm
2
, B
0
=
diag[1000 1000 1000 5.75 0.192] Ns/m and Ns/rad and
α
=
diag[100 100 100 8.64 0.152]. Parameters of constant
admittance are set as: low admittance B
max
= B
0
(the
maximum value of B
d
(t)); high admittance B
min
= diag[500
500 500 1.8. 0.1] Ns/m and Ns/rad with M
0
= M
max
= M
min
.
Moreover, the hand tremor in one of the gestures, such as
tool spin, is not significant [3]. Hence, for better stability,
the rotational axes are conducted with constant inertia and
variable damping [14]. A low pass filter of 5Hz cut-off
frequency with dead-zone and saturation [13] is used for
noise cancellation. The condition (14) is guaranteed by
α
that is chosen empirically.
1 mm
Targeted point 1
(Insertion)
Targeted point 2
(Extraction) Artificial cut
Robot
starting point
Above tissue
Beneath tissue
Fig 4. The path of the needle from insertion (targeted point 1) to extraction
(targeted point 2) under the microscope imaging.
B. Best Trade-off Between Compliance and Accuracy
In the experiments, compliance and accuracy of the robot
have been evaluated here. Two experiments have been
tested: (1) an operator extracts needle from point 2 to test the
performance of compliance, and (2) a subject test for the
proposed system in suturing, from insertion to extraction, is
reported to compare the compliance and accuracy.
(1). Compliance Test - Extraction (Path 3)
In this experiment, in order to demonstrate the compliance
of the system, an extraction experiment is designed as this
hand motion that involves large movement. Before the test,
the needle is inserted in the artificial surface and the tip of it
is at the other side of point 2. To simply the test, the operator
extracts the needle out of the targeted point with the robot
along the x-axis rather than along multi-DOF. The force
exerted on the x-axis is shown in Fig 5. A largest force is
demanded in the low admittance and a smaller force is
applied for the high admittance. The result of variable
admittance control is the same with [13]–[15], i.e., the force
is closer to the result of the high admittance when it involves
large movement. Based on this conclusion, subject tests are
evaluated next.
Compliance
Fig 5. Compliance Test, the lower admittance the higher compliant of the
robot.
(2). Subjects Test – Insertion and extraction (Path 1 to 3)
The paths including needle insertion and extraction are
carried out in this experiment with the same parameters
above. The paths involve a) compliant movement when the
2018 3rd International Conference on Advanced Robotics an
d
Mechatronics (ICARM)
978-1-5386-7066-8/18/$31.00 ©2018 IEEE 233
needle is approaching the insertion point and b) fine
movement when two of the points are being inserted or
extracted. During the experiments, the position of the robot
begins at the starting point. Point 1 is targeted for needle
insertion and point 2 is desired for the operators to extract
the needle. Once the needle is pulled out, the task is
completed.
The experiments are performed by 5 subjects aged
between 23 and 29, who have no experience in suturing.
After some practice runs, subjects perform the task using
their dominate hand for 3 times for each admittance
controller. The subjects are asked to perform the task as
accurately as possible. They have to drive the needle till it is
inserted into the correct points. Given this accuracy task that
involve fine movement (insertion and extraction), the
operational forces are recorded with respect to each
admittance to check the compliance. In addition, another
performance parameter is execution time of the task, defined
as when the subject starts from path (1) to approach the
insertion point till the needle is extracted in path (3).
The properties and performance of the controllers for a
single test can be seen in Fig 6. Because low admittance
only provides stiff co-manipulation with any human
intention (blue), the operational forces in our proposed
controller (red) is smaller as the damping is varying resulting
in minimum human efforts. While high admittance (green)
ensures the compliance of the operation, it takes longer time
than the other two methods.
Fig 6. The results in x direction for subject #1. (Top) Time history of
damping, and (bottom) the absolute value of external force with respect to
each admittance controller.
We found out that during the paths that require
compliance and large movement, the subjects tend to drive
the needle with the maximum speed of the robot (as a
normal free hand motion is usually faster than the maximum
speed of a micromanipulator). This cause similar execution
time of these paths with all three admittance controllers.
However, during insertion and extraction, the execution time
is significantly affected as a low admittance provides stiff
co-manipulation with tremor suppression resulting in easier
operation for the operators to target the precise points. This
explains why low and variable admittance control achieves
better performance in terms of time.
Fig 7 and Fig 8 show the task completion time and the
magnitude of resultant forces of translational axes and
torques of rotational axes. The large marks represent the
average of five subjects. As seen from the figures, the low
admittance engenders less total execution time as the
operators have more precise control on targeting fine
movement, but compliance is compromised and the
operators need to apply larger forces or torques on the robot.
On the contrary, the compliance is achieved in high
admittance but it takes longer execution time since the robot
is sensitive to the applied force and the operators now have
the difficulty on fine movement. Finally, with the proposed
variable admittance controller, the operator is able to target a
precise movement and is also able to conduct a large motion
with relatively small force. Hence, the results conclude that
the robot now is controlled in decent trade-off between
compliance and accuracy based on the operator’s intentions.
Fig 7. Magnitude of resultant translational forces in the experiments on five
subjects using constant and variable admittance control. Larger marks
represent the average of the corresponding group of results.
Fig 8. Magnitude of resultant rotational torques in the experiments on five
subjects using constant and variable admittance control. Larger marks
represent the average of the corresponding group of results.
C. Discussion
Fig 9 depicts the paths (side view) that require accuracy or
compliance. During the experiments, the most critical points,
where precision and tremor suppression are required, are
insertion and extraction. At these two points (path (1b) to
(2b)), the desired human intention is to guide the instrument
in a low acceleration. Low admittance provides a stiff co-
manipulation with tremor suppression for the operator to
target the point precisely. On the contrary, the motions that
involve approaching or pulling require (path (1a) and (3))
large acceleration. High admittance ensures compliance of
robot for the operator to execute the tasks with smaller
forces, such as the result in Fig 5, where the accuracy is less
required. The experimental results show the performance in
micro-scale operation with our robot achieves the best trade-
off between compliance and accuracy which concludes our
contribution in this paper.
2018 3rd International Conference on Advanced Robotics an
d
Mechatronics (ICARM)
978-1-5386-7066-8/18/$31.00 ©2018 IEEE 234
(3)
Extraction
Tissue surface
(Side View)
Insertion
(1a)
(1b)
(2a)
(2b)
Pulling
approaching
(Compliance) (Compliance)
(Accuracy)
Path
Fig 9. Path (side view) of microsurgical insertion and extraction with
respect to compliance and accuracy. Based on the function, path 1 is
separated into 1a & 1b and path 2 is 2a & 2b. Path 1a & 3 involve large
motion such as pulling out the needle and the compliance is recommended.
Path 1b, 2a and 2b which are critical for precision require tremor
suppression for accuracy.
VI.
C
ONCLUSION
In this paper, a cooperative robotic system is developed
for micro-scale operation. An instrument at the end-effector
of the robot is controlled in response to the operator’s force.
With the proposed variable admittance control, compliance
or accuracy of co-manipulation is achieved with respect to
different human’s intended motion. The experiments of a
micro-suturing that involves insertion and extraction are
illustrated. Experimental results support that the
performance of the robot achieves the best trade-off between
compliance and accuracy based on operator’s intentions. In
addition, though the controller is specifically applied to
micro-suturing in this paper, it can also be applied into other
application such as the development of micro-devices
assembly that also requires dexterity and accuracy of hand
motion.
With the proposed system, we hope to contribute towards
the efficiency and performance enhancement of micro-scale
operations. The goal is to assist the operators by increasing
the accuracy for the operations with minimum human
efforts.
R
EFERENCES
[1] R. Taylor et al., “A Steady-Hand Robotic System for
Microsurgical Augmentation,” Int. J. Rob. Res., vol. 18, no. 12,
pp. 1201–1210, 1999.
[2] R. A. MacLachlan, B. C. Becker, J. C. Tabarés, G. W. Podnar, L.
A. Lobes, and C. N. Riviere, “Micron: An actively stabilized
handheld tool for microsurgery,” IEEE Trans. Robot., vol. 28, no.
1, pp. 195–212, 2012.
[3] B. Mitchell et al., “Development and application of a new steady-
hand nanipulator for retinal surgery,” IEEE Int. Conf. Robot.
Autom., no. April, pp. 623–629, 2007.
[4] S. Yang, S. Member, R. a Maclachlan, C. N. Riviere, and S.
Member, “Manipulator Design and Operation of a Microsurgical
Instrument,” IEEE/ASME Trans. Mechatronics, vol. 20, no. 2, pp.
1–12, 2014.
[5] B. C. Becker, S. Voros, R. a Maclachlan, G. D. Hager, and C. N.
Riviere, “Active Guidance of a Handheld Micromanipulator using
Visual Servoing.,” IEEE Int. Conf. Robot. Autom., vol. 2009, pp.
339–344, 2009.
[6] J. Cecil, M. B. Bharathi Raj Kumar, Y. Lu, and V. Basallali, “A
review of micro-devices assembly techniques and technology,”
Int. J. Adv. Manuf. Technol., vol. 83, no. 9–12, pp. 1569–1581,
2016.
[7] X. He, M. Balicki, P. Gehlbach, J. Handa, R. Taylor, and I.
Iordachita, “A novel dual force sensing instrument with
cooperative robotic assistant for vitreoretinal surgery,” IEEE Int.
Conf. Robot. Autom., pp. 213–218, 2013.
[8] A. Üneri, M. A. Balicki, J. Handa, P. Gehlbach, R. H. Taylor, and
I. Iordachita, “New steady-hand eye robot with micro-force
sensing for vitreoretinal surgery,” 2010 3rd IEEE RAS EMBS Int.
Conf. Biomed. Robot. Biomechatronics, BioRob 2010, pp. 814–
819, 2010.
[9] A. Gijbels, N. Wouters, P. Stalmans, H. Van Brussel, D.
Reynaerts, and E. Vander Poorten, “Design and realisation of a
novel robotic manipulator for retinal surgery,” IEEE Int. Conf.
Intell. Robot. Syst., pp. 3598–3603, 2013.
[10] A. Degirmenci, F. L. Hammond, J. B. Gafford, C. J. Walsh, R. J.
Wood, and R. D. Howe, “Design and control of a parallel linkage
wrist for robotic microsurgery,” IEEE Int. Conf. Intell. Robot.
Syst., vol. 2015–Decem, pp. 222–228, 2015.
[11] T. Tsumugiwa, R. Yokogawa, and K. Hara, “Variable impedance
control with regard to working process for man-machine
cooperation-work system,” Proc. 2001 IEEE/RSJ Int. Conf. Intell.
Robot. Syst., vol. 3, pp. 1564–1569, 2001.
[12] T. Tsumugiwa, R. Yokogawa, and K. Hara, “Variable impedance
control based on estimation of human arm stiffness for human-
robot cooperative calligraphic task,” Proc. 2002 IEEE Int. Conf.
Robot. Autom., vol. 1, no. May, pp. 644–650, 2002.
[13] A. Lecours, B. Mayer-St-Onge, and C. Gosselin, “Variable
admittance control of a four-degree-of-freedom intelligent assist
device,” Proc. - IEEE Int. Conf. Robot. Autom., no. 2, pp. 3903–
3908, 2012.
[14] F. Ficuciello, L. Villani, and B. Siciliano, “Variable Impedance
Control of Redundant Manipulators for Intuitive Human – Robot
Physical Interaction,” IEEE Trans. Robot., vol. 31, no. 4, pp. 1–
14, 2015.
[15] S. Grafakos, F. Dimeas, and N. Aspragathos, “Variable
Admittance Control in pHRI using EMG-based Arm Muscles Co-
Activation,” IEEE Int. Conf. Syst. Man, Cybern., pp. 1–6, 2016.
[16] C. T. Landi, F. Ferraguti, L. Sabattini, C. Secchi, and C. Fantuzzi,
“Admittance Control Parameter Adaptation for Physical Human-
Robot Interaction,” IEEE Int. Conf. Robot. Autom., pp. 2911–
2916, 2017.
[17] F. Ferraguti, C. Secchi, and C. Fantuzzi, “A tank-based approach
to impedance control with variable stiffness.,” IEEE Int. Conf.
Robot. Autom., pp. 4948–4953, 2013.
[18] F. Ferraguti et al., “An Energy Tank-Based Interactive Control
Architecture for Autonomous and Teleoperated Robotic Surgery,”
IEEE Trans. Robot., vol. 31, no. 5, pp. 1073–1088, 2015.
[19] H.-Y. Li et al., “Stable and Compliant Motion of Physical
Human-Robot Interaction Coupled with a Moving Environment
using Variable Admittance and Adaptive Control,” Robot. Autom.
Lett. IEEE, 2018.
[20] M. Balicki, A. Uneri, I. Iordachita, J. Handa, P. Gehlbach, and R.
Taylor, “Micro-force sensing in robot assisted membrane peeling
for vitreoretinal surgery,” Int. Conf. Med. Image Comput.
Comput. Interv., vol. 6363 LNCS, no. PART 3, pp. 303–310,
2010.
[21] X. He, M. Balicki, P. Gehlbach, J. Handa, R. Taylor, and I.
Iordachita, “A multi-function force sensing instrument for
variable admittance robot control in retinal microsurgery,” Proc. -
IEEE Int. Conf. Robot. Autom., pp. 1411–1418, 2014.
[22] E. Beretta, E. De Momi, F. Rodriguez Y Baena, and G. Ferrigno,
“Adaptive hands-on control for reaching and targeting tasks in
surgery,” Int. J. Adv. Robot. Syst., vol. 12, pp. 1–9, 2015.
[23] S. A. Pedram, P. Ferguson, J. Ma, E. Dutson, and J. Rosen,
“Autonomous Suturing Via Surgical Robot: An Algorithm for
Optimal Selection of Needle Diameter , Shape , and Path ,”
IEEE Int. Conf. Robot. Autom., pp. 2391–2398, 2017.
[24] V. Duchaine, B. Mayer St-Onge, D. Gao, and C. Gosselin,
“Stable and intuitive control of an intelligent assist device,” IEEE
Trans. Haptics, vol. 5, no. 2, pp. 148–159, 2012.
[25] N. Hogan, “Impedance control: An approach to manipulation:
Part I—Theory,” J. Dyn. Sys., Meas., Control, vol. 107, pp. 1–7,
1985.
2018 3rd International Conference on Advanced Robotics an
d
Mechatronics (ICARM)
978-1-5386-7066-8/18/$31.00 ©2018 IEEE 235
