22 Introduction to Robotics

2.1 ROBOT SUBSYSTEMS

As illustrated in Fig. 2.2, a robotic system generally consists of three subsystems,
namely, a motion subsystem, a recognition subsystem, and a control subsystem.
While some aspects of hardware implementations are explained in Chapter 13, their
functions are described below.

Fig. 2.2 Robot subsystems and their interactions

[Courtesy: Yoshikawa (1990)]

1. A Motion Subsystem The motion subsystem is the physical structure of

the robot that carries out desired motion similar to human arms, as illustrated in
Fig. 2.3.

Fig. 2.3 A robot manipulator and its equivalent human parts

 Industrial Robots and Their Applications 23

2. A Recognition Subsystem The recognition subsystem uses various sensors to

gather information about the robot itself and any object being acted upon, and about
the environment. Based on sensor data, it recognizes the robot’s state, the objects,
and the environment.
3. A Control Subsystem The control subsystem influences the robot’s motion to
achieve a given task using the information provided by the recognition subsystem.
It may be useful to point out here that a
What is PUMA? person with mechanical engineering background
PUMA stands for Programmable normally works on motion subsystem area,
Universal Manipulator for
whereas the people with computer science and
Assembly.
electrical engineering knowledge focus on the
recognition and control subsystems, respectively. This is due to the subjects taught in
their academic curricula. However, robotics is an interdisciplinary area and a
comprehensive knowledge of all three will certainly help to design and develop a
better robotic system. As a result, it is not
uncommon to see people crossing their Why Robotics is interdisciplinary?
boundaries of specialization. It is often seen Since it requires the knowledge of
that a mechanical engineering specialist Mechanical Engineering, Electrical
works on Artificial Intelligence (Recognition Engineering, Computer Science En-
gineering, and Information Technol-
Subsystem), while the one with electrical ogy.
engineering or computer science background
deals with dynamic simulation and design of
robots (motion subsystem).
2.1.1 Motion Subsystem
The elements of the motion subsystem are as follows:
1. Manipulator This is the physical structure, which moves around. It comprises
of links (also referred as bodies) and joints (also called kinematic pairs) normally
connected in series, as shown in Figs. 1.1,
1.3 and 2.1 or the PUMA robot as shown in Robot or Manipulator?
Fig. 2.3(a). Each link is either made of steel In this book, many times both the
or aluminum. Other materials can also be words are used interchangeably.
used depending on the requirements. The However, as explained in this chap-
ter, manipulator is the mechanical
joints are generally rotary or translatory
structure of the robot.
types. In the study of robotics and
mechanisms, these joints are referred to as
revolute and prismatic joints, as explained in Chapter 5. Whereas an example of a
revolute joint is the hinge of a door, a prismatic joint is the piston-cylinder arrangement
of an Internal Combustion (IC) engine used in automobiles.
Like a human arm, wrist, and hand arrangement of Fig. 2.3(b), a robot manipulator
has also three parts. The first two, i.e., the arm and wrist, are shown in Fig. 2.3(a),
respectively, whereas the third one, i.e., the hand, is shown in Fig. 2.4(a). More
gripper or end-effectors performing the task of a hand are explained in Chapter 3. The
function of an arm is to place an object in a certain position in the three-dimensional
Cartesian space, where the wrist orients it. For a typical six degrees-of-freedom
 24 Introduction to Robotics

(DOF) robot, as shown in Fig. 1.1, 1.3(a), 2.1 and 2.3(a), the first three links and
joints form the arm, and the last three mutually intersecting joints make the wrist.

Fig. 2.4 Robot hands

2. End-effector This is the part attached at the end of a robot manipulator. Hence,
the name follows. This is equivalent to the human hand. An end-effector could be a
mechanical hand that manipulates an object or holds it before they are moved by the
robot arm. Two typical hands are shown in Figs. 2.4(a-b). Figure 2.4(a) shows a
simple two-fingered gripper that holds simple objects, whereas a multi-fingered hand
shown in Fig. 2.4(b) can perform complex
tasks. More grippers of the type shown in End-effector or Hand?
Fig. 2.4(a) are explained in Chapter 3. Also, Mechanical hands (Fig. 2.4) are
basically the end-effectors of a robot.
the specialized tools like welding electrode,
End-effector is a more generic term
gas-cutting torch, painting brush, debarring used to specify what is attached at
tool, or grinding wheel attached to the end of the end of a robot manipulator. Other
a manipulator arm to perform specific tasks, examples of end-effectors are
are also considered end-effectors. welding gun, paintbrush, etc.

3. Actuator The actuators of a robot actually provide motion to the manipulator

links and the end-effecter. They are classified as pneumatic, hydraulic, or electric,
based on their principle of operation, which are explained in Chapter 3. Note here
that an electric motor, e.g., dc or ac, when coupled to motion transmission elements,
e.g., gears, etc., is called an actuator. However, a pneumatic or a hydraulic system
which can directly impart motions to the robot links and the end-effectors is called
an actuator, not motor.
4. Transmission As the term conveys, these elements transmit motion from
the electric motors and pneumatic/hydraulic actuators to the actual links of the
manipulator. With electric motors these elements, mainly, the gears, are used to step
down the speed. Note that electric motors are efficient at higher speeds. However,
 Industrial Robots and Their Applications 25

the robot links move at relatively slow speed. Hence, the transmission gears are used
to reduce the speed of the electric motors. Typical transmission elements are the
following:
Belt and Chain Drives Belt drives are widely used in robotics, particularly, the
synchronous belt shown in Fig. 2.5(a). However, their life is short as they rely on belt
tension to produce grip over the pulley. Alternatively, chains shown in Fig. 2.5(b)
are generally cheaper. They have higher load capacities and service lives compared
to belt drives, but lower in relation to the gears.

Fig. 2.5 Belt and chain drives

Gears Of all mechanical transmissions, gears What is a mechanism?

shown in Fig. 2.6 are the most long-lasting and A series of links, generally rigid,
reliable, although their backlash must be coupled by joints that allow relative
carefully taken into account during the design motions between any two links
stage. form a mechanism.

Fig. 2.6 Gears

Link Mechanisms In order to reduce the flexibility and weight of the above
transmission elements, link mechanisms shown in Fig. 2.7(a) are used.
Screw Mechanism Figure 2.7(b) shows how ball-screws with a four-bar mechanism
(indicated with 1, 2, 3 and 4) bars are used to transmit motion.
 26 Introduction to Robotics

Fig. 2.7 Mechanisms and their use in a robot manipulator

Example 2.1 Gear Ratio

Assume a pair of spur gears shown in Fig. 2.6 has 20 and 100 teeth, respectively. If
the smaller gear (pinion) rotates at 200 rpm (revolutions per minute), the speed of the
larger gear or gear, denoted with wg, can be obtained as
wg = (1/h)wp = (20/100)200 = 40 rpm (2.1)
where h is the gear ratio defined as the ratio
between the number of teeth of pinion and gear,
Pinion vs. Gear?
respectively, i.e., h = ng/np—np and ng being the
In a pair of two gears, the smaller number of teeth in pinion and gear, respectively,
gear is called ‘pinion,’ whereas the
larger one is simply referred as only
whereas wp is the speed of the pinion. Such
‘gear’. gear transmission is called step-down gear
arrangement and specified as 1:5.

Example 2.2 Screw Transmission

Figure 2.7(b) shows a screw mechanism for the transmission of motion from the
motor to link 4 via link 3. If the nut on Link 3 has to be translated by 50 mm while
moving on the screw of 5 mm pitch, the screw should be turned by q times. The value
of q can be obtained from the following formula:
x = lq, where l = ap (2.2)
In Eq. (2.2), x is the displacement of the nut for q rotations of the screw, and l
is the lead of the screw which is equal to pitch p multiplied by the number of starts
a. Assuming, a = 1, i.e., single start screw, the number of times the screw needs to
rotate can be easily obtained from Eq. (2.2) as q = 10.

2.1.2 Recognition Subsystem

The most important element in the recognition subsystem is the sensor, which is
like our eyes or nose. Inclusion of sensors to a robot changes its dumb nature to
 Industrial Robots and Their Applications 27

an intelligent one. In order to process the sensed signal, which is mostly analog,
by a digital controller, an Analog-to-Digital Converter (ADC) is required. Some of
the aspects of signal conditioning will be elaborated in Chapter 4. A recognition
subsystem typically consists of the following elements.
1. Sensors Most of the sensors are essentially transducers. Transducers convert
one form of signal to another. For example, the human eye converts light patterns
into electrical signals. Sensors fall into one of the several general areas: vision, touch,
range and proximity detection, navigation, speech recognition, etc. Each of these
areas is an individual research area in itself. However, some of them are explained in
relative details in Chapter 4.
2. Analog-to-Digital Converter

[Courtesy: http://www.eeci.com/adc-16p.htm]
(ADC) This electronic device
interfaces the sensors with the robot’s
controller. For example, the ADC
converts the voltage signal due to the
strain in a strain gauge to a digital
signal, i.e., 0 or 1, so that the digital
robot controller can process these
information. They physically look
like any other computer interface card
inside the Central Processing Unit
(CPU) box, as depicted in Fig. 2.8.
Fig. 2.8 An analog-to-digital converter
2.1.3 Control Subsystem
The role of control in a robot is depicted in Fig. 2.9(a). It primarily consists of the
following items.
1. Digital Controller The digital controller is a special electronic device that
has a CPU, memory, and sometimes hard disk to store programmed data. In robotic
systems, these components are kept inside a box referred as controller, as shown in
Fig. 2.9(b). Figure 2.9(b) is the controller KR C2 for KUKA robots whose inside
is shown in Fig. 2.9(c). It is used to control the movement of the manipulator and
end-effecter. A robot controller is like the supervisor in a factory. Since a computer
has the same characteristics as those of a digital controller, it can be used as a robot
controller. A controller processes the user-programmed commands and sends signals
to the actuators through the Digital-to-Analog Converters (DAC). The programming
languages can be same as computers, i.e., BASIC, Fortran, C, and C++. However,
commercial robots use their domain-specific languages based on their manufacturers.
For example, KUKA, Germany, uses Kuka Robot Language (KRL), whereas Fanuc,
Japan, uses Karel robot programming language. This is mainly to introduce specific
features into the robotic systems so that the products are different.
2. Digital-to-Analog Converter (DAC)
ADC vs. DAC
A Digital-to-Analog Converter, or DAC, converts
the digital signal from the robot controller to A DAC serves the purpose opposite
to an ADC.
an analog signal to activate the actuators. In
 28 Introduction to Robotics

order to actually drive the actuators, e.g., a dc electric motor, the digital controller is
coupled with a DAC to convert its signal back to an equivalent analogue signal, e.g.,
the electric voltage for the dc motor. Physical appearance of a DAC is shown in Fig.
2.10.

Fig. 2.9 Control subsystem

[Courtesy: http://www.eeci.com]

Fig. 2.10 A digital-to-analog converter

 Industrial Robots and Their Applications 29

3. Amplifier Since the control commands from the digital controller converted to
analog signals by the DAC are very weak, they need to be amplified to drive, e.g., the
electric motors of the robot manipulator.

2.2 CLASSIFICATION OF ROBOTS

As mentioned in the introduction of this chapter, there are several ways to classify a
robot, for example, based on a coordinate system, i.e., Cartesian or cylindrical, etc.,
or control of a robot or based on its applications, i.e., assembling or welding, etc.
Whereas classifications of robots based on the coordinate system, actuators, control
scheme, and programming methodology are explained in this section, those based on
industrial applications are separately covered in Section 2.3.
2.2.1 Coordinate Systems
Classification of a robot based on its coordinate system is also referred to as
classification by arm configuration or geometric work envelope. It actually classifies
the arm of a robot without considering the wrist and its end-effector or hand. It tells
the volume of reachable coordinates of a point on the end-effecter, rather than its
orientations. There are four such fundamental types, namely, Cartesian, cylindrical,
spherical or polar, and articulated or revolute. They are explained next.
1. Cartesian When the arm of a robot moves in a rectilinear mode, that is, to
the directions of x, y, and z coordinates of the rectangular right-handed Cartesian
coordinate system, as shown in Fig. 2.11(a), it is called Cartesian or rectangular type.
The associated robot is then called Cartesian robot. The movements are referred to as
travel x, height or elevation y, and reach z of the arm. Its workspace has the shape of a
rectangular box or prism, as indicated in Fig. 2.11(b). A Cartesian robot needs a large
volume to operate. It has, however, a rigid structure and provides an accurate position
of the end-effecter. Maintenance of such robots is difficult, as the rectilinear motions
are generally obtained through the sets of rotary electric actuators coupled with nut-
and-ball screws. Dust accumulated on the screws may jam the smooth motion of
the robot. Hence, they have to be covered with bellows. Moreover, maintaining the
straightness of the screw demands higher rigidity in those components. Hence, such
robots tend to be more expensive.

Fig. 2.11 A Cartesian robot arm with its workspace

 30 Introduction to Robotics

2. Cylindrical When the arm of a robot possesses one revolute and two prismatic
joints, i.e., the first prismatic joint of the Cartesian type, Fig. 2.11(a), is replaced by
a revolute one with its axis rotated by 90o
Why different coordinate systems about the reach z-axis, the points that it can
to define robot architectures? reach conveniently be specified by the
No special reason except the cylindrical coordinates, i.e., angle θ, height
convenience of defining a point in the y, and radius z, as in Fig. 2.12(a). A robot
three-dimensional Cartesian space. with this type of arm is termed as cylindrical
robot whose arm moves, θ, y, and z, i.e., it
has base rotation, elevation, and
reach, respectively. Since the
coordinates of the arm can
assume values between specified
upper and lower limits, its end-
effector can move in a limited
volume that is a cut section from
the space between the two
concentric cylinders, as shown in
Fig. 2.12(b). Note that for a
Cartesian arm, this is not the
case. The workspace is a solid
box given by Fig. 2.11(b). The
dotted line in Fig. 2.12(b) Fig. 2.12 A cylindrical robot arm
completes the boundary of the with its workspace
workspace volume for better visualization. It has no other purpose. A robot of this
type may have difficulties in touching the floor near the base. Cylindrical robots are
successfully used when a task requires reaching into small openings or working on
cylindrical surfaces, e.g., welding pipes.
3. Spherical or Polar When the arm of a robot can change its configuration by
moving its two revolute joints
and one prismatic joint, i.e., the
second prismatic joint along the
height y of the cylindrical type is
replaced by a revolute joint with
its axis rotated by 90o about the
reach z-axis, the arm position is
conveniently described by means
of the spherical coordinates, i.e.,
θ, φ, and z. The arm is shown in
Fig. 2.13(a), and is termed a
spherical or polar robot arm. The
arm movements represent the
base rotation, elevation angles,
and reach, respectively. Its
workspace is indicated in Fig. Fig. 2.13 A spherical robot arm
2.13(b). with its workspace
 Industrial Robots and Their Applications 31

4. Articulated or Revolute When a Why some workspaces,

robot arm consists of links connected by e.g., Fig. 2.12(b), are hollow?
revolute joints only, i.e., the third prismatic Hollow portions are those where the
joint of the spherical type is also replaced by end-effector of the robot cannot reach.
another revolute joint with its axis rotated They are called singular zones (see
by 90° about q-axis, it is called articulated Chapter 6 for the definition).
or revolute jointed arm. It is shown
in Fig. 2.14(a). Its spherelike
workspace is shown in Fig. 2.14(b)
whose internal surface is difficult
to determine. Such robots are
relatively simple to fabricate and
maintain, as the robot’s actuators
are directly coupled through a set
of rotary gear or belt elements.
However, achieving a task of the
Cartesian coordinates requires
mathematical transformations.
These aspects are discussed in Fig. 2.14 An articulated robot arm with
Chapter 6. its workspace
Table 2.2 Transformation of robot-arm types

Type Joints
1 (base): Motion 2 (elevation): Motion 3 (reach): Motion
Cartesian P: travel x P: height y P: reach z
Ø Ø – P+R+90°@z Ø Ø
Cylindrical R: rotation q P: -do- P: -do-
Ø Ø Ø –P+R+90°@z Ø
Spherical R: -do- R: rotation φ P: -do-
Ø Ø Ø Ø –P+R+90°@q-axis
Revolute R: -do- R: -do- R: rotation y
–P: Remove prismatic joint; +R: Add revolute joint; +90o@z: Rotate the revolute joint axis
about z-axis by 90o.
It is interesting to note here that how the above four fundamental arm architectures
can be derived from one another. This is explained in Table 2.2, whereas Table 2.3
provides the advantages and disadvantages of those basic robot arms. Some literatures
also classify robots as Gantry and SCARA (Selective Compliance Assembly Robot
Arm), as shown in Figs. 2.15 and 2.16, respectively, and others. This is truly not
required, as the fundamental types will help one to understand such types. For
example, the arm in the gantry robot is a Cartesian type whose base is mounted
overhead, i.e., the robot is placed upside down. This robot is large, versatile in its
operation, and expensive. The SCARA, on the other hand, is a cylindrical type whose
reach is obtained using a revolute, instead of a prismatic joint [compare Figs. 2.12(a)
and 2.16]. A SCARA robot is very suitable for assembly operations, and is, therefore,
extensively used in several industries.
 32 Introduction to Robotics

Table 2.3 Comparison of fundamental robot arms [Courtesy: Fuller (1999)]

Configuration Advantages Disadvantages
Cartesian (3 linear axes) - Easy to visualize - Reach only front and back
x: base travel - Rigid structure - Requires large floor space
y: height - Easy offline programming - Axes are hard to seal
z: reach - Easy mechanical stops - Expensive

Cylindrical (1 rotation and - Can reach all around - Cannot reach above itself
2 linear axes) - Rigid y, z-axes - Less rigid q-axis
q : base rotation - q-axis easy to seal - y, z-axes hard to seal
y: height - Won’t reach around obstacles
z : reach - Horizontal motion is circular
Spherical (2 rotating and - Can reach all around - Cannot reach above itself
1 linear axes) - Can reach above or below - Short vertical reach
q: base rotation obstacles
f: elevation angle - Large work volume
z: reach
Articulated (3 rotating - Can reach above or below - Difficult to program off-line
axes) objects - Two or more ways to reach
q: base rotation - Largest work volume for a point
f: elevation angle least floor space - Most complex robot
y: reach angle

Fig. 2.15 A Gantry robot [Courtesy: Koivo (1989)]

2.2.2 Actuation Systems

Robots are driven by either electric power or
fluid power. The latter category can be further
subdivided into pneumatic and hydraulic.
Today, the most common drive method is
electric with various types of motors, e.g.,
stepper, dc servo, and brushless ac servo.
Pneumatic robots are used in light assembly
or packing work but are not usually suitable
for heavy-duty tasks or where speed control Fig. 2.16 A SCARA arm
is necessary. On the other hand, hydraulic [Courtesy: Fuller (1999)]
 Industrial Robots and Their Applications 33

robots are used in heavy payload applications because of their high power to size
ratios. Actuators are explained in detail in Chapter 3.
2.2.3 Control Methods
Here, control could mean two things. One is motion control strategy, i.e., whether a
robot is servo controlled or not, and the other one is how the motion path is achieved,
i.e., point-to-point or continuous.
1. Servo/Non-servo Control Robots are either servo controlled (closed loop) or
non-servo controlled (open loop). To gain full advantage of digital or microprocessor
control, achieve good precision under heavy load conditions, and to carry out
complex tasks with confidence, full servo control is necessary. In this method of
control, commands are sent to the arm drives to move each axis the requisite amount.
The actual movement is monitored for both displacement and velocity and compared
with the command signal. The difference between the command and the action,
defined as the error, is used as feedback to the controller to enable further commands
to be modified accordingly. Most electric and hydraulic robots are servo controlled.
Pneumatic robots are usually non-servo controlled. In this case, a command signal
is sent and it is assumed that the robot arm reaches its intended position. Non-servo
control is adequate where position control of light loads only is required. However,
if velocity, acceleration, and torque are to be controlled or if the movement against
heavy loads is necessary then non-servo control is usually not possible. The majority
of industrial robots today are servo controlled. This control problem requires
knowledge of Proportional-Derivative (PD), Proportional-Integral (PI), Proportional-
Integral-Derivative (PID), fuzzy, neural network, and other control theories. Some of
the control algorithms are presented in Chapter 10.
2. Motion Control In Point-To-Point (PTP) motion control, the robot arm moves
from one desired point to the next without regard to the path taken between them.
The actual path taken may be the result of a combination of arm link movements
calculated to provide the minimum travel time between the points. Point-to-point
control is widely used for assembly, palletizing, and machine-tool loading/unloading.
Spot-welding robots also use point-to-point control.
In Continuous Path (CP) control, a robot moves along a continuous path with
specified orientations. For example, for welding two metal parts along a straight line
or a specified curve, CP control is used. Signals from the sensors located at the joints
are constantly monitored by the robot controller for appropriate motion control.
Theoretical development of motion control is explained in Chapter 12.
2.2.4 Robot Programming
Industrial robots can be programmed by various means. For example, they can be
programmed either on-line or off-line. On-line methods require the direct use of the
robot and utilize teach pendant for point-to-point or PTP programming, and slave
arms or pistol grip attachments for continuous path or CP programming. More recent
robots have the ability to be programmed off-line, i.e., the robot can continue working
on a particular task while a program for a new task is prepared on a computer terminal
using the robot programming language, for example, VAL, ALU, KRL, and others.
Detail descriptions on robot programming appear in Chapter 13 of this book.
 34 Introduction to Robotics

2.3 INDUSTRIAL APPLICATIONS

As mentioned in Section 1.3.1, serial robots are mainly used in industries for welding,
assembling, machining, etc. As more and more robots are designed for a specific
task, they can be classified based on their tasks, e.g., ‘welding robot’, ‘assembly
robot’, etc. Since these robots are designed for a specified work, they cannot be made
readily adaptable to other applications. For example, an assembly robot cannot be
used for welding purposes unless a major change is done. Note that some suppliers
provide a complete robot system, e.g., for welding, with welding equipment and other
material-handling facilities like turntables, etc., as an integrated unit. Such integrated
robotic system is also called a ‘welding robot’ even though its discrete manipulator
unit could be adapted to a variety of tasks. Some robots are specifically designed for
heavy load manipulation, and are labeled as ‘heavy-duty robots’.
There are certain industrial tasks which robots do very well like material transfer,
machine loading, spot welding, continuous arc welding, spray coating, material
removing, cutting, assembling, inspection, sorting, cleaning and polishing parts, and
a dozen more specialized tasks. In this section, some of these applications will be
examined as per their actions of end-effector (Groover et al., 2012), namely, the
following:
Material Handling In these applications, the robot grasps an object, e.g., a
machined component, with the help of a gripper, and moves it to another location,
say, on a conveyor belt.
Processing Here, the robot uses a tool, e.g., an electrode, rather than a gripper in
order to perform some processing task, say, welding.
Assembling These tasks are more complex than other two, as the robot has to
interact with a fixture or another robot in order to put together two or more components
or sub-assemblies to complete an assembly task.
2.3.1 Material Handling
During the sixties and the seventies, automation affected primarily the manufacturing
process but not the auxiliary functions such as handling, set-up, loading, unloading,
etc. The time spent to transfer a workpiece from one station to the next is still high.
Up to 95 percent of the time involved in manufacturing a part is composed of transfer
and waiting time, and only about 5 percent of the total time is actual processing.
Whereas the processing time has been reduced considerably by automation, much
less progress has been made in handling and loading. The fully automatic systems that
were developed for mass production (e.g., transfer lines in the automobile industry)
are rigid (hard automation) and not suitable for batch production (in the order of
50 to 100,000 parts annually). A more flexible automation technology that takes
into account frequent changes in production is needed for this category of industrial
production, which accounts for about 75 percent of the manufactured parts. A new
solution was offered to the handling and machine-tool loading of small and medium-
size parts with the development of industrial robots. Actually, loading and unloading
of machine tools are the major applications of robots. Robots are utilized to load and
unload machine tools in two basic configurations:
 Industrial Robots and Their Applications 35

(a) A robot tending a single machine,

[Courtesy: http://www.packworld.com/machinery/palletizing/
and
(b) a robot serving several machines.
The first configuration is applied
when typical machining times per part
is short. The second configuration is
applicable when a chain of operations

hassia-usa-inc-robotic-palletizer]
must be executed to complete a part.
For many applications, the cylindrical,
polar, and revolute types are equally
suitable whereas in some light-duty
work in restricted areas, rectangular
types may be used. For material-
handling applications, point-to-point
control is necessary. A typical material-
handling robot is shown in Fig. 2.17. Fig. 2.17 A pick-and-place material-
handling robot
2.3.2 Welding (Processing)
Welding is a manufacturing process in which two metal pieces are joined usually
by heating and fusing. The welding operations performed by robots are thermal
processes in which the metal
pieces are joined by melting or

[Courtesy: http://www.kukarobotics.com/en/
solutions/solutions_search/L_R142_Body_
fusing their contacting surfaces.
These processes can be grouped

shop_Mercedes_Benz_A_Class.htm]
under two classes, namely, where
no filler material is added to the
joint interface, and in which a
filler material of the same parent
metal is added. Accordingly,
there are two types of welding
operations performed by the
robots, namely, spot and arc
welding, respectively, as shown (a) Spot welding of car parts in an assembly plant
in Fig. 2.18.
en/newmachines/907/jointed-arm-robot/

1. Spot Welding
[Courtesy: http://www.cnc-arena.com/

In spot welding, two metal pieces

are joined only at certain points
by melting or fusing their
fanucrobotics/m-10ia]

contacting surfaces. The required

heat is generated by the passage
of an electric current through the
metals at the point where they are
to be joined. It is frequently used
in the automotive industries to
(b) A pair of robots performing arc welding
join thin sheet metals. A spot-
welding robot has to carry the Fig. 2.18 Welding robots
 36 Introduction to Robotics

welding gun, which consists of electrodes, cables required to conduct the high
current, and sometimes a water-cooling system for the electrodes. Since the welding
gun is relatively heavy (10 to 80 kg), most of the spot-welding robots are hydraulically
powered as dc-motor-driven robots cannot
handle such heavy loads. The control system Accuracy vs. Repeatability
for spot-welding robots is of a point-to-point Accuracy is defined as the difference
(PTP) type with reasonable positional between the commanded and actual
positions attained by the robot end-
accuracy and repeatability of ±1 mm. This
effector, whereas repeatability is
repeatability of a robot is much higher than defined as the closeness between the
that of a human welder. Hence, a robotic actual positions attained by the end-
welding is preferred over manual one for effector for the same command.
uniformity of the products at hand. Further,
the operation of robotized spot-welding is very fast.
Spot-welding robots, as shown in Fig. 2.18 (a) for a car assembly, affected the car
industry most. However, they can be used in fabrication of metal products, domestic
appliances, metal furniture, containers which do not require liquid-tight joints, etc.
The first spot-welding robots were installed in 1969 at a General Motor’s plant for
welding of car bodies (Engelberger, 1980). Since then spot-welding robots have
proved to be very profitable. Moreover, parts of the robots can be suspended from the
ceiling, which saves expensive floor space. Several robots can operate simultaneously
on the same car body, which increases the efficiency of the assembly line. A human
operator might miss a weld or make it in an incorrect location, and, therefore, many
times extra welds are added at the design stage. With robot operation, the work is
consistent and all the welds are placed in the right location and, therefore, the required
body strength can be achieved by specifying fewer welds. A typical assembly line
produces between 50 to 90 cars per hour, and the work is performed while the car
bodies are continuously moving on conveyors, which means that the weld locations
specified by the task programs should be synchronized with the velocity of the
assembly line.
2. Arc Welding Arc welding falls in the category in which two metals pieces are
joined along a continuous path by adding a filler material of the same type as the
parent metal. The required heat is provided by an electric arc generated between the
electrode and the metals. It is needed, for example, in sealing a container against
leakage. While most robotic-arc welding uses a consumable wire electrode, e.g., in
MIG welding, with an automatic wire feeder, welding with non-consumable tungsten
electrodes with shielding gas, as in TIG welding, is also in use.
In arc welding, the robot shown in Fig. 2.18(b) uses the welding gun as a tool.
The consumable electrode, which provides the filler material, is in the form of a
wire (coiled on a drum) of the same composition as the material to be welded.
Wire diameters of 1/32 to 3/16 in (0.8 to 4.8 mm) are commonly used. The wire is
automatically fed by a motor with adjustable speed at a preset rate that is determined
by the arc voltage. The wire feed increases with an increase in the voltage applied
between the work and the electrode. This voltage can be monitored and used to
maintain a constant arc length by varying the speed of the motor which feeds the
wire. In order to keep the electrode cooler and permit higher currents to be used, the
 Industrial Robots and Their Applications 37

shielding gas flows in a tube along the electrode. The tube is terminated in a nozzle
at the end of the gun from which the gas flows into the arc region. Robotic welding
systems sometimes use water-cooled guns. The weight of the welding gun is usually
not heavy (unless the water-cooled type is used) and, therefore, dc servomotor-driven
robots are typically used in arc welding, although hydraulically drive robot are also
sometimes found. Welding speeds range from about 10 to over 120 in/min (0.25 to
3 m/min). The welding current usually ranges between 100 and 300 A, but with the
larger electrodes (3/16 in or about 4.75 mm) the current may be as high as 1200 A,
resulting in a very deep penetration of the weld. The control system for robots in
arc welding is usually of a Continuous Path (CP) type. Nevertheless, PTP control
systems are also used.
2.3.3 Spray Painting (Processing)
Spray painting is also a kind of processing used in manufacturing industries. The
unhealthy and unpleasant environment of the painting booth in industry made this
process an ideal candidate for the application of robots. The solvent materials that
are used in spray painting are toxic, and, therefore, the operators must be protected
by masks and be provided with fresh-air ventilation. The painting area must be dust-
free and temperature-controlled, and consequently the painting booth is small in
size and inconvenient for the operators. Furthermore, the noise arising from the air
discharge through the painting nozzles can cause irreversible damage to the ears.
For all these reasons, spray painting became one of the first applications of robots.
The requirement for robots in spray painting are different from those of other robot
applications, and, therefore, many robot manufacturers offer a robot dedicated to
this one application. The spray-painting robots are of CP capability and have the
following characteristics:
(a) high level of manipulator dexterity,
(b) large working volume for small-base manipulator,
(c) compact wrist,
(d) small payload, and
(e) low accuracy and repeatability.
Figure 2.19 shows the spray painting of a car chassis by robots. The painting
robot must be able to carry any type of spray gun. Spray guns, however, are light
weight and, therefore, painting robots are designed for small payloads (e.g., 1 kg).
Finally, the requirements for repeatability and resolution are the least severe in
painting robots. Therefore, a repeatability of 2 mm throughout the working volume
is regarded as sufficient for spray-painting robots.
2.3.4 Machining (Processing)
Machining is another type of manufacturing process. There are five basic types of
machine tools to perform machining, namely, drilling machine, lathe or turning
machine, milling machine, shaper, and grinder. Out of all these machining operations,
only drilling is being successfully done with robots, and mainly in the aircraft
industry. Another application related to machining which is performed by robots is
deburring of metal parts.
 38 Introduction to Robotics

[Courtesy: http://www.roboticsbible.com/
spray-painting-robot.html]
Fig. 2.19 Spray-painting robots with car chassis

1. Drilling Robots can replace the manual operators if the template hole is provided
with a chamfered guide. The gripper of the robot holds a portable drill and guides
from hole to hole. At each hole, a fixed drill cycle is performed, and then the robot
moves the drill to the next hole. Programming the robot to perform the task is quite
simple. Since drilling is a PTP operation, the manual teaching method is appropriate.
Figure 2.20(a) shows a robotized drilling operation.

[Courtesy: http://www.romheld.com.au/
sub_products.php]

(a) Drilling (b) Deburring

Fig. 2.20 Machining robots

2. Deburring Burrs are generated almost always when machining is performed

on metal parts. Burrs can be generated between a machined surface and a raw surface
or at the intersection between two machined surfaces. The removal of these burrs is
an expensive operation. Most deburring is performed manually by workers equipped
with appropriate tools. By closely following the manual method, the industrial robots
can solve most deburring problems. Figure 2.20(b) shows deburring operation by
a robot. There are two basic ways to perform robotized deburring. If the part is
relatively lightweight, it can be picked up by the robot and brought to the deburring
tool. If the part is heavy then the robot holds the tool. The support of the tool is
very important, whether it is held by the robot or mounted on the work table. In
both cases, the relative motion between the tool and the part is of a CP type with
high repeatability (approximately 0.2 mm) and highly controlled speed. Therefore,
deburring is one of the most difficult tasks for robots.
 Industrial Robots and Their Applications 39

2.3.5 Assembling
Assembling with industrial robots is mainly used for small products such as electrical
switches and small motors. Robotized assembly systems are programmable and,
therefore, provide a cost-effective solution for the assembly of small batch sizes
and for batches containing different products. Although industrial robots require the
same fixtures, feeders, and other equipment for positioning the parts as conventional
assembly machines, simpler workpiece feeder and fixtures may be used because of
the robots’ programmability feature. Furthermore, tactile or optical sensors may be
added to the assembly robot to tackle more complex assembly tasks. Some assembly
tasks require the participation of more
than one robot. In order to reduce the

[Courtesy: http://www.fanucrobotics.com]
cost per arm, there are systems in which
several Cartesian arms can use the same
base and share the same controller.
Assembly robots, e.g., the one
similar to Fig. 2.21, can be designed in
any coordinate system: Cartesian,
cylindrical, spherical, or articulated.
However, many tasks require only
vertical assembly motions, such as the
assembly of printed circuit boards. For
these applications, the 4-DOF robot Fig. 2.21 Assembly robot
shown in Fig. 2.22 can be adequate. Its

[Courtesy: http://www.systemantics.
arm has two articulated motions, and
the wrist has two axes of motion: a
linear vertical displacement and a roll
movement. This robot can pick up parts com/products.htm]
located on the horizontal plane, bring
them to the assembly location, orient
them with the roll motion of the wrist,
and finally insert them in a vertical
motion. This class of robot is known as
the SCARA type which was first Fig. 2.22 A 4-DOF robot from
Systemantics India
developed in Japan.

Example 2.3 Robot Considerations for an Application

There are several points which should be taken into account while choosing a robot
for an application. Some of the important ones are listed below:
∑ Degrees-of-freedom (DOF) of the robot for the planned application
∑ Robot work volume
∑ Payload capacity
∑ Type of end-effector, i.e., gripper for material handling, etc.
∑ For grippers, its total DOF, number of fingers or jaws
∑ Actuator type, i.e., electric, hydraulic, or pneumatic
∑ Point-to-point or PTP (mostly used in material-handling applications) or
continuous-path or CP (welding, spraying and similar applications)