National Conference on Technological Advancements in Mechanical Engineering ISBN : 978-93-85100-57-4

Modeling and Simulation of 7-dof Robotic Manipulator

Md Nizamuddin Ahmed(M.E)1, K.Veladri2
1
Department Mechanical Engineering,Vasavi College of Engineering, (Autonomous)Hyderabad, India
2
Associate Professor MED, Vasavi College of Engineering(Autonomous)Hyderabad, India
Email: nizamuddin8019@gmail.com1

Abstract the robotic arm so it can acquire the desired position in its
In this work, a 7 degrees of freedom redundant manipulator work space. The arm is then fitted with a wrist having three
is modeled and simulated. The CAD model of the robot is joints to acquire the desired orientation. Such manipulators
developed in Siemens UG-NX and simulation of the manipulator are kinematically simple to design and solve, but they lead to
is performed in Virtual Robot Experimentation Platform (V- two fundamental problems, singularity and inability to avoid
REP) software. The inverse kinematic model is developed and obstacles. A manipulator is termed kinematically redundant if
is tested for some typical applications like pick-and-place, peg the number of degrees of freedom (DOF's) is higher than the
in hole and welding operations. number of task space coordinates.
Keywords-V-rep; simulation; robot; redundant manipulator; B. Inverse Kinematic Analysis
performance; Inverse Kinematics.
An Inverse kinematics problem determines the possible
I. INTRODUCTION
sets of joint variables for any specified end effector location.
The inverse kinematics problem for redundant manipulator This problem of inverse kinematic solution is generally more
has been solved by Lorenzo Sciavicco [1]. The process difficult than the forward kinematic problem. The inverse
undertaken in the design and development of an 8-DOF kinematic solution is more important because manipulation
humanoid robotic arm using specifically made parts has been tasks are naturally formulated in terms of desired tool positions
presented by Le Bang Duc [2]. Localization and navigation, and orientation of the objects to be manipulated. Forward
manipulation, human-robot communication and interaction, kinematics problem always has a unique solution that can be
adaptability and learning are presented by Rainer Bischoff [3]. obtained by simple evaluation of forward equations; the
The Introduction to the pseudo inverse method and damped inverse kinematics problem may or may not have a solution.
least squares method for inverse Kinematics are presented by
C. Damped Least Squares Method
Samuel R.Buss and J.S.Kim [4]. The reformulated inverse
kinematic problem as a damped least squares problem to The damped least squares method avoids many of the
overcome the difficulties encountered near kinematic pseudo inverse method's problems with singularities and can
singularities is presented by Charles W.Wampler [5]. The give a numerically stable method of selecting . It is also called
inverse kinematics of 7-DOF manipulator is done using Closed the Levenberg-Marquardt method. Rather than just finding
Loop Inverse Kinematics, redundancy resolution at the the minimum vector that gives a best solution to e = J†θ ∆
acceleration and velocity level is presented by Jingguo Wang , we find the value of that minimizes the quantity
[6]. A strategy for bipedal robot walking on inclined surfaces
|| J θ ∆ - e ||2 + λ 2 || θ ∆ ||2 (1)
using position and orientation based inverse kinematics
algorithm is introduced by Fariz Ali [7]. Eric Rohmer [8] whereλΕ R is a non-zero damping constant. This is
discussed the utility of a portable and flexible simulation equivalent to minimizing the quantity
framework that allows for direct incorporation of various
control techniques. A versatile, scalable, powerful general (2)
purpose robot simulation framework called V-REP is presented
in this paper. The corresponding normal equation is
A. Redundant Manipulators T
∆θ = T
(3)
A manipulator is required to have a minimum of six degree
of freedom if it needs to acquire any random position and This can be equivalently rewritten as
orientation in its operational space or work space. Assuming
that one joint is required for each degree of freedom, then (JTJ +λ2I)∆θ = JTe (4)
such a manipulator must be composed of minimum of six joints. It can be shown that J J +λ I is non-singular. Thus, the damped
T 2

In standard practice three degree of freedom is implemented in least squares solution is equal to

22nd & 23rd July 2016, University College of Engineering Kakinada(A) JNTUK Kakinada A.P. India. 303
National Conference on Technological Advancements in Mechanical Engineering

= (JTJ + λ2I)-1 JTe (5) purpose robot simulator developed by Coppelia Robotics. The
robot simulator V-REP, with integrated development
Now J J is an n x n matrix, where n is the number of degrees
T
environment, is based on a distributed control architecture:
of freedom. It is easy to show that
each object/model can be individually controlled via an
(JTJ + λ2I)-1 JT = JT(JJT + λ 2I)-1. Thus, embedded script, a plugin, a ROS node, a remote API client, or
∆θ= JT (JJT + λ2I)-1 e (6) a custom solution. Controllers can be written in C/C++, Python,
Java, Lua, Matlab, Octave or Urbi. Lua is a powerful, fast,
The advantage of this equation with the previous one is lightweight, embeddable scripting language designed to
that the matrix being inverted is only m x m where m=3k is the support procedural programming. The Lua script interpreter is
dimension of the space of target positions, and m is often embedded in V-REP, and extended with several hundreds of V-
much less than n. Additionally eq.(6) can be computed without REP specific commands. Scripts in V-REP are the main control
needing to carry out the matrix inversion, instead row mechanism for a simulation. A simulation is handled when the
operations can find such that client application calls a main script, which in turn can call
(JJT + λ2I) = e and then JT f is the solution. child script. Child scripts are divided in two categories. They
can be either non-threaded or threaded child scripts. Non-
The damping constant depends on the details of the threaded child scripts are pass-through scripts. Which means
multibody and the target positions and must be chosen they return control to the caller directly after each simulation
carefully to make eq.(6) numerically stable. The damping pass they execute. Non-threaded child scripts are normally
constant should be large enough so that the solution for are called at each simulation step. Threaded child script are scripts
well-behaved near singularities, but if it is chosen too large, in which a call to a threaded child script will launch a new
then the convergence rate is too slow. There have been a thread, which will execute in parallel and then directly return.
number of methods proposed for selecting damping constants Both the non-threaded and threaded child scripts are
dynamically based on the configuration of the articulated segmented in three parts.
multibody.
1. The initialization part: Executed the first time the child
II. MODELLING THE REDUNDANT MANIPULATOR script is called.
The objective of this project is to model and analyze the 7- 2. The regular part: Executed at each simulation pass. This
DOF robot manipulator. In the first stages of design, rigid code is in charge of handling a specific part of a simulation.
links of the robot arm are modelled using solid modelling in
SIEMENS UG-NX 10. These links are imported to V-REP and 3. The restoration part: Executed one time just before a
assembled. simulation ends.
IV KINEMATIC SIMULATIONS
V-REP uses IK groups and IK elements to solve inverse
and forward kinematics tasks. An IK group contains one or
more IK elements:
IK groups: IK groups group one or more IK elements. To
solve the kinematics of a simple kinematic chain, one IK group
containing one IK element is needed. The IK group defines
the overall solving properties (like what solving algorithm to
use, etc.) for one or more IK elements.
IK elements: IK elements specify simple kinematic chains.
One IK element represents one kinematic chain. A kinematic
chain is a linkage containing at least one joint object.
The IK element is made up by: a base, several links, several
joints, a tip, a target.
Fig 1: Assembly of the manipulator in V-rep A. Robot Model in V-Rep
Figure 1 shows the final assembly of the manipulator in V- The Manipulator is modelled in a way that the end effector
rep. All the links are named in the above figure. is used to grasp the cube and can also hold arc welding holder.
III. INTRODUCTION TO VIRTUAL ROBOT This redundant robotic manipulator has 7-DOF in which 3
EXPERIMENTATION PLATFORM (V-REP) DOF are on the shoulder, 1 DOF on the elbow and 3 DOF on
Virtual Robot Experimentation Platform (V-Rep) is a general the wrist. Out of the 7 DOF, 6 joints are revolute, meaning they

304 ISBN : 978-93-85100-57-4

National Conference on Technological Advancements in Mechanical Engineering ISBN : 978-93-85100-57-4

can rotate around a single axis. The gripper at the end of arm 5. Adding new IK group and IK element.
can open and close linearly through a prismatic joint.
6. Configuring the base of the Kinematic chain.
B. Assembling the Robot
7. Selecting the Constraints.
Steps for assembling the robot in V-rep:
8. Choosing the Calculation method (DLS method).
1. Import STL files into V-rep main scene.
9. Defining a path for its travel.
2. Rename all the imported components.
10. Adding script.
3. The imported individual components are first placed
11. Simulating the scene.
correctly by using manipulation toolbar buttons (translate
and rotate).
4. Now add joints (revolute and prismatic) to each link and
position it accordingly.
C. Designing the Simulation Scene
The assembly contains robotic arm and a gripper. Figure 9
represents the structure of the V-rep scene hierarchy.
1. After assembling the parts, the next step is to connect the
links by the corresponding joints.
2. The model tree was adjusted to form the kinematic chain of
the manipulator.
3. The same steps are to be repeated for gripper also.
4. Add the script to the robot.
Fig 3: Linking dummy in Scene Object Properties

After Linking the dummies in the Scene Object Properties,

a red coloured dotted line will appear in the scene hierarchy
indicating the dummy object that have been linked.
The linked dummy in the scene hierarchy can be seen in
Fig 4.

Fig 2: Structure of the V-rep Scene hierarchy

D. Configuring the Inverse Kinematics Module

After the assembly of the robot and modelling the tree
structure in scene hierarchy, Inverse Kinematics module is to
be defined.
1. Enable the joints in IK mode.
2. Add target dummy and tip dummy to the scene.
3. Position and orient the dummies accordingly. Fig 4: Linked dummy objects
4. Specifying a target to follow (linking Dummy target with
tip).

22nd & 23rd July 2016, University College of Engineering Kakinada(A) JNTUK Kakinada A.P. India. 305
National Conference on Technological Advancements in Mechanical Engineering

Fig 6: Pick and place task: snapshots

Lua script is the program language used in V-REP. It is

used to program the robot for performing simulation. Figure.
14 is a control script for the robot to perform pick-and-place
Fig 5: Path planning dialog box
operation.
Path is calculated using Path planning module where
The respective results from the simulation are plotted which
Obstacles are taken as one entity using Collection module.
can be seen in Section-VI.
The path is calculated by pressing Compute path button and
the tip of the robot follows the computed path avoiding B. Peg-in-Hole
obstacles Peg-in-hole task is a process of grasping a cube through
• Revolute Joint 1 - Position is cyclic suction pad and placing it carefully in a target position which
is a hole of square cross section. The robot hand is moved
• Revolute Joint 2 - Position range: 20 degrees towards the grasp location and the suction pad is switched on
through script in a way that it can grasp the object. Then the
• Revolute Joint 3 - Position range: 200 degrees
robot gets in contact with the object. The suction applies as
• Revolute Joint 4 - Position range: 240 degrees much force as needed to be able to hold it. Now the object(cube)
is held by the gripper and it is moved towards the drop location.
• Revolute Joint 5 - Position is cyclic As the manipulator reaches the goal position, it places the
• Revolute Joint 6 - Position range: 40 degrees cube in a hole and suction ends and releases the object.

• Prismatic Joint - Position range: 0.1 metre

IV.SIMULATING THE SCENE FOR INDUSTRIAL
APPLICATIONS
The robotic arm was modelled with seven degrees of
freedom and script is written to accomplish pick-and-place,
peg-in-hole and welding tasks to assist in the production line
in any industry.
A. Pick-And-Place
Pick-and-place task is a process of grasping an object,
lifting it and dropping it at a target position. The robot hand is
Fig 7: Peg-in-hole task: snapshots
moved towards the grasp location and the gripper is kept open
in a way that it can grasp the object. Then the robot gets in The respective results from the simulation are plotted which
contact with the object. The gripper is closed and applies as can be seen in Section-VI.
much force as needed to be able to hold it. Now the object(cube) The Lua script used to program the robot for performing
is held by the gripper and it is moved towards the drop location. simulation is represented in Fig. 19. It is the control script for
As the manipulator reaches the goal position, the gripper opens the robot to perform welding operation.
and releases the object.
The respective results from the simulation are plotted which
can be seen in Section-VI.

306 ISBN : 978-93-85100-57-4

National Conference on Technological Advancements in Mechanical Engineering ISBN : 978-93-85100-57-4

C. Welding Task From Fig 10, the prismatic joint starts to close at 4.3 sec
from the start of simulation and it opens at 14.2 sec. The initial
The weld holder is attached to the robot arm and is allowed
joint position (open position) is -0.1m and while it grips the
to weld. The robotic Manipulator used has 7 degrees of
box, its position is -0.06m which is due to the size of the box.
freedom. The welding occurs when the electrode comes in
contact with the component.

Fig 11: velocities of revolute joints in degrees/sec

From Fig 11, the velocities of all the joints plotted reveals
that the motion is smooth with no jerks during the operation.
The positive value in the velocity indicate that the joint motion
is clockwise and negative value indicate anticlockwise.
Fig 8: Welding : snapshots

VI. RESULTS AND DISCUSSION

A. Pick-And-Place

Fig 12: Prismatic joint velocity in meters/sec

From Fig 12, the prismatic joint velocity records highest

value at 4.35 sec which is 0.56 m/s. It remains in hold position
Fig 9: Joint positions of Revolute joints of manipulator in till it holds the block and the velocity becomes negative when
deg re es.
it comes back to initial position.
From Fig 9, it is evident that revolute1 joint position has
maximum rotation to about -100 degrees and minimum rotation
is for revolute2. It is because in pick and place, the box
orientation doesn't change and is simply translated from the
pick point to the goal point.

Fig 13: Prismatic joint acceleration in meters/sec2

Fig 10: Prismatic joint position in meters

22nd & 23rd July 2016, University College of Engineering Kakinada(A) JNTUK Kakinada A.P. India. 307
National Conference on Technological Advancements in Mechanical Engineering

Fig 17: Suctionpad link force in kg*m/sec2

Fig 14: prismatic joint force
From Fig 17, the suction pad link needs a force of 235 N to
Figure 14 shows the force exerted by the prismatic joint
hold the box of weight 150grams initially.
during the operation. The joint records a maximum force of 100
N in gripping the cube.
B. Peg-In-Hole

Fig 18: cube vs tip orientation

From Fig. 18, the tip orientation and cube orientation match
from 18.75 sec to 36 sec. This is due to the holding of the cube
Fig15: Joint positions of Revolute joints of manipulator in firmly by the suction pad without slipping it. After 36th sec,
de gr e e s
the cube is placed into the peg and orientation inline is lost
From Fig 15, the prismatic joint position is fixed and is zero between these two objects.
because it holds the suction pad firmly. Rest all joints have
movement while performing the operation.

Fig 19: Welding Tip absolute velocity meters/sec

Figure 19 shows the welding tip absolute velocity of the

welding tip of the robot during welding operation performed.
Fig 16: velocities of revolute joints in degrees/sec The X axis represents time in seconds and is plotted for 100
From Fig 16, the velocities of all the joints plotted reveals seconds which is the time taken for performing welding task.
that the motion is smooth with no jerks during the operation. VII. CONCLUSIONS
The positive value in the velocity indicates that the joint motion
The objective of this project is to model and analyze a 7-
is clockwise and negative value indicate anticlockwise
DOF robot redundant manipulator. Computer aided design and
engineering tools, which are proved to be valuable tools in
the field of robotics have been used for this purpose. Initially,
modeling of rigid links of the robot arm was carried out using
308 ISBN : 978-93-85100-57-4
National Conference on Technological Advancements in Mechanical Engineering ISBN : 978-93-85100-57-4

solid modeling techniques available through Siemens NX 10 7 . Fariz Ali, et al, "Bipedal robot walking strategy on inclined surfaces
package. These solid models are then imported in to V-REP using position and orientation based inverse kinematics algorithm",
Control Automation Robotics & Vision (ICARCV), 11th
(Virtual Robot Experimentation Platform). The simulation is International Conference, page-181-186, ISBN- 978-1-4244-7814-
carried out in order to check the performance of the design. 9.
The modelled manipulator is implemented for following 8 . Eric Rohmer, et al, "V-REP: a Versatile and Scalable Robot Simulation
applications: Framework", 2013, IEEE/RSJ International Conference on
Intelligent Robots and Systems, Date of Conference-3-7 Nov. 2013
• Pick-and-Place: Picking a cube that is moving on the Page-1321-1326 ISSN-2153-0858.
conveyor belt and placing it on the desired target location.
9 . Adria Colome and Carme Torras, "Redundant Inverse Kinematics:
• Peg-in-hole: Picking a cube through suction pad and Experimental Comparative Review and Two Enhancements", IEEE/
placing in a square hole. RSJ International Conference on Intelligent Robots and Systems,
2012, Page-5333-5340, ISSN-2153-0858.
• Welding: Performing arc welding on a component 10 . Le Minh Phuoc, et al, "Damped least square based genetic algorithm
with Gaussian distribution of damping factor for singularity-robust
The robot can avoid obstacles with the extra degree of inverse kinematics", Journal of Mechanical Science and Technology,
freedom and considering of joint limits. The pay load that the 2008, ISSN-1330~1338.
robot can carry is 10 kg in the case of pick and place task and
11. Haslina Arshad, et al, "Teaching Robot Kinematics in Virtual
1.5 kg in the case of grasping of falling object. Environment", 2010, Proceedings of World Conference on
Engineering & Computer Science, vol-1, ISBN: 978-988-17012-0-
In all these applications, the simulation has been
6.
successfully carried out. The results clearly show that the
model performs all operations effectively without any sudden 12 . Wong Guan Hao, et al, "6-DOF PC-Based Robotic arm With Efficient
Trajectory Planning and Speed Control", 2011, 4th International
or unaccounted movements. Conference on Mechatronics, pp-17-19 May, Kuala Lumpur,
Malaysia.
ACKNOWLEDGMENT
13 . Donghun Lee, et al, "Development and application of an intelligent
The authors would like to thank the management of Vasavi welding robot system for shipbuilding", Robotics and Computer
College of Engineering, Hyderabad for extending support and Integrated Manufacturing, Elsevier. pp-377-388.
resources towards this work. 14 . Stefano Baglioni, et al, "Multibody modelling of N DOF robot arm
assigned to milling manufacturing. Dynamic analysis and position
REFERENCES errors evaluation", Journal of Mechanical Science and Technology,
1 . Lorenzo Sciavicco & Bruno Siciliano, "A solution Algorithm to the 2016, Springer, page-405-420, ISSN-405-420.
inverse kinematic Problem for redundant Manipulators", IEEE
15 . D. Prattichizzo, J.C. Trinkle, "Grasping", in Springer Handbook of
Journal of robotics and Automation Vol-4, No-4.
Robotics, 2008, Springer, Heidelberg(Eds), pp. 671-700.
2 . Le Bang Duc, et al, "Designing 8-dof Humanoid Robotic Arm", 16. K.Karteek Reddy, K.Praveen, "Path Planning using V-rep",
International Conference on intelligent and Advanced Systems, International Journal of Research in Engineering and Technology,
pp-1 069-10 74. Vol- 02, Issue- 09 Sep-2013, eISSN: 2319-1163.
3 . Rainer Bischoff, Volker Graefe, "HERMES- a Versatile Personal 17 . Marc Freese, et al, "Virtual Robot Experimentation Platform V-
Robotic Assistant". Proceedings of IEEE, Human Interactive robots, REP: A Versatile 3D Robot Simulator", 2010, Simulation, Modelling
pp-1759-1779, ISSN-0018-9219. and Programming for Autonomous Robots - Second International
4 . Samuel R. Buss, "Introduction to inverse kinematics with jacobian Conference, pp-51-62, SIMPAR.
transpose, pseudoinverse and damped least squares methods", IEEE
18. V-REP simulator : http://www.coppeliarobotics.com
Journal of Robotics and Automation, Vol-3, page-681-685.
19. Robert J. Schilling. "Fundamentals of Robotics, Analysis and control".
5 . Charles W. Wampler, "Manipulator Inverse Kinematic Solutions PHI publications, 2000, ISBN-81-203-1047-0.
Based on Vector Formulations and Damped Least-Squares Methods",
IEEE Transactions on systems, man and cybernetics, Vol-16 page- 20. Mark W. Spong, M. Vidyasagar, "Robot Dynamics and Control",
93 . John Wiley & sons (ASIA) pte Ltd, 2004, ISBN-9812-53-078-9.
6 . Jingguo Wang, "International Journal of Advanced Robotic 21. R K Mittal, I J Nagrath, "Robotics and Control". Tata McGraw Hill
Systems", Vol- 7, No. 4, pp.1?10, ISSN: 1729?8806. Publications, 2005, ISBN-0-07-048293-4

22nd & 23rd July 2016, University College of Engineering Kakinada(A) JNTUK Kakinada A.P. India. 309