Proceedings of the 2018 13th World

Congress on Intelligent Control and Automation

July 4-8, 2018, Changsha, China

Development of Robot Gripper and Force Control

Anan Suebsomran, Member, IEEE

Abstract— Generally, robotic gripper enhances the efficiency simulations in [3]. In other views, grip force and slippage
and the ability to grasp and to control robot’s manipulator controller for robotic object manipulation based on
rapidly to objects in different shapes and dimensions. In this mechanical friction model are presented in [4] and [5]. Their
paper, a new design to control robotic gripper is proposed by approaches allow a load to be held without excessive force
controlling based on grasping force method. The gripper is for a variety of weighting objects. However, the main
designed with two jaws driven symmetrically by the actuator drawback of those gripper designs is that, it is detected the
to grasp and to release the object. Basically, its force is slippage in only one axis. Subsequently, other slippage axes
directly measured by force sensing resistance (FSR) type. mostly extended to gain such information to fulfill those
Force controller is designed by using PID control algorithm drawbacks. Moreover, the intelligent robotic griper is
with different control gains and objects tuning by experiment designed and implemented to grasp unknown different
methods. Results are shown by experiment by using GUI objects with masses, dimensions, and friction coefficient
system. Responses of force controller in different grasping using fuzzy control during handling object shown in [2] and
objects are identical. In addition, the virtual motion of robot [6]. Also, an interesting adaptive neuro-fuzzy inference
gripper can be displayed on computer. So this work will lead system designed for relation of input and output for force
to improve the grasping force controller design and to control in order to overcome the slipping of object, are
enhance working performance in remote environment during proposed in [2], [7], and [8]. Lastly, grip force control based
handling operation of robot gripper. on the degree of slippage using optical tactile sensor is
surveyed showing in [9]. Eventually from above reviewing
I. INTRODUCTION research literatures, there is a promising room on a new
design and development of force control for robot gripper.
Recently, a number of applications of robotic and automation Due to different objects can be grasped with different gains
for industrial sectors have been increasing rapidly. A of force controller. The knowledge of control gains could be
numerous industries apply robotics and automation for obtained by experimental method. In addition, the robotic
manufacturing shop floor such as automotive, electronics and gripper can display kinematic motion of grasping and
hard disk drive, electrical appliance, etc. Task allocation in releasing jaw mechanism of its gripper. Therefore, it is
production line is ensemble to a variety of each work focused as this research target.
stations. Operation of some task is dependent to specific job
at shop floor control. A simple task of robot and automation II. DESIGN STAGE
such as on and off event driving is carried out by machine
without human interaction with highly dexterous automation A. Design of Robotic Gripper
manufacturing system. Other simple tasks are picking, Basically, the weight and dimension of the grasping objects
placing, loading and unloading, material transfer line system must be known roughly prior to system building as the
and other simple operations that operated by automatic important parameters. Thus, different objects or those
machine or robot. The higher sophisticated tasks are more parameters are chosen to validate the efficiency of the
carefully designed with a certain high level of automation proposed robotic gripper. They are kitchen egg, plastic
system. Those complex tasks are in assembling process such bottle, and metal load. The main reason of choosing such
as pegging, screwing, tightening and grasping, they are all the sensitive to hard through flexible objects, is that the privilege
challenge of highly automation level. to designing a complexity of robot gripper and to be
Previously, related researches in literature such as the sliding customized through the design of force controller. With
mode control of shape memory alloys actuated gripper are choosing such objects, the types of robot gripper are
presented in [1] and [2]. Those proposed researches show the consecutively selected for holding and grasping objects
significant advantage of lower weight and smaller size of without damage as the aim. Base on various kinds of robot
robot gripper, but its accuracy of position control has still to gripper those found, two fingers robot gripper types are
be developed further. Next, parallel proportional and integral chosen in this work. Under grasping and holding without
(PI) force controllers are applied to predict the contact force damage situation for all of those objects, it is the challenge
for robotic manipulation of deformable objects. The efficacy condition to build up satisfied robot gripper accordingly.
of those parallel (PI) force controllers is demonstrated by Principally, general two fingers gripper is firstly depicted in
its structure as shown in Fig. 1. It consists of the main
*Research supported by Contract no. KMUTNB-GOV-59-14.
A. Suebsomran is with the King Mongkut’s University of Technology
components including parallel gripper jaw, linkage
North Bangkok, Bangkok, 1080 THAILAND (corresponding author to mechanism and transmission system, and actuator. At the
provide phone: +66 92 249-3070; fax: +66 2 587-8261; e-mail: asr@ jaw, it is implemented with a force sensor for measuring the
kmutnb.ac.th). contact force between object and force sensor. It is the force

978-1-5386-7346-1/18/$31.00 © 2018 IEEE 433

Authorized licensed use limited to: University of Melbourne. Downloaded on December 14,2024 at 09:48:06 UTC from IEEE Xplore. Restrictions apply.
sensing resistance (FSR) type) which its resistance will jaw for grasping and releasing object.
change to voltage output across a bridge circuit. Its basic of
output signal from system calibration can be compared to the Computer

calculation as shown in Table 1 in order to initial the system. Gripper

motion Microcontroller
Next, the material of gripper structure is industrial graded display
Robot
ߑ
aluminum. Then, a DC motor is applied to the actuator for PID controller Driver DC motor gripper

driving linkage and transmission mechanism for grasping and Force

command

releasing objects. It is noted that diameter of those three Proximity sensor

objects is limited at 38 mm and their weighs must not heavier

than 5kg.
In software development part, graphic user interface is Force sensor

conducted by using MATLAB GUIDE development. The

control algorithm is implemented by using ARM7
Fig. 3. Force controller of robotic gripper work flow
microcontroller to control robot gripper with desired force
response. The more applicable creative of this design and
development is a motion monitoring on computer system. Let’s start with PID controller that can be implemented by (1)
Then, SolidWork software with soft motion module is also or (2).
applied for visualizing the motion asynchronous with the real t
gripper motion during its operation as shown in Fig. 2. de(t )
u (t ) = K p e(t ) + K i ³ e(t )dt + K d (1)
0
dt
or in another form as:
i
Δe
u (t ) = K p e(t ) + K i ¦ eΔ(t ) + K d (2)
i =0 Δt
Next, error of force control can be obtained by (3):

e f = Fref − F (3)

Figure 2. The proposed robot gripper where e f is the error of force controller; Fref is a reference
signal (N); and F is a force feedback signal (N).
Calculation of grasping force to grasp objects is expressed as:
Robot gripper simulation design
F = mg (4)

where F is the force (N); m is the mass (kg); and g is the

2
gravity constant (g = 9.81 m / s ).
From those three types of grasping objects of kitchen egg,
plastic bottle and metal load with different dimensions and
weights, grasping force is simply calculated by using this
equation (4) to apply in the design with that calibration step
as shown in Table I.
Table I. Testing result of force sensing for
calibration

Weight of objects of Calculated force (N)

force sensing
Figure 2. Simulation design measurement (g)
0 0
B. Controller Design 100 0.981
Designing and implementing force controller is illustrated 200 1.962
with a work flow as shown in Fig. 3. Basically, the feedback 300 2.943
force is measured by using force sensing resistance (FSR) 400 3.924
type for actually force measurement. Proportional-integral-
500 4.905
derivative (PID) controller is applied to control grasping
force. It is calculated by equation (1) to (3). A proximity
sensor is applied in order to detect location of the gripper at
the initial position, and a DC motor is applied to drive the

434
Authorized licensed use limited to: University of Melbourne. Downloaded on December 14,2024 at 09:48:06 UTC from IEEE Xplore. Restrictions apply.
 III. EXPERIMENT
There are three experiments conducting to measure the
grasping force of robotic gripper when grasping the different
objects. Those three kinds of grasping objects of kitchen egg,
plastic bottle and weighting metal are used accordingly under
the limitation of mentioned sizes and weights.

Fig. 5. Force response of grasping kitchen egg with

applied force 8N

Figure 2. GUI design interface

A. Experiment No.1 (50g kitchen egg, 43 mm-diameter,

and 69 mm-height)
a) It is applied with 4N grasping force. In this
experiment, we specify the grasping force reference
is 4N for the force command to microcontroller. The
Fig. 6. Experimental testing of grasping kitchen egg
PID controller requires user to set control gain: Kp
is10, Ki is 5, and Kd is 20. Its goal is to move two B. Experiment No.2(100g plastic bottle, 53mm-diameter,
fingers to grasp the object with the grasping force and 74 mm-height)
reaching to the force reference value. The responses
a) This first case is applied with 2N grasping force. The
are measured showing performance of controller of
PID controller requires to user to set control gain:
time response (Tresp) is 16ms; rising time (Tr) is
Kp, Ki, and Kd at 10, 1, and 4 respectively. The
12ms; delay time (Td) is 8ms, and settling time (Ts)
measured responses are with the time response
is 60 ms. The absolute error of force controlling is
(Tresp) at 19 ms; rising time (Tr) at 17 ms; delay time
0.6 N as illustrated in Fig. 4.
(Td) at 3 ms; settling time (Ts) at 18 ms, and
b) Latter case, 8N grasping force is applied. The PID
overshooting of 33.33%. The absolute error of force
controller requires user to set control gain: Kp, Ki, and
controlling is 0.2 N as illustrated in Fig. 7.
Kd at 50, 10, and 30 respectively. The performance
b) Plastic bottle is applied with 4N grasping force.
shows; time response (Tresp) is 34 ms, rising time (Tr)
Control gains: Kp; Ki, and Kd are set at 10, 2, and 6
is 2 ms, delay time (Td) is 18 ms, and settling time
respectively. The absolute error of force control is
(Ts) is 31ms. The absolute error of force controlling is
0.4 N as illustrated in Fig. 8 (time response (Tresp) =
0.8 N as illustrated in Fig. 5. Its scenario of grasping
20 ms; rising time (Tr) = 10 ms; delay time (Td) = 8
is given in Fig 6.
ms; and settling time (Ts) = 24 ms).
c) 6N grasping force is used in this last case (with
control gain; Kp = 80; Ki = 15 and Kd = 30). The
absolute error of force controlling is 0.6 N as
illustrated in Fig. 9 (time response (Tresp) = 20 ms;
rising time (Tr ) = 10 ms; delay time (Td) = 7 ms;
and settling time (Ts) = 40 ms). Scenario of grasping
and its dimension are given in Fig 10.

Fig. 4. Force response of grasping kitchen egg

with applied force 4N

435
Authorized licensed use limited to: University of Melbourne. Downloaded on December 14,2024 at 09:48:06 UTC from IEEE Xplore. Restrictions apply.
 C. Experiment No.3 (500g metal load, 45mm-diameter,
69mm-height)
a) The first case is applied with 3N grasping force (with
PID control gains: Kp = 25, Ki = 5 and Kd = 10). The
absolute error of force controlling is 0.4 N as
illustrated in Fig. 11 (with time response (Tresp) = 36
ms; rising time (Tr) = 8 ms; delay time (Td) = 6 ms;
. settling time (Ts) = 90 ms).
b) Another, grasping force at 7N is applied with PID
control gains as: Kp = 70; Ki = 45 and Kd = 20. The
Fig. 7. Force response of grasping plastic bottle with absolute error of force controlling is 0.6 N as
2N applied force illustrated in Fig. 12 (with time response (Tresp) = 32
ms; rising time (Tr ) = 17 ms; delay time (Td) =
11ms, settling time (Ts) = 28 ms). Scenario of
grasping of this metal load is shown in Fig 13.

Fig. 8. Force response of grasping plastic bottle with

applied force 4 N

Fig. 11. Force response of grasping metal load with

applied force 3N

Fig. 9. Force response of grasping plastic bottle

with 6N applied force

Fig. 12. Force response of grasping metal weighting with

applied force 7 N

Fig. 10. Experimental testing for grasping plastic Fig. 13. Experimental testing for grasping metal load
bottle
In this experiment No3, five times of grasping to metal load
in order to investigate the error of force controlling and to
figure out the threshold values are described in Table II. It

436
Authorized licensed use limited to: University of Melbourne. Downloaded on December 14,2024 at 09:48:06 UTC from IEEE Xplore. Restrictions apply.
shows the error of force controller during experiment in case disturbance”, 2014 IEEE International Conference on Systems, Man
and Cybernetics, pp. 3193-3198.
of grasping a metal weighing object.
[9] M-J. E. Salami, N. Mir-Nassiri and S.N. Sidek, “Design of intelligent
Table II. Error of force controlling multi-finger gripper for a robotic ARM using a DSP-based fuzzy
Number of Experiments 1 2 3 4 5 controller”, Proceedings TENCON 2000, pp. 348 – 353 vol.3.

Error 0.3 0.4 0.2 0.6 0.4

IV. CONCLUSION
Robotic grippers attached with end effector of industrial
robot can be separately design as tool tip of industrial robot
to working in several tasks such as loading, unloading, pick
and place, gluing, welding, and other automation
applications. A new design of robotic gripper is proposed in
this research. Objects which are directly grasped by robotic
gripper maybe damaged by grasping operation and subjected
to contact force of a robotic gripper. Thus the prevention of
object damage is very critical design stage. Force controller
using Proportional – Integral - Derivative controller (PID) is
chosen in design of force control with different control gains
of different objects achieved by experiments. Responses of
force controller can be achieved and gains of force controller
of different objects can be identified in experiment methods.
Moreover, robotic gripper motion in the proposed design
system can be virtually displayed by computer during gripper
operation.
Acknowledgment
This work was supported by the King Mongkut's University of Technology
North Bangkok. Contract no. KMUTNB-GOV-59-14.

REFERENCES
[1] S Krishna Chaitanya and K Dhanalakshmi, “Control of Shape
Memory Alloy Actuated Gripper using Sliding Mode Control”, 2013
IEEE International Conference on Control Applications (CCA) Part of
2013 IEEE Multi-Conference on Systems and Control Hyderabad,
India, August 28-30, 2013, pp. 1247-1252
[2] Pavel Dzitac, and Abdul Md Mazid, “A method to control grip force
and slippage for robotic object grasping and manipulation”, 2012 20th
Mediterranean Conference on Control & Automation (MED),
Barcelona, Spain, July 3-6, 2012, pp. 116-121
[3] Xuelin Wang, Xinjian Fan, Yongguo Zhao, Shiguo Ma, “Parallel
Force Control for a Robot Gripper Based on Grey Prediction Models”,
2012 IEEE Power Engineering and Automation Conference (PEAM),
Wuhan, China, pp. 1 – 5
[4] A. M. Zaki , A. M. Soliman, A. M. Zaki , A. M. Soliman, “Design and
Implementation of Efficient Intelligent Robotic Gripper”, Proceedings
of the 2010 International Conference on Modelling, Identification and
Control, Okayama, Japan, July 17-19, 2010, pp. 710-716
[5] Aziza. M. Zaki and Ayman. M. Soliman, Osama. A. Mahgoub and
Abdellatif.M. El-Shafei, “High Performance Robotic Gripper Based on
Choice of Feedback Variables”, Proceedings of the 2010 IEEE
International Conference on Information and Automation June 20 - 23,
Harbin, China, pp. 54-59
[6] A. M. Soliman, A. M. Zaki, A.M. El-Shafei , O. A. Mahgoub, “A
Robotic Gripper Based on Advanced System Set-up and Fuzzy
Control Algorithm”, Proceedings of the IEEE International Conference
on Automation and Logistics, Shenyang, China August 2009, pp. 1-6
[7] N. Watanabe and G. Obinata, “Grip force control based on degree of
slippage using optical tactile sensor”, 2007.International Symposium
on Micro-Nano Mechatronics and Human Science, MHS '07, pp 466 –
471
[8] J.A. Dominguez-Lopez, R.M. Crowder, R.I Damper, and C.J. Harris,
“Adaptive neurofuzzy control of a robotic gripper with external

437
Authorized licensed use limited to: University of Melbourne. Downloaded on December 14,2024 at 09:48:06 UTC from IEEE Xplore. Restrictions apply.