Robot kinematics

Kinematics is the study of relationship between robot’s joint coordinates and its spatial layout.
Kinematics can yield very accurate calculations in many problems such as positioning a gripper
at a place in space, designing mechanism that can move a tool from point A to point B, or
predicting whether a robot’s motion could collide with obstacles.

Figure 1 Robot
Kinematics is about how you convert the position and velocity of an end effector (gripper,
robotic arm hand) to position and velocity of the joints (i.e the servo motors along robotic arm),
and vice versa. The motion of the bodies in robotic mechanism without regard to the
forces/torque that cause the motion is kinematics.

An end effector
This is a component on robot arm that has effect on the environment. For example, robotic
gripper.

Joint of robots
Joint in a robot is like joint in a human body (shoulder, elbow etc). A joint is a movable part of
a robot. In robots there are three types of robots namely, revolute, prismatic, spherical joints.

1
 i. Revolute joint
Is a one degree of freedom kinematic pair used frequently in mechanism and machines. The
joint constrains the motion of two bodies to pure rotation along a common axis.

Figure 2 Revolute Joints

ii. Prismatic joint
is a one degree of freedom kinematic pair which constrains the motion of two bodies to
sliding along common axis, without rotation, for this reason it is called a slider or a sliding
pair. They often utilized in hydraulic and pneumatic cylinders.

Figure 3 Prismatic Joint

iii. Spherical joint

This also known as ball and socket joint, which allow three relative rotation between the two
connected segments and is frequently used in model the hip and glenohumeral joints.

2
 Figure 4 Revolute Joint
There are two kinematics of robot
i. Forward kinematics and
ii. Inverse kinematics

Forward kinematic of robot

forward kinematics asks where the end effector of the arm will be following a sequence of
rotation of the joints of the arm. A rotation of robotic manipulator is described by a rotational
matrix whose elements are trigonometric function of the angle rotation. The rotation matrix for
planar rotation is derived followed by an overview of three-dimension rotation.

In forward kinematics you are given length of each joint and angle of each joint. Then you are
required to find the position of any points coordinates. (x, y, z)
Inverse kinematics
Inverse kinematics asks what rotation of the joints will bring the end effector to a specific
position.
In inverse kinematics you are given the length of each link and the position of some points of
the robot. Then you are required to find the angle of each joints needed to obtain the position.

Basic Transformations Moving Between Coordinate Frames

Translation Along the X-Axis

3
Px = distance between the XY and NO coordinate planes

𝑋 𝑁 𝑃
𝑉 𝑋𝑌 = [𝑉 𝑌 ], : ̅̅̅̅̅̅̅̅
Notation: ̅̅̅̅̅̅̅̅ 𝑉𝑁𝑂 = [𝑉 𝑂 ], : 𝑃̅ =
̅ [ 𝑋]
𝑉 𝑉 0

Writing 𝑉̅ 𝑋𝑌 in terms of 𝑉̅ 𝑁𝑂

𝑁
𝑃 +𝑉
𝑉̅ 𝑋𝑌 = [ 𝑋 𝑂 ]= 𝑃̅𝑋 + 𝑉̅ 𝑁𝑂
𝑉

Translation along the X-Axis and Y-Axis

𝑃
𝑃̅ 𝑋𝑌 = [ 𝑋 ],
𝑃𝑌

𝑃𝑋 + 𝑉 𝑁
𝑉 𝑋𝑌 = 𝑃 𝑋𝑌 + 𝑉 𝑁𝑂 =[ ]
𝑃𝑌 + 𝑉 0

4
Using Basis Vectors
Basis vectors are unit vectors that point along a coordinate axis

Where,
𝑛̅ − 𝑢𝑛𝑖𝑡 𝑣𝑒𝑐𝑡𝑜𝑟 𝑎𝑙𝑜𝑛𝑔 𝑁 𝑎𝑥𝑖𝑠
𝑜̅ − 𝑢𝑛𝑖𝑡 𝑣𝑒𝑐𝑡𝑜𝑟 𝑎𝑙𝑜𝑛𝑔 𝑂 𝑎𝑥𝑖𝑠

‖𝑉̅ 𝑁𝑂 ‖ − 𝑚𝑎𝑔𝑛𝑖𝑡𝑢𝑑𝑒 𝑜𝑓 𝑉̅ 𝑁𝑂 𝑉𝑒𝑐𝑡𝑜𝑟

𝑁 ‖𝑉̅ 𝑁𝑂 ‖ cos 𝜃 𝑁𝑂
𝑉̅ 𝑁𝑂 = [𝑉 0 ]=[ ̅ 𝑁𝑂 ]=[𝑉 𝑁𝑂 . 𝑛̅]
𝑉 ‖𝑉 ‖ cos(90 − 𝜃) 𝑉 . 0̅

Rotation (around the Z-Axis)

5
Y = Angle of rotation between the XY and NO coordinate axis

𝑋 𝑁
𝑉 𝑋𝑌 = [𝑉 𝑌 ], ̅̅̅̅̅̅̅̅
: ̅̅̅̅̅̅̅̅ 𝑉𝑁𝑂 = [𝑉 𝑂 ]
𝑉 𝑉

𝑋 −Unit vector along X-Axis

‖𝑉̅ 𝑋𝑌 ‖ = ‖𝑉̅ 𝑁𝑂 ‖

𝑉 𝑋 = ‖𝑉̅ 𝑋𝑌 ‖𝑐𝑜𝑠𝛼 = ‖𝑉̅ 𝑁𝑂 ‖𝑐𝑜𝑠𝛼 = 𝑉̅ 𝑁𝑂 . 𝑥

𝑉 𝑋 = 𝑉 𝑁 (𝑥̅ . 𝑛̅) + 𝑉 0 (𝑥̅ . 𝑜̅ )

𝑉 𝑋 = 𝑉 𝑁 cos 𝜃 + 𝑉 0 cos(𝜃 + 90)

𝑉 𝑋 = 𝑉 𝑁 cos 𝜃 − 𝑉 0 sin 𝜃

similary

𝑉 𝑌 = ‖𝑉̅ 𝑁𝑂 ‖ sin 𝛼 = 𝑉̅ 𝑁𝑂 cos(90 − 𝛼)=𝑉̅ 𝑁𝑂 . 𝑦̅

𝑉 𝑋 = (𝑉 𝑁 ∗ 𝑛̅ + 𝑉 0 ∗ 𝑜̅ ). 𝑦̅

𝑉 𝑦 = 𝑉 𝑁 (𝑦̅. 𝑛̅) + 𝑉 0 (𝑦̅. 𝑜̅ )

𝑉 𝑦 = 𝑉 𝑁 cos(90 − 𝜃) + 𝑉 0 cos 𝜃

𝑉 𝑌 = 𝑉 𝑁 sin 𝜃 + 𝑉 0 cos 𝜃

So,

6
𝑉 𝑋 = 𝑉 𝑁 cos 𝜃 − 𝑉 0 sin 𝜃 and
𝑉 𝑌 = 𝑉 𝑁 sin 𝜃 + 𝑉 0 cos 𝜃
To write in matrix form,
𝑋
𝑉̅ 𝑋𝑌 = [𝑉 𝑌 ],
𝑉
𝑋 cos 𝜃 − sin 𝜃 𝑉 𝑁
𝑉̅ 𝑋𝑌 = [𝑉 𝑌 ]=[ ][ ]
𝑉 sin 𝜃 cos 𝜃 𝑉 𝑂

𝑋 𝑃 cos 𝜃 − sin 𝜃 𝑉 𝑁
𝑉̅ 𝑋𝑌 = [𝑉 𝑌 ]=[ 𝑋 ] [ ][ ]
𝑉 𝑃𝑌 sin 𝜃 cos 𝜃 𝑉 𝑂
(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 (X0,Y0).
Homogeneous representation
Putting it all into a Matrix
𝑋 𝑃 cos 𝜃 − sin 𝜃 𝑉 𝑁
𝑉̅ 𝑋𝑌 = [𝑉 𝑌 ]=[ 𝑋 ] [ ][ ]
𝑉 𝑃𝑌 sin 𝜃 cos 𝜃 𝑉 𝑂
𝑉 𝑋 𝑃𝑋 cos 𝜃 − sin 𝜃 0 𝑉 𝑁
𝑉̅ 𝑋𝑌 =[ 𝑉 𝑌 ]=[𝑃𝑌 ] [ sin 𝜃 cos 𝜃 0] [ 𝑉 0 ]
1 1 0 0 1 1
𝑉 𝑋 cos 𝜃 − sin 𝜃 𝑃𝑋 𝑉 𝑁
𝑉̅ =[ 𝑉 𝑌 ]=[ sin 𝜃 cos 𝜃 𝑃𝑌 ] [ 𝑉 0 ]
𝑋𝑌

1 0 0 1 1
cos 𝜃 − sin 𝜃 𝑃𝑋
H=[ sin 𝜃 cos 𝜃 𝑃𝑌 ] but,H is homogenous Matrix for a Translation in XY plane,
0 0 1
followed by a Rotation around the z-axis.

7
References

References
Angelea, J. (1998). Fundamental Of Robotic Mechanical system: Theory,methods, and algorithms
(Second Edition ed.). New york: TLFeBOOk.

Todd, D. J. (1989). Fundamental Of Robot Technology. Kogan Page LTD.

1.Saha, S.K.( 2014) “Introduction to Robotics, 2nd Edition, McGraw-Hill Higher Education,
New Delhi..
Ashitava Ghosal.( 2006), “Fundamental Concepts and Analysis”, Oxford, New Delhi,
Mikell.P.Groover, Mitchell Weiss, Roger.N.Nagel, Nicholas.G.Odrey(2008), “Industrial
Robotics-Technology, Programming and Applications”, Tata McGraw-Hill Publishing
Company Limited, New Delhi, Third Reprint