Determination of object stiffness control parameters in robot manipulation using a prototype optical three-axis tactile sensor
ABSTRACT This paper presents experimental results to define suitable parameters in object stiffness control using a prototype optical three-axis tactile sensor mounted on robotic fingers. We have developed a novel optical three-axis tactile sensor system based on an optical waveguide transduction method applying image processing technique. We conducted a series of calibration experiments with soft and hard objects to define suitable parameters in object stiffness control. We analyzed normal and shearing forces data detected in the experiments and compiled suitable parameters in an algorithm inside the robot control system. Verification experiment using robotic fingers to manipulate soft object was conducted whose result revealed that the fingerpsilas system managed to recognize the stiffness and safely manipulate the object.
- SourceAvailable from: Rolf P. Würtz[show abstract] [hide abstract]
ABSTRACT: We present a novel tactile sensor, which is applied for dextrous grasping with a simple robot gripper. The hardware novelty consists of an array of capacitive sensors, which couple to the object by means of little brushes of fibers. These sensor elements are very sensitive (with a threshold of about 5 mN) but robust enough not to be damaged during grasping. They yield two types of dynamical tactile information corresponding roughly to two types of tactile sensor in the human skin. The complete sensor consists of a foil-based static force sensor, which yields the total force and the center of the two-dimensional force distribution and is surrounded by an array of the dynamical sensor elements. One such sensor has been mounted on each of the two gripper jaws of our humanoid robot and equipped with the necessary read-out electronics and a CAN bus interface. We describe applications to guiding a robot arm on a desired trajectory with negligible force, reflective grip improvement, and tactile exploration of objects to create a shape representation and find stable grips, which are applied autonomously on the basis of visual recognition.Robotics and Autonomous Systems. 01/2006;
- [show abstract] [hide abstract]
ABSTRACT: This paper describes precision enhancement of an optical three-axis tactile sensor capable of detecting both normal force and tangential force. The sensor's single cell consists of a columnar feeler and 2-by-2 conical feelers. We have derived equations to precisely estimate the three-axis force from the area-sum and area-difference of the conical feelers’contact areas by taking into account wrench-length shrinkage caused by a vertical force. To evaluate the equations and determine constants included in the equations, we performed a series of calibration experiments using a manipulator-mounted tactile sensor and a combined load-testing machine. Subsequently. to evaluate the tactile sensor's practicality. it was mounted on the end of a robotic manipulator which rubbed flat specimens such as brass plates with step-heights of δ=0.05, 0.1, 0.2 mm and a brass plate with no step-height. We showed from the experimental data that the optical three-axis tactile sensor can detect not only the step-heights but also the distribution of the coefficient of friction, and that the sensor can detect fine plate inclination with accuracy to about ±0.4°. Robotica. v.22, n.2, 2004, p.213-221Robotica 01/2004; · 0.88 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: This paper presents development of tactile sensing-based control algorithm for humanoid robot finger system with optical three-axis tactile sensor mounted on fingertips. Our aim is to develop an intelligent control system that can recognize stiffness of unknown objects and respond to sudden changes of objectpsilas weight during object manipulation. For this purpose, we developed a novel optical three-axis tactile sensor system based on an optical waveguide transduction method capable of acquiring normal and shearing forces. We proposed a control algorithm in the finger control system based on tactile and slippage sensations, and analyzed real-time grasp synthesis in object manipulation tasks. The control algorithm was designed to control fingertips movements by defining optimum grasp pressure and perform re-push movement when slippage was detected in object manipulation tasks. Verification experiments using humanoid robot fingers were conducted whose results revealed that the fingerpsilas system managed to recognize the stiffness of unknown objects and complied with sudden changes of the objectpsilas weight during object manipulation tasks.Robot and Human Interactive Communication, 2008. RO-MAN 2008. The 17th IEEE International Symposium on; 09/2008
Determination of Object Stiffness Control Parameters
in Robot Manipulation Using a Prototype Optical
Three-Axis Tactile Sensor
Hanafiah Yussof, Masahiro Ohka
Graduate School of Information Science
Furo-cho Chikusa-ku, Nagoya, Aichi, Japan
firstname.lastname@example.org / email@example.com
Abdul Rahman Omar, Muhammad Azmi Ayub
Faculty of Mechanical Engineering
University Teknologi MARA
Shah Alam, Selangor, Malaysia
Abstract—This paper presents experimental results to define
suitable parameters in object stiffness control using a
prototype optical three-axis tactile sensor mounted on robotic
fingers. We have developed a novel optical three-axis tactile
sensor system based on an optical waveguide transduction
method applying image processing technique. We conducted a
series of calibration experiments with soft and hard objects to
define suitable parameters in object stiffness control. We
analyzed normal and shearing forces data detected in the
experiments and compiled suitable parameters in an algorithm
inside the robot control system. Verification experiment using
robotic fingers to manipulate soft object was conducted whose
result revealed that the finger’s system managed to recognize
the stiffness and safely manipulate the object.
The ability to sense hardness and/or softness will be
particularly important in
developmental robots that apply tactile sensing. To
successfully manipulate objects in the real world, robot
systems require some form of tactile feedback to distinguish
the object’s stiffness. Unfortunately, so far, no specific
parameters in developmental robots can measure the
hardness sensation. In robot manipulation, researchers are
addressing this problem through a novel sensor design
considering intelligent object exploration algorithms.
However, it seems not enough since so far manipulation
tasks are mostly demonstrated using solid and hard objects,
and questions of low force interaction and stiffness
distinction for the safe manipulation of soft and fragile
objects have still not been fully researched.
future applications of
In this research, we aimed to clarify control parameters
involved in stiffness control of robot manipulation based on
tactile sensing. We developed a prototype optical three-axis
tactile sensor capable of acquiring normal and shearing
forces to mount on the fingertips of robotic fingers. We used
this system to perform calibration experiments and analysis.
Figure 1. Optical three-axis tactile sensor mounted on robotic fingers.
OPTICAL THREE-AXIS TACTILE SENSOR
A tactile sensor is a device that can measure a given
property of an object or contact event through physical
contact between the sensor and the object. Tactile sensors
have been developed using measurements of strain produced
in sensing materials that are detected using physical
quantities such as electric resistance and capacity, magnetic
intensity, voltage and light intensity . Research on tactile
sensor is basically motivated by the tactile sensing system
of the human skin. In humans, the skin’s structure provides
a mechanism to simultaneously sense static and dynamic
pressure with extremely high accuracy. Meanwhile in
robotics, several tactile sensing principles are commonly
used, such as capacitive, piezoelectrical, inductive,
piezoresistive, and optoelectrical sensors .
In this research, to establish object manipulation ability
in a real humanoid robot, we developed an optical three-axis
tactile sensor capable of acquiring normal and shearing
forces. This tactile sensor uses an optical waveguide
transduction method and applies image processing
techniques . This type of sensing principle is
comparatively provides better sensing accuracy to detect
contact phenomena from acquisition of three axial directions
of forces. Fig. 1 shows the optical three-axis tactile sensor
mounted on robotic fingers.
1-4244-2581-5/08/$20.00 ©2008 IEEE992IEEE SENSORS 2008 Conference
Figure 2. Structure of optical three-axis tactile sensor.
Figure 3. Image taken by CCD camera inside the optical three-axis tactile
sensor, and configuration of 41 sub-regions of sensing elements.
The optical three-axis tactile sensor developed in this
research is designed in a hemispherical dome shape that
consists of an array of sensing elements. This shape is to
mimics the structure of human fingertips for easy
compliance with various
miniaturization of the tactile sensor, measurement devices
are placed outside the sensor. The hardware novelty consists
of an acrylic hemispherical dome, an array of 41 pieces of
sensing elements made from silicon rubber, a light source,
an optical fiber-scope, and a CCD camera, as shown in Fig.
2. The optical fiber-scope is connected to the CCD camera
to acquire image of sensing elements touching acrylic dome
inside the tactile sensor. Meanwhile, the silicone rubber
sensing element is comprised of one columnar feeler and
eight conical feelers which remain in contact with the
acrylic surface while the tip of the columnar feeler touches
an object. The sensing elements are arranged on the
hemispherical acrylic dome in a concentric configuration
with 41 sub-regions as shown in Fig. 3. Such orientation is
expected to provide good indication of contact pressure.
shapes of objects. For
When an object contacts the columnar feelers, resulting
in contact pressure, the feelers collapse. At the points where
the conical feelers collapse, light is diffusely reflected out of
the reverse surface of the acrylic surface because the rubber
has a higher reflective index. Contact phenomena consisting
of bright spots caused by the collapse of the feelers are
observed as image data, which are retrieved by the optical
fiber-scope connected to the CCD camera and transmitted to
computer. Sensing program inside the computer is using
Visual C++ and we utilized image analysis software
Cosmos32 to analyze and measure the image data.
It is important to define object stiffness control to
enhance performance of the tactile sensor system, and to
improve its robustness during real-time object manipulation.
In robotic finger’s control system, normal and shearing
forces, and slippage sensation detected by the tactile sensor
are used to measure stiffness of the manipulated object.
Consequently, suitable control parameters are required to
specify so that robotic fingers can respond correctly to any
objects with different stiffness conditions. However, serious
conflicts are remained to specify such parameters, especially
during contact point changes. To solve this problem, at first
the finger must perform a soft touch on the object’s surface
and detect forces that occurred during the soft contact event.
At this moment, the detected forces are definitely low and
difficult to measure. However, this low force must be
utilized to distinguish the object’s stiffness so that the finger
control system can conduct grasping motions without
crushing the object or damaging the sensor elements.
Therefore, the tactile sensor must be highly sensitive enough
to detect a very low force. Furthermore, the sensor system
must be able to detect not only normal force but also
shearing force so that a slippage sensation that may occur
during soft contact can be detected.
METHODOLOGY OF OBJECT STIFFNESS CONTROL
The optical three-axis tactile sensor used in this research
is capable to satisfy the above requirements because the
sensing principle, which utilized an optical waveguide
transduction method, permits highly sensitive force
detection from the acquisition of the three axial directions of
forces; thus normal force and shearing force can be
measured simultaneously with high accuracy . In the
current object manipulation scheme, motion planning is
divided into two modes: grasping and moving. In the
grasping mode, both fingers move slowly to grasp the object
to define the optimum grasp pressure, while controlling the
pushing velocity of both fingers to grip the object. When the
optimum grasp pressure is defined, both fingers are
automatically shifted to the moving mode and together
manipulate the object.
IV. DETERMINATION OF IMPORTANT PARAMETERS
A. Calibration Experiment
We conducted a series of calibration experiments with
soft and hard objects using the multi-fingered robotic
system to determine the important parameters in object
stiffness control. The hard object was an aluminum block,
and the soft object was a paper box, as shown in Fig. 4. In
this experiment, both fingers move along the x-axis to softly
grip the object and define the optimum grasp pressure for
the grasping mode. Then both fingers lift up the object along
the z-axis in the moving mode.
In experiment with hard object, the reaction force
applied toward the tactile sensor elements was large because
the elasticity coefficient for the hard object is high.
Therefore, the detected normal force becomes high. On the
other hand, the object’s weight caused great slippage.
Figure 4. Calibration experiments with aluminium block and paper box.
Figure 5. Relationship between normal (top) and shearing (bottom) forces
with fingertip position at x-axis for experiment with aluminum block.
For soft object, small reaction force is applied to the
sensing elements because the elasticity coefficient for soft
objects is low. Accordingly, the detected normal force
becomes low. Therefore, to correlate the stiffness distinction
of both hard and soft objects, we utilized the increment of
normal force ΔF, which was calculated within a specified
progress time, as a stiffness distinction parameters.
To comply with the slippage that particularly occurred
for hard object; we considered the amount of centroid
change dr for x-directional (dxG) and y-directional (dyG) of
the fingertip coordinate frame, by means of shearing force
distribution. If slippage is over the dr value, the finger re-
pushes toward the object to prevent it from slipping.
However, if the detected ΔF was lower than a specified
value (i.e., a soft object), the finger system uses the dr value
Figure 6. Relationship between normal (top) and shearing (bottom) forces
with fingertip position at z-axis for experiment with paper box.
to control the finger’s re-push velocity so that the grasping
motion becomes gentler and finally stops when the centroid
change is over a specified dr value. On the other hand, the
fingertips movements are basically controlled by the
thresholds of normal force F1 and F2. If the normal force is
over the F1 value, both fingers will not further re-push
toward the object. F2 is used for emergency stops in case of
over push towards the object occurred.
B. Experimental Result
Based on the above control scheme, the results of
calibration experiments were compiled in graphs and
analyzed to define suitable parameters values. For example,
Fig. 5 shows the relationship between normal and shearing
forces with fingertips movements at x-axis during
experiment with aluminum block for left finger. Meanwhile,
for experiment with paper box, the relationship between
normal and shearing forces with fingertips movements at z-
axis for right finger are shown in Fig. 6.
To define increment of normal force ΔF, we measured
amount of maximum normal force increments in specific
progress time as shown in Fig. 7. For reference, we also
conducted experiment using styrofoam as shown in this
figure. These increments of normal force values are used to
Progress ti m e from touchi ng, Δt
Progress time from touching, Δt
Amount of normal force change, Δ
Amount of maximum normal force change, ΔFmax N
Figure 7. Amount of maximum normal force increments in specific
Figure 8. Verification experiment of robotic fingers manipulate paper box.
specify object’s stiffness distinction to control fingertips
movements during object manipulation tasks. After
analyzing the calibration experiment results, we determined
the parameter values as shown in Table 1. These control
parameters enable the finger system to realize object’s
stiffness, even when the detected forces are very low, and
then adjust grasp pressure to manipulate the object. These
parameters are applied in the finger’s control algorithm .
We conduct verification experiments based on the
results of calibration experiment as shown in Table 1. In the
control algorithm, at first the control system sees the
threshold of centroid change dr before performing stiffness
distinction using the increment of normal force ΔF. Then
the fingers reinforce the grasping pressure to re-push the
object according to the velocity ratio. Regarding the
threshold of normal force, if the object was detected as soft
object, the finger will no longer re-push the object when the
normal force detected exceeds F1. Meanwhile, F2 is used for
emergency case so that the finger will not over-push the
object, especially when handling hard object. Fig. 8 show
photographs of robotic fingers manipulate paper box. In this
experiment, both fingers move along x-axis direction to
grasp the paper box. The robot recognized stiffness of the
object, and then both fingers lift up the paper box along y-
axis. Then the fingers perform twisting motion. The fingers
managed to manipulate the paper box without crushing it.
PARAMETERS OF STIFFNESS CONTROL
We have presented experimental results to define
suitable parameters in object stiffness control using a
prototype optical three-axis tactile sensor mounted on
robotic fingers. We analyzed normal and shearing forces
data detected in the experiments. To correlate the normal
force characteristics of soft and hard objects, we measured
the increment of maximum normal force in specific progress
time to classify the stiffness of objects. Meanwhile, the
shearing force is utilized to define re-pushes velocity of the
robot fingers when grasping the object. Verification
experiment using robotic fingers to manipulate soft object
was conducted whose result revealed that the finger’s
system managed to recognize the stiffness and safely
manipulate the object.
This research is partially supported by a fiscal 2006
Grant-in-Aid for Scientific Research in Exploratory
Research from the Japan Ministry of Education, Culture,
Sports, Science and Technology (Grant no. 18656079), and
Grant-in-Research by Japan Society for the Promotion of
Science (JSPS) for fiscal year 2008-2010.
 H. R. Nicholls, Tactile sensing using an optical transduction method,
Traditional and Non-Traditional Robot Sensors (Edited by T. C.
Henderson), Springer-Verlag, 1990, pp. 83-99.
 P. A. Schmidt, E. Mael and R. P. Wurtz, “A sensor for dynamic
tactile information with applications in human-robot interaction and
object exploration,” Journal Robotics & Autonomous Systems, vol.
54, issue 12, Dec. 2006, pp. 1005-1014.
 M. H. Lee and H. R. Nicholls, “Tactile sensing for mechatronics – a
state of the art survey,” Journal Mechatronics, vol. 9, 1999, pp. 1-31.
 M. Ohka, Y. Mitsuya, Y. Matsunaga and S. Takeuchi, “Sensing
characteristics of an optical three-axis tactile sensor under combined
loading,” Robotica, vol. 22, 2004, pp. 213-221.
 H. Yussof, M. Ohka, J. Takata, Y.Nasu and M. Yamano, “Low force
control scheme for object hardness distinction in robot manipulation
based on tactile sensing,” In proc. of ICRA2008, 2008, pp. 3443-3448.
 H. Yussof, M. Ohka, H. Suzuki and N. Morisawa, “Tactile Sensing-
Based Algorithm for Real-Time Grasp Synthesis in Object
Manipulation Tasks of Humanoid Robot Fingers,” In proc. of RO-
Sensor 100 ms
Finger 25 ms
Threshold of normal force
F1 0.5 N
F2 1.8 N
Threshold of shearing force dr 0.004 mm
Velocity of re-push vp 2 mm/s
Increment of normal force soft< 0.08 N <hard
Progress time Δt 0.1 s