Introduction to Mechatronics

Lecture on Robotics

Mohammad Shahbazi

School of Mechanical Engineering

Iran University of Science and Technology

* The slides are partly based on the course “Robot Modeling and Control” by Prof. Anton Shiriaev. Written permission to re-
use the material has been obtained.
Lecture1: Introduction and Conceptual Problems
● Introduction
● Robot Types
● Conceptual Problems:
● Forward Kinematics
● Inverse Kinematics
● Velocity Kinematics
● Dynamics
● Path Planning and Trajectory Generation
● Motion Control
● Force Control
● Computer Vision and Vision Based Control
Lecture1: Introduction and Conceptual Problems
● Introduction
● Robot Types
● Conceptual Problems:
● Forward Kinematics
● Inverse Kinematics
● Velocity Kinematics
● Dynamics
● Path Planning and Trajectory Generation
● Motion Control
● Force Control
● Computer Vision and Vision Based Control
What Do We Mean by Robot?
● Virtually anything that operates with some degree of autonomy,
usually under computer control, has at some point been called a
robot;
● In this course, the term robot will mostly! mean a computer
controlled industrial manipulator

–p.1/20
Components and Structure of Robots
Robot Manipulators are composed of links connected by joints into a
kinematic chain

Components of a robotic system Symbolic representation of robot joints

–p.2/20
Degrees of Freedom and Workspace
● The degrees-of-freedom (DOF) of a rigid body is defined as the number of
independent movements it has
● For a robot, the number of joints determines the DOF
● When DOF < 6: the arm cannot reach every point in its work environment with
arbitrary orientation
● When DOF > 6: kinematically redundant manipulator

● The workspace of a manipulator is the total volume swept out by the end-
effector as the manipulator executes all possible motions
● Reachable Workspace: the entire set of reachable points
● Dextrous Workspace: those points that the manipulator
can reach with an arbitrary orientation of
the end-effector

Workspace of the SCARA manipulator

–p.3/20
Wrists and End-Effectors
● Wrist: the joints in the kinematic chain between the arm and hand
● The most common wrists are spherical;
● Spherical wrist greatly simplifies the kinematic analysis.

● End-Effector: the “hand” or tool of robot that actually performs the

work End-Effector
● The simplest end-effectors are grippers;
● Design of special purpose and modular
end-effectors is an active field of research;
e.g., development of anthropomorphic hands;
● We will not discuss end-effector design or the study of grasping
and manipulation.

–p.4/20
Lecture1: Introduction and Conceptual Problems
● Introduction
● Robot Types
● Conceptual Problems:
● Forward Kinematics
● Inverse Kinematics
● Velocity Kinematics
● Dynamics
● Path Planning and Trajectory Generation
● Motion Control
● Force Control
● Computer Vision and Vision Based Control
Robot Types
Fixed-base Robots:

Serial Parallel Hybrid

Larger workspace Higher stiffness Combined

–p.5/20
Robot Types
Mobile Robots:

Ground Robots Aerial Robots Underwater Robots

–p.6/20
 Robot Types
Ground Robots

Wheeled Legged Tracked Combinatorial

Engineering solution Nature solution Engineering solution The future!

–p.7/20
Lecture1: Introduction and Conceptual Problems
● Introduction
● Robot Types
● Conceptual Problems:
● Forward Kinematics
● Inverse Kinematics
● Velocity Kinematics
● Dynamics
● Path Planning and Trajectory Generation
● Motion Control
● Force Control
● Computer Vision and Vision Based Control
Conceptual Robotics Application:

Example: Two-link planar robot used for grinding a wall

– p. 8/20
Forward and Inverse Kinematics:

To perform the task, we need to know

• a parametrization of a wall
• configurations of the robot for which it touches the wall, i.e.
Forward Kinematics: the position (x, y) of end-effector as a
function of joint angles (θ1 , θ2 );
Inverse Kinematics: angles (θ1 , θ2 ) as functions of (x, y)

– p. 9/20
Forward Kinematics:

x = x1 + x2 = a1 cos θ1 + a2 cos(θ1 + θ2 )

– p. 10/20
Forward Kinematics:

x = x1 + x2 = a1 cos θ1 + a2 cos(θ1 + θ2 )

y = y1 + y2 = a1 sin θ1 + a2 sin(θ1 + θ2 )
– p. 10/20
Forward Kinematics (Tool Frame Orientation):

 
x02 = cos(θ1 + θ2 ), sin(θ1 + θ2 )

– p. 11/20
Forward Kinematics (Tool Frame Orientation):

 
x02 = cos(θ1 + θ2 ), sin(θ1 + θ2 )
 
y20 = − sin(θ1 + θ2 ), cos(θ1 + θ2 )
– p. 11/20
Inverse Kinematics:

If the tool position (x, y) is given, but the orientation is not

defined, then, except particular cases, there are two
configurations: Elbow Up and Elbow Down.

– p. 12/20
Inverse Kinematics:

x2 + y 2 − a21 − a22 n p o
cos θ2 = := D, ⇒ sin θ2 = ± 1 − D 2
2a1 a2

– p. 13/20
Inverse Kinematics:

x2 + y 2 − a21 − a22 n p o
cos θ2 = := D, ⇒ sin θ2 = ± 1 − D 2
2a1 a2
  p !
2
sin θ2 ± 1−D
⇒ θ2 = tan −1
= tan −1
cos θ2 D
– p. 13/20
Inverse Kinematics:

p !
2
± 1−D
θ2 = tan−1
D
   
y a2 sin θ2
⇒ θ1 = tan −1
− tan −1
x a1 + a2 cos θ2
– p. 13/20
Velocity Kinematics:
Geometrical relations between (x, y) and (θ1 , θ2 )

x = a1 cos θ1 +a2 cos(θ1 +θ2 ), y = a1 sin θ1 +a2 sin(θ1 +θ2 )

– p. 14/20
Velocity Kinematics:
Geometrical relations between (x, y) and (θ1 , θ2 )

x = a1 cos θ1 +a2 cos(θ1 +θ2 ), y = a1 sin θ1 +a2 sin(θ1 +θ2 )

imply the relations between velocities

 
d d d d
dt
x(t) = −a1 sin θ1 dt− a2 sin(θ1 + θ2 )
θ1 + θ
dt 1
θ
dt 2
 
d d d d
dt
y(t) = a1 cos θ1 dt θ1 + a2 cos(θ1 + θ2 ) dt θ1 + dt θ2

– p. 14/20
Velocity Kinematics:
Geometrical relations between (x, y) and (θ1 , θ2 )

x = a1 cos θ1 +a2 cos(θ1 +θ2 ), y = a1 sin θ1 +a2 sin(θ1 +θ2 )

imply the relations between velocities

 
d d d d
dt
x(t) = −a1 sin θ1 dt− a2 sin(θ1 + θ2 )
θ1 + θ
dt 1
θ
dt 2
 
d d d d
dt
y(t) = a1 cos θ1 dt θ1 + a2 cos(θ1 + θ2 ) dt θ1 + dt θ2

In compact form it is
" # " #" #
ẋ −a1 sin θ1 − a2 sin(θ1 + θ2 ) −a2 sin(θ1 + θ2 ) θ̇1
=
ẏ a1 cos θ1 + a2 cos(θ1 + θ2 ) a2 cos(θ1 + θ2 ) θ̇2
| {z }
=: J (θ1 , θ2 )

– p. 14/20
Velocity Kinematics:
Geometrical relations between (x, y) and (θ1 , θ2 )

x = a1 cos θ1 +a2 cos(θ1 +θ2 ), y = a1 sin θ1 +a2 sin(θ1 +θ2 )

imply the relations between velocities

 
d d d d
dt
x(t) = −a1 sin θ1 dt− a2 sin(θ1 + θ2 )
θ1 + θ
dt 1
θ
dt 2
 
d d d d
dt
y(t) = a1 cos θ1 dt θ1 + a2 cos(θ1 + θ2 ) dt θ1 + dt θ2

In compact form it is
" # " #" #
ẋ −a1 sin θ1 − a2 sin(θ1 + θ2 ) −a2 sin(θ1 + θ2 ) θ̇1
=
ẏ a1 cos θ1 + a2 cos(θ1 + θ2 ) a2 cos(θ1 + θ2 ) θ̇2
| {z }
=: J (θ1 , θ2 )
The matrix J (·) is called the Jacobian of the manipulator.
– p. 14/20
Velocity Kinematics:
The relation between the joint velocities and the tool velocity
" # " #
ẋ θ̇1
= J (θ1 , θ2 )
ẏ θ̇2

allows to compute the joint velocities θ̇1 (t), θ̇2 (t) to achieve the
particular velocity of the tool!

– p. 15/20
Velocity Kinematics:
The relation between the joint velocities and the tool velocity
" # " #
ẋ θ̇1
= J (θ1 , θ2 )
ẏ θ̇2

allows to compute the joint velocities θ̇1 (t), θ̇2 (t) to achieve the
particular velocity of the tool!

Indeed, θ̇1 (t), θ̇2 (t) are found by

" # " #
θ̇1 ẋ
= J (θ1 , θ2 )−1
θ̇2 ẏ

with
 
1 a2 cos(θ1 + θ2 ) a2 sin(θ1 + θ2 )
J −1 =  
a1 a2 sin θ2 −a1 cos θ1 − a2 cos(θ1 + θ2 ) −a1 sin θ1 − a2 sin(θ1 + θ2 )

– p. 15/20
Velocity Kinematics:

It is clear that the inverse of the Jacobian

 
1 a2 cos(θ1 + θ2 ) a2 sin(θ1 + θ2 )
J −1 =  
a1 a2 sin θ2 −a1 cos θ1 − a2 cos(θ1 + θ2 ) −a1 sin θ1 − a2 sin(θ1 + θ2 )

is not defined when


sin θ2 = 0 ⇒ θ2 = π · k, k = · · · − 1, 0, 1, . . .

– p. 16/20
Velocity Kinematics:

It is clear that the inverse of the Jacobian

 
1 a2 cos(θ1 + θ2 ) a2 sin(θ1 + θ2 )
J −1 =  
a1 a2 sin θ2 −a1 cos θ1 − a2 cos(θ1 + θ2 ) −a1 sin θ1 − a2 sin(θ1 + θ2 )

is not defined when


sin θ2 = 0 ⇒ θ2 = π · k, k = · · · − 1, 0, 1, . . .

What does this mean?

– p. 16/20
Singular Configurations

If θ2 = 0, then the Jacobian J (·) looses the rank and cannot be

inverted. It means that the tool velocity cannot be any, but
should belong to particular line irrespective of the joint velocities.

– p. 17/20
Dynamics:

• Kinematics and Velocity Kinematics define the relations

between variables irrespective of actuation, and physical
(Newton) laws;

– p. 18/20
Dynamics:

• Kinematics and Velocity Kinematics define the relations

between variables irrespective of actuation, and physical
(Newton) laws;
• Dynamics are equations, which allow to introduce effect of
control action and define variables as functions of time;

– p. 18/20
 Dynamics:

• Kinematics and Velocity Kinematics define the relations

between variables irrespective of actuation, and physical
(Newton) laws;
• Dynamics are equations, which allow to introduce effect of
control action and define variables as functions of time;
• For instance, for the double pendulum the dynamics are
     
p1 + p2 + 2p3 cos θ2 p2 + p3 cos θ2 θ̈1 −θ̇2 −θ̇2 − θ̇1 θ̇1
   +p3 sin θ2   +
p2 + p3 cos θ2 p2 θ̈2 θ̇1 0 θ̇2
   
p4 g cos θ1 + p5 g cos(θ1 + θ2 ) u1
+ = ,
p5 g cos(θ1 + θ2 ) u2

where
◦ p1 -p5 are constants defined by physical parameters;
◦ u1 , u2 are control torques.

– p. 18/20
Path Planning and Trajectory Generation:

• Path is a curve in the configuration space of the robot;

– p. 19/20
Path Planning and Trajectory Generation:

• Path is a curve in the configuration space of the robot;

• Trajectory is the path augmented with the information on

velocities how the system links will travel along the path;

– p. 19/20
Path Planning and Trajectory Generation:

• Path is a curve in the configuration space of the robot;

• Trajectory is the path augmented with the information on

velocities how the system links will travel along the path;

• Planning trajectory might include also specifications on

accelerations

– p. 19/20
Motion Control:

Basic structure of a feedback control system.

– p. 20/20