MEC 401

Advances in Robotics Lecture 5

ELEMENTS OF ROBOTS: JOINTS, LINKS,
ACTUATORS & SENSORS

By Dr.A Chandrashekhar
Department of Mechatronics
 Position analysis-Coordinate
transformation from one frame
to another in 2D
 Overview
• Why transformations?

• Basic transformations:
– translation, rotation, scaling

• Combining transformations
– homogenous coordinates, transform. Matrices

• First 2D, next 3D

 Transformations

instantiation…
world

train animation…

viewing…
wheel
modelling…
image
 Why transformation?
• For the purpose of controlling a robot, it is
necessary to know the relationships between the
joints motion (input) and the end-effector
motions (output), because the joint motions
control the end-effector movements.

• Thus, the study of kinematics is important, where

transformations between the coordinate frames
attached to different robot links of the robot need
to be performed
What is Transformation
Transformation
 Robot Architecture
• A robot is made up of several links connected serially by joints.
• The robot’s degree of freedom (DOF) depends on the number of links and
joints, their types, and the kinematic chain of the robot.
 Components of Industrial Robot
• An industrial robot consists of

– A manipulator that consists of a sequence of rigid bodies

(links) connected by means of articulation (joints). It is
characterized by an arm (for mobility), a wrist (for dexterity)
and an end effector (to perform a task).

– Actuators that set the manipulator in motion by actuating the

joints (electric, hydraulic or pneumatic)

– Sensors that measure the manipulators status (proprioceptive

sensors) or the status of the environment (exteroceptive
sensors).

– A control system that controls and supervises manipulator

motion.
 An industrial robot has three
fundamental capacities
– Material handling: objects are transported from one
location to another, their physical characteristics are
not altered (e.g. pallettizing, warehouse
loading/unloading, sorting, packaging)

– Material Manipulation: physical characteristics of

objects are changed or their physical identity is lost
(e.g. welding, painting, gluing, cutting, drilling,
screwing, assembly, …)

– Measurement: object properties are measured by

sensors on the end effector, so they are able to
explore a 3D environment (e.g. quality control by
image processing, tactile sensors,…)
 Basic Terms
• Kinematic chain: sequence of links and joints connecting
the end effector to the base. Can be open (single chain) or
closed (loop within the chain).
• Joints:
– Revolute: rotational joint
– Prismatic: translational joint

• Degree of mobility: each joint type provides an

additional DOM (e.g. 6DOM necessary for 3D motion)
• Accuracy: “How close does the robot get to a desired point?”,
distancebetween desired point and actual point.

• Repeatability: “How close does the robot get to a position it has

reached before?”.
 Repeatability and Accuracy
Repeatability: the measurement of how closely a robot returns to
the same position n number of times.
Accuracy: the measurement of how closely the robot moves to a
given target coordinate.
Z
P ( x , y ,z )
a a a a

Pt ( x t , y t , z t )
X

Where: Pt is the target position,

Y Pa the average achieved position,
R is the repeatability, and
A is the accuracy.
 Many factors can effect the accuracy of a robot such
as:
– environmental factors (eg temperature, humidity and electrical noise);

– kinematic parameters (eg robot link lengths etc);

– dynamic parameters (eg structural compliance, friction);

– measurement factors (eg resolution and non-linearity of encoders and

resolvers);

– computational factors (eg digital round off, steady-state control errors); and

– application factors (eg installation errors and errors in defining work-piece

coordinate frames).
 Why?
Each of these can effect the robot kinematics
– depending on the robot model design and manufacturing process
the kinematics can be the most significant source of error.
– if the internal kinematic model of the robot does not match the
actual model we can not accurately predict where the end-
effector is.

Make them match by either

– tighter manufacturing tolerances; or
– by measuring actual kinematics and incorporating this
information into the kinematic model.
 Basic Terms
• Joint space: end effector position and orientation given in
joint coordinates.

• Operational space: end effector position and orientation

given in world coordinates (e.g. cartesian coordinate
system)

• Workspace:
– Reachable WS: set of positions the EE can reach
– Dexterous WS: set of positions the EE can reach with a given
orientation

• Redundancy: more DOM than needed are available

 Basic Joints
Revolute Joint
1 DOF ( Variable - θ)

Prismatic Joint
1 DOF (linear) (Variables - d)
 An Example - The PUMA 560

2
3

4
1 There are two more
joints on the end
effector (the gripper)

The PUMA 560 has SIX revolute joints

A revolute joint has ONE degree of freedom ( 1 DOF) thatis
defined by its angle
Manipulator Structures
Manipulator Structures
Forward and Inverse
Kinematics Joint 2
2

1 Joint 3
Joint 1
yt

Tool Coordinate Frame

zt

z1 Link 1

zw
World (Base) Coordinate Frame
xw

Link Space Tool Space

Forward K
n variables 6 variables
Inverse K
(1 … n) (x,y,z,x,y,z)
 Co-ordinate Frames
• right handed coordinate frames y
x
• used as reference frames

yt
Link frame zt Tool Coordinate
Frame

zw
World (Base) Coordinate
xw Frame
 We are interested in two
kinematics topics
Forward Kinematics (angles to position)
What you are given: The length of each link
The angle of each joint

What you can find: The position of any point

(i.e. it’s (x, y, z) coordinates

Inverse Kinematics (position to angles)

What you are given: The length of each link
The position of some point on the robot

What you can find: The angles of each joint needed to

obtain that position
The Pose of a Robot
 Pose of a Robot
• To control the movement of a robot, we need to understand
and use some equations related to the position of the robot
and its parts relative to both the world and to itself. This
position is called the pose of the robot.
• The pose of a robot tells us location of the robot in either two
or three dimensions and also its orientation. That is, in what
direction is it facing. For mobile robots, the pose can be a
simple three element vector, but for arm-based robots, we
often use a matrix to describe the pose.
• You are requested to go through the links for information on
vectors, matrices, matrix multiplication, and transforming
vectors:
– Vector introduction for linear algebra
https://www.khanacademy.org/math/linear-algebra/vectors-and-spaces/vectors/v/vector-introduction-linear-algebra
– Introduction to matrices
https://www.khanacademy.org/math/precalculus/precalc-matrices/intro-to-%20matrices/v/introduction-to-the-matrix
– Introduction to matrix multiplication
https://www.khanacademy.org/math/precalculus-2018/precalc-matrices/multiplying-matrices-by-matrices/v/matrix-multiplication-
intro
– Introduction to identity matrix
https://www.khanacademy.org/math/precalculus-2018/precalc-matrices/properties-of-matrix-multiplication/v/identity-matrix
– Transforming vectors using matrices
https://www.khanacademy.org/math/precalculus/precalc-matrices/matrices-as-transformations/v/transforming-position-vector
 Quick Math Review
Dot Product: a x 
a 
Geometric Representation:  y

A •B = A B cosθ A bx 
b 
θ  y

Matrix Representation: B

a x  b x 
A•B =  • b  = a xb x + a yb y
a y   y

Unit Vector
Vector in the direction of a chosen vector but whose magnitude
is 1. B
uB =
B B
uB
 Quick Matrix Review

Matrix Multiplication:
An (m x n) matrix A and an (n x p) matrix B, can be multiplied
since the number of columns of A is equal to the number of rows of B.
Non-Commutative Multiplication
AB is NOT equal to BA

a b  e f  (ae + bg ) (af +bh)

c    =
 d  g h  (ce + dg ) (cf + dh)

Matrix Addition:
a b  e f  (a + e) (b + f )
c  +  =
 d  g h  (c + g ) (d + h)
 Coordinate system
• Global coordinate system - corresponds to the
entire body and is used to define nodes in the
entire body.
• Local coordinate system - corresponds to a
particular element in the body , and the
numbering is done to that particular element
neglecting the entire body .
• WORLD-coordinate system is a
Cartesian coordinate system for describing the
location of the points within the workspace.
 Tracking the Robot
• One of the gotchas of robotics is keeping track of which frame of
reference is being used. The global coordinate frame is relative to the
robot’s surrounding. We start robotics experiments with the robot at
position (0, 0) and facing in the direction of the axis
• The local coordinate frame is relative to the robot. We use the local
coordinates when measuring the movement or setting the velocity of the
wheels.
 The Robot’s Global Position
• We call the robot’s location and orientation in the
global coordinate frame the position, or pose, of the
robot. The robot’s position is represented as a matrix
(P) containing its x and y Cartesian coordinate pair
and its angle of orientation (Ɵ).
 Transformations in 2-D
• For robots that move in two dimensions, such as a
mobile robot or arm type robots that move in a plane,
the robot’s pose is a three element vector:
• An alternative expression of the pose is to define it as a
coordinate frame. The coordinate frame consists of two
perpendicular (orthogonal) axis and an (x, y) point
location of the origin of the coordinate frame. The
coordinate frame pose is usually expressed by a matrix.
This allows us to express the coordinate frame of
connected parts and joints of the robot with respect
other coordinate frames. Then by matrix multiplication
and linear algebra, we can easily determine a coordinate
frame with respect to the world frame, or any other
frame of interest.
 Basic Transformations
Moving Between Coordinate Frames

Translation Along the X-Axis

Y O

(VN,VO)

VO

P
N
X VN
Px

Px = distance between the XY and NO coordinate planes

 V X
  V N
 Px 
V =
XY
V =
NO
P= 
Y O
Notation:
V  V  0
 Writing VXY in terms oVf NO

Y O

VO

P θ
N
X VN

PX + V N 
V XY
= O  = P +V NO

 V 
 O
Translation along the X-Axis and Y-Axis

VO

N
VN

Px 
X P XY
= 
PY 

 P + V N

V = P +V = 
XY NO X
O
 Y
P + V 
 Using Basis Vectors
Basis vectors are unit vectors that point along a coordinate
axis
O

n Unit vector along the N-Axis

o Unit vector along the O- Axis

VO
VNO Magnitude of the VNO vector
o
N
n VN

 V N
  VNO cosθ  V NO
cosθ   V NO
• n
V =
NO
 =  NO  =  NO  =  NO 
 V  
O V sinθ  V cos(90 − θ)  V • o 
 Y
Rotation (around the Z-Axis)
Z X
Y

VY

θ
X
VX

θ = Angle of rotation between the XY and NO coordinateaxis

 V X
  V N

V =
XY
V =
NO
Y O
V  V 
 Y X Unit vector along X-Axis
V Can be considered with respect t
the XY coordinates or NO coordinate
V
V X Y
= V N O

VY
 θ

X
VX

V X = V X Y cosα = V N O cosα = V N O • x

= (V  n + V  o) • x
X N O (Substituting for VNO using the N and O
V components of the vector)

V X = V N (x • n) + V O (x • o)
= V N (cosθ) + V O (cos(θ + 90))
= V N (cosθ) − V O (sinθ)
Similarly….
V Y = V N O sinα = V N O cos(90 − α) = V N O • y
V Y = (VN  n + V O  o) • y
V Y = V N (y • n ) + V O (y • o )
= V N (cos(90 − θ)) + V O (cosθ)
= V N (sinθ) + V O (cosθ)
So….
V = V (cosθ) − V (sinθ)
X N O VX 
V XY =  Y
V Y = V N (sinθ) + V O (cosθ) V 
Written in Matrix Form

 V X
  cosθ − sinθ   V N
 Rotation Matrix about the z-
V =
XY
=  O  axis
V  
Y sinθ cosθ V 
 Y1

(VN,VO)
Y0
VNO

VXY

X1
P
Translation along P followed by rotation by 
X0

 V X   P x   cosθ − sinθ   V N 
V XY
= =   +
Y   
cosθ   V O 
 V   P y   sinθ
(Note : Px, Py are relative to the original coordinate frame. Translation followed
by rotation is different than rotation followed by translation.)

In other words, knowing the coordinates of a point (VN,VO) in some coordinate

frame (NO) you can find the position of that point relative to your original
coordinate frame (X0Y0).
 HOMOGENEOUS REPRESENTATION
Putting it all into a Matrix

 V X   P x   cosθ − sinθ   V N  What we found by doing a

V XY
= =   +
Y   
cosθ   V O 
translation and a rotation
 V   P y   sinθ

VX   Px   cosθ − sinθ 0VN  Padding with 0’s and 1’s

 Y  
= V  = P  +  si nθ

cosθ 0   V O 
 y
 1  1   0 0 1 1 

VX   cosθ − sinθ Px   V N 

 Y=
= V 
 si nθ cosθ Py   V O  Simplifying into a matrix form
  
 1   0 0 1  1 

 cosθ − sinθ Px 
H =  sinθ cosθ Py  Homogenous Matrix for a Translation in
 XY plane, followed by a Rotation
 0 0 1  around the z-axis
 Rotation Matrices in 3D – OK,lets return from
 cosθ −si nθ 0  homogenous repn
Rotation around the Z-
R z =  sinθ cosθ 0  Axis
 0 0 1

 cosθ 0 sinθ 
Ry =  0 1 0  Rotation around the Y-
  Axis
−sinθ 0 cosθ

1 0 0 
R z =  0 cosθ − sinθ  Rotation around the X-
 Axis
0 sinθ cosθ 
 Homogeneous Matrices in 3D
H is a 4x4 matrix that can describe a translation, rotation, or both in one
matrix O
1 0 0 Px 
0 1 0 Py 
Y

N H=
P 0 0 1 Pz 
X A 0 1 
Translation without  0 0 
Z rotation

Y
nx ox ax 0
n oy ay 0
H= y
O
N
n z oz az 0
 
X 0 0 0 1
Rotation part:
Z Rotation without Could be rotation around z-
A axis, x-axis, y-axis or a
translation
combination of the three.
 Homogeneous Continued….
VN 
 O The (n,o,a) position of a point relative to the
V 
V =H
XY  current coordinate frame you are in.
VA 
 
 1 

nx ox ax Px   V N 
n oy ay Py V O  VX = n VN + o VO + a VA + P
V XY =  y   x x x x

nz oz az Pz   V A 
  
0 0 0 1  1 

The rotation and translation part can be combined into a single

homogeneous matrix IF and ONLY IF both are relative to the same
coordinate frame.
 Finding the Homogeneous Matrix
EX.

Y WN 
WO 
 
WA
T
P
X

Z
WX   WI  WN 
 Y  Point relative to the  WJ  Point relative to the WO  Point relative to the
W  X-Y-Z frame   I-J-K frame   N-O-A frame
WZ  WK  WA 
 

 WI  ni oi ai Pi   W N 
 W   Pi   n i
I
oi aiW  N
 J  n
 J     O  Pj   W O 
 W  =  Pj  +  n j oj aj W  W = j oj aj  
 W K  Pk  n k  W K  nk Pk   W A 
ok a k  W A    
ok ak
   
 1   0 0 0 1  1 
 W N 
Y WO 
 
WA 

T
P
X

Z
 WX  ix jx kx Tx   W I 
 W  Tx  i x
X
kx W  I
 W Y  i Ty   W J 
jx
 Y      J  = y jy ky  
 W  = Ty  + i y jy k y  W 
 W Z  i z jz kz Tz   W 
K
 W Z  Tz  i z jz k z  W k      
   1   0 0 0 1  1 

 W X   i x jx k x Tx  n i jy k y oi ai Pi   W N 
 WI   Y   
 WJ  i Ty  n j jz k z oj aj Pj   W O 
Substituting    W = y
for WK  W Z   i0z 0 0 Tz  n k ok ak Pk   W A 
     
 1  1  0 0 0 1  1 
  W 
X
WN 
 O
i x jx kx Tx  n i oi ai Pi 
Y   n Pj 
 W  = HW  i jy ky Ty  oj aj
H = y
j
WZ  WA  i z jz kz Tz n k ok ak Pk 
      
 1   1  0 0 0 1  0 0 0 1
Product of the two
matrices
Notice that H can also be written as:

1 0 0 Tx  i x jx kx 01 0 0 Pi  n i oi ai 0
0  0  0  0
1 0 Ty i y jy ky 1 0 Pj n j oj aj 
H =  
0 0 1 Tz  i z jz kz 00 0 1 Pk  n k ok ak 0
0    1
 0 0 1  0 0 0 1  0 0 0 1  0 0 0

H = (Translation relative to the XYZ frame) * (Rotation relative to the XYZ

frame)
* (Translation relative to the IJK frame) * (Rotation relative to the IJK
frame)
 The Homogeneous Matrix is a concatenation of
numerous translations and rotations
W N 
Y WO 
 
WA 

T
P
X

Z
One more variation on finding H:

H= (Rotate so that the X-axis is aligned with T)

* ( Translate along the new t-axis by || T || (magnitude of T))
* ( Rotate so that the t-axis is aligned with P)
* ( Translate along the p-axis by || P || )
* ( Rotate so that the p-axis is aligned with the O-axis)

This method might seem a bit confusing, but it’s actually an easier
way to solve our problem given the information we have.
Coordinate management
Kinematics of Robot Arms
 Kinematics of Robot Arms
• The study of moving bodies and parts is
called Kinematics. In robotics, we use kinematics
to facilitate the discussion of the pose and also
velocity of robots, parts of robots (sensors, arms,
joints, and end effectors), and objects of interest.
• When we know the joint angles and link lengths
of a robot arm, we use forward kinematics to
calculate the pose of parts of the arm. When we
know what pose is needed, we use inverse
kinematics to calculate the joint angles that will
yield the desired pose.
Forward and Inverse
Kinematics Joint 2
2

1 Joint 3
Joint 1
yt

Tool Coordinate Frame

zt

z1 Link 1

zw
World (Base) Coordinate Frame
xw

Link Space Tool Space

Forward K
n variables 6 variables
Inverse K
(1 … n) (x,y,z,x,y,z)
 Co-ordinate Frames
• right handed coordinate frames y
x
• used as reference frames

yt
Link frame zt Tool Coordinate
Frame

zw
World (Base) Coordinate
xw Frame
 We are interested in two
kinematics topics
Forward Kinematics (angles to position)
What you are given: The length of each link
The angle of each joint

What you can find: The position of any point

(i.e. it’s (x, y, z) coordinates

Inverse Kinematics (position to angles)

What you are given: The length of each link
The position of some point on the robot

What you can find: The angles of each joint needed to

obtain that position
Forward Kinematics
 The Situation:
You have a robotic arm that
starts out aligned with the xo-axis.
You tell the first link to move by 1
and the second link to move by 2.

The Quest:
What is the position of the
end of the robotic arm?
Solution:
1. Geometric Approach
This might be the easiest solution for the simple situation. However,
notice that the angles are measured relative to the direction of the previous link.
(The first link is the exception. The angle is measured relative to it’s initial
position.) For robots with more links and whose arm extends into 3 dimensions
the geometry gets much more tedious.

2. Algebraic Approach
Involves coordinate transformations.
 Example Problem:
You are have a three link arm that starts out aligned in the x- axis.
Each link has lengths l1, l2, l3, respectively. You tell the first one to move by 1
, and so on as the diagram suggests. Find the Homogeneous matrix to get
the position of the yellow dot in the X0Y0 frame.
Y3

3

Y2 2 X3 3
2 X2

H = Rz(1 ) * Tx1(l1) * Rz(2 ) * Tx2(l2) * Rz(3 )

i.e. Rotating by 1 will put you in the X1Y1 frame.

Y0
1 Translate in the along the X1 axis by l .1
Rotating by 2 will put you in the X2Y2 frame.
and so on until you are in the X3Y3 frame.
1
X0 The position of the yellow dot relative to the X3Y3 frame is
(l1, 0). Multiplying H by that position vector will give you the
coordinates of the yellow point relative the the X0Y0 frame.
 Slight variation on the last
solution:
Make the yellow 4 4
Y dot the origin of a new coordinate X Y frame
3
Y4
3

Y2 2 X3 3
2 X2
X4

H = Rz(1 ) * Tx1(l1) * Rz(2 ) * Tx2(l2) * Rz(3 ) * Tx3(l3)

Y0 This takes you from the X0Y0 frame to the X4Y4 frame.
1
The position of the yellow dot relative to the X4Y4 frame is (0,0).
1
X0

X  0 
Y  0
  = H  
Z  0  Notice that multiplying by the (0,0,0,1) vector
 1    will equal the last column of the H matrix.
  1
More on Forward Kinematics…
Denavit - Hartenberg
Parameters
 Denavit-Hartenberg
Notation

Z(i - 1)

Y(i -1) Yi Zi

Xi ai
a(i - 1 ) di
X(i -1)
i
( i - 1)

IDEA: Each joint is assigned a coordinate frame. Using the Denavit-

Hartenberg notation, you need 4 parameters to describe how a frame (i)
relates to a previous frame ( i -1 ).
THE PARAMETERS/VARIABLES: , a , d, 
 The Parameters
You can
Z(i - 1) align the two
axis just
Y(i -1) Yi Zi using the 4
Xi ai parameters
a(i - 1 ) di
X(i -1)
i
( i - 1)

1) a(i-1)
Technical Definition: a(i-1) is the length of the perpendicular between the
joint axes. The joint axes is the axes around which revolution takes place
which are the Z(i-1) and Z(i) axes. These two axes can be viewed as lines
in space. The common perpendicular is the shortest line between the two
axis-lines and is perpendicular to both axis-lines.
a(i-1) cont...
Visual Approach - “A way to visualize the link parameter a(i-1) is to imagine
an expanding cylinder whose axis is the Z(i-1) axis - when the cylinder just
touches the joint axis i the radius of the cylinder is equal to a(i-1).” (Manipulator
Kinematics)

It’s Usually on the Diagram Approach - If the diagram already specifies

the various coordinate frames, then the common perpendicular is usually
the X(i-1) axis. So a(i-1) is just the displacement along the X(i-1) to move from
the (i-1) frame to the i frame.

Z(i - 1)

Y(i -1) Yi Zi
Xi ai
a(i - 1 ) di
X(i -1)
i
( i - 1)
2) (i-1)
Technical Definition: Amount of rotation around the common perpendicular so that
the joint axes are parallel.

i.e. How much you have to rotate around the X(i-1) axis so that the Z(i-1) is pointing in
the same direction as the Zi axis. Positive rotation follows the right hand rule.

3) d(i-1)
Technical Definition: The displacement
along the Zi axis needed to align the a(i-1) Z(i - 1)
common perpendicular to the ai common
Y(i -1) Yi Zi
perpendicular.
In other words, displacement along the Xi ai
Zi to align the X(i-1) and Xi axes. a(i - 1 ) di
X(i -1)
If the link is prismatic, then d(i-1) i
is a variable, not a parameter ( i -
1)

4) i or i
Amount of rotation around the Zi axis needed to align the X(i-1) axis with the Xi axis.
The Denavit-Hartenberg Matrix
 cosθi − sinθi 0 a(i−1) 
sinθ cosα cosθ cosα − sinα − sinα d 
 i (i −1) i (i−1) (i−1) (i −1) i 

sinθisinα(i −1) cosθisinα(i −1) cosα(i −1) cosα(i −1)di 

 
 0 0 0 1 
• Just like the Homogeneous Matrix, the Denavit-Hartenberg
Matrix is a transformation matrix from one coordinate frame to
the next. Using a series of D-H Matrix multiplications and the
D-H Parameter table, the final result is a transformation matrix
from some frame to your initial frame.
Z(i - 1)
Y(i -1) Yi Zi
X i ai
a(i - 1 ) d i
X(i -1)
i
( i - 1)
 3 Revolute JoY ints 2

Z0 Z1

X2
d2
X0 X1
Denavit-Hartenberg Link
Y0 Y1
Parameter Table
a0 a1 i (i-1) a(i-1) di i

Notice that the table has two uses: 0 0 0 0 0

1) To describe the robot with its
variables and parameters. 1 0 a0 0 1
2)To describe some state of the
robot by having a numerical values 2 -90 a1 d2 2
for the variables.
 Y2 i (i-1) a(i-1) di i

Z0 Z1
0 0 0 0 0
X2
d2 1 0 a0 0 1
X0 X1

Y0 Y1 2 -90 a1 d2 2

a0 a1

T =( T)(0 T)(1T)
V  X2
0 1 2
 Y2 
 V 
V X 0 Y0 Z 0 =T
V Z2  Note: T is the D-H matrix with (i-1) = 0 and i = 1.
 
 1 
 i (i-1) a(i-1) di i
cosθ0 −sinθ0 0 0
0 0 0 0 0 sinθ cosθ0 0 0
T =  0

1 0 a0 0 1
0
 0 0 1 0
 
 0 0 0 1
2 -90 a1 d2 2
This is just a rotation around the Z0
axis

cosθ1 − sinθ1 0 a0   cosθ2 − sinθ2 0 a1 

sinθ
 cosθ1 0 0   0
 0 1 d 2
1T = 2T =
0 1 1

 0 0 0 0 − sinθ 2 − cosθ 2 0 0
   0 
 0 0 0 1  0 0 1

This is a translation by a0 followed by a This is a translation by a1 and then d2

rotation around the Z1 axis followed by a rotation around the X2 and Z2
axis
T =( T)(0 T)(1T)
0 1 2
 Inverse Kinematics
From Position to Angles
K-1

(q1 … qn) (x,y,z,x,y,z)

 Inverse Kinematics
• What we have:
– Geometric layout of robot (constant under robot motion)
– End Effector (EE) position and orientation wrt base coordinate frame (variable under robot
motion)

• What we want:
– Joint coordinates qnx1 (variable under robot motion).
 A mapping f from EE position and orientation to joint coordinates:

q1 
 ...  = f (x, y,z, , , )
  x y z
• Problematic, because: qn
– EE position may be out of reachable workspace (no solution)
– EE position and orientation may be out of dexterous workspace (no solution)
– EE position may be reachable in several ways (multiple solutions, redundancy)
– EE position may be reachable in infinitely many ways (infinite number of solutions,
redundancy)
– f is highly nonlinear, no closed form solution for complicated configurations.
 Inverse Kinematics
• How we get there:

– Calculate closed-form solution, if possible

• fast
• accurate
• unique for every arm configuration (no general concept)

– Use Iterative Method based on differential kinematics

• slower
• gives approximate solution
• converges to a single solution (even if multiple solutions exist)
• general algorithm, applicable to all arm configurations
Multiple Solutions

•Because we are dealing with

transcendental functions there is
often more than one solution.

eg: cos-1(0.5) = 60º and 300º

Insoluble
 A Simple Example
Revolute and Finding :
Prismatic θ = arctan( y )
Joints x
Combined More Specifically:
(x , y) θ = arctan 2( y ) arctan2() specifies that it’s in
x the first quadrant

S Finding S:
1

X S = (x 2 + y 2 )
Inverse Kinematics of a Two Link Manipulator

(x , y) l
Given: 1, l2 , x , y
Find:  1,  2
2

l2 Redundancy:

A unique solution to this problem does

not exist.
l1 Sometimes no solution is possible.
1

(x , y)
 The Geometric Solution
l2 Using the Law of Cosines:
2
(x , y)
c 2 = a 2 + b 2 − 2ab cos C
(x 2 + y 2 ) = l 2 + l 2 − 2l l cos(180 − θ )
1 2 1 2 2

cos(180 − θ 2 ) = −cos(θ2 )
l1 x12 + y 2 − l 22 − l 2

 cos(θ2 ) =
2l1l2
1 x 21 + y 2 − l 22 − l 2 
θ 2 = arccos 
 2l1l2 
Using the Law of Cosines:
Redundant since 2 could be in
sin B sin C
= the first or fourth quadrant.
b c
sinθ1 sin(180 −θ 2 ) sin(θ2 ) Redundancy caused since 2 has two
= =
l2 x 2 + y2 x 2 + y2 possible values
θ1 = θ1 +α  l sin(θ )   y
 y θ1= arcsin  2 2 
+ arctan 2 
α = arctan2   x 2 + y2   x
 x
 
 (x , y)
l2 
2
The Algebraic Solution

c 1 = co sθ 1
l1 c1+2 = co s (θ 2 + θ1 )
(1) x = l 1 c 1 + l 2 c 1 + 2
1
(2) y = l 1 s 1 + l 2 sin 1+ 2
(3) θ = θ 1 + θ 2
(1)2 + (2)2 = x 2 + y 2 =
( 2 2
)( 2 2
)
= l1 c1 + l2 2 (c 1+2 ) 2 + 2l1l 2 c1 (c 1+2 ) + l1 s1 + l2 2 (sin 1+2 ) 2 + 2l1l2 s1 (sin1+2 )
= l1 2 + l2 + 2l1 l2 (c(c
1 1+2 ) + s1 (sin 1+2 ))
2

= l1 2 + l2 + 2l1l 2 c 2
2
Only Unknown Note:
 x21 + y 2 − l22 − l 2  cos(a+b) = (cos a)(cosb)− (sin a)(sinb)
− +
θ 2 = arccos 
 + +
2l1l2 sin(a b) = (cos a)(sinb) (cosb)(sin a)
  − −
 x = l1 c 1 + l 2 c 1 + 2 Note:
= l1 c 1 + l 2 c 1 c 2 − l 2 s 1 s 2 cos(a+b) = (cosa)(cosb)− (sin a)(sinb)
− +
= c 1 (l 1 + l 2 c 2 ) − s 1 (l 2 s 2 )
sin(a+b) = (cosa)(sinb)+ (cosb)(sin a)
− −
y = l 1 s 1 + l 2 sin 1 + 2
We know what  2 is from the previous
= l1 s1 + l 2 s 1 c 2 + l 2 s 2 c1
slide. We need to solve for 1 . Now
= c 1 (l 2 s 2 ) + s 1 (l 1 + l 2 c 2 ) we have two equations and two
unknowns (sin 1 and cos 1 )
x + s 1 (l 2 s 2 )
c1 = Substituting for c1 and
(l 1 + l 2 c 2 ) simplifying many times
x + s 1 (l 2 s 2 )
y= (l 2 s 2 ) + s 1 (l 1 + l 2 c 2 )
(l 1 + l 2 c 2 )
Notice this is the law of cosines
=
1
(l 1 + l 2 c 2 )
(
xl 2 s 2 + s1 (l1 2 + l 2 2 + 2l1l 2 c 2 ) ) and can be replaced by x2+ y2

y ( l + l 2 c 2 ) − x l2 s  y(l 1 + l 2 c 2 ) − x l 2 s 2 
s1 =  1 2 θ 1 = arcsin  
 x 2 + y2 
x 2 + y2  
  cosθ −si nθ 0
Rotation around the Z-
R z =  sinθ cosθ 0  Axis
 0 0 1