Advanced Robotics

ENGG5402

Fei Chen
Topics:
• Direct kinematics

Readings:
• Siciliano: Sec. 2.8
 Kinematics of Robot Manipulators

• Study of…
geometric and timing aspects of robot motion,
without reference to the causes producing it

• Robot seen as…

an (open) kinematic chain of rigid bodies
interconnected by (revolute or prismatic) joints

2
 Motivations
• Functional aspects
• Definition of robot workspace

• Operational aspects (Operational Space)

task execution task definition and

(actuation by motors) performance
Actuator Space -> Joint Space Task Space / Cartesian Space

two different “spaces” related by kinematic (and dynamic) maps

• trajectory planning
• programming
• motion control

3
 Kinematics
Kinematics (formulation and parameterizations)

𝑟 = 𝑓(𝑞)

DIRECT TASK
JOINT
INVERSE (Cartesian)
space
space

𝑞 = 𝑞1 , … , 𝑞𝑛 𝑞 = 𝑓 −1 (𝑟) 𝑟 = 𝑟1 , … , 𝑟𝑚

• choice of parameterization 𝑞
• unambiguous and minimal characterization of robot configuration
• 𝑛 = degrees of freedom (𝐷𝑂𝐹) = robot joints (rotational or translational)
• choice of parameterization 𝑟
• compact description of position and/or orientation (pose) variables of interest to the
required task
• usually, 𝑚 ≤ 𝑛 and 𝑚 ≤ 6 (but none of these is strictly necessary)
4
 Open Kinematic Chains

𝑅𝐹𝐸
𝑞4
𝑞𝑛 𝑟 = (𝑟1, … , 𝑟 𝑚 )
𝑞3
𝑞1 𝑞2

e.g., it describes the

e.g., the relative angle
pose of frame 𝑅𝐹𝐸
between a link and the
following one

• 𝑚=2 • 𝑚=5
• pointing in space • positioning and pointing in space (like for spot welding)
• positioning in the plane • 𝑚=6
• 𝑚=3
• positioning and orientation in space
• orientation in space
• positioning and orientation in the plane • positioning of two points in space (e.g., end-effector and elbow)

5
 Classification
Classification by kinematic type (first 3 dofs only)

SCARA
cylindric (RRP)
(RPP)

Cartesian or
gantry (PPP)

articulated or
anthropomorph
ic (RRR)

polar or
𝑅 = 1 − 𝑑𝑜𝑓 rotational (revolute) joint
spherical
(RRP) 𝑃 = 1 − 𝑑𝑜𝑓 translational (prismatic) joint

6
 Direct Kinematic Map

• The structure of the direct kinematics function depends

on the chosen 𝑟
𝑟 = 𝑓𝑟 (𝑞)
• Methods for computing 𝑓r (𝑞)
• geometric/by inspection
• systematic: assigning frames attached to the robot
links and using homogeneous transformation matrices

7
 Direct Kinematic Map
Direct kinematics of 2R planar robot (just using inspection…)
𝑦
𝑃
𝑝𝑦
∙ 𝑞1
𝑞= 𝑞 𝑛=2
𝑙2 2
𝑞2

𝑙1 𝑝𝑥
𝑞1
𝑟 = 𝑝𝑦 𝑚=3
𝑝𝑥 𝑥
𝜙

𝑝𝑥 = 𝑙1 cos 𝑞1 + 𝑙2 cos 𝑞1 + 𝑞2
for more general cases,
𝑝𝑦 = 𝑙1 sin 𝑞1 + 𝑙2 sin 𝑞1 + 𝑞2 we need a ‘method’!
𝜙 = 𝑞1 + 𝑞2

8
 Direct Kinematic Map
Numbering links and joints

joint 2 joint 𝑖 + 1 joint 𝑛

link 𝑖 − 1
link 1
link 𝑖 link 𝑛
joint 1
joint 𝑖 − 1
joint 𝑖
link 0
(end effector)
(base)

icon representation of joint types

for the manipulator skeleton

revolute prismatic

J. Denavit and R.S. Hartenberg, “A kinematic notation for lower-pair mechanisms based on matrices,” Trans. ASME J. Applied Mechanics, 23: 215–221, 1955

9
 Denavit-Hartenberg (DH) Layout
Spatial relation between joint axes

axis of joint 𝑖 axis of joint 𝑖 + 1

90°
𝜋
𝛼𝑖
𝐴

90°
common normal (axis
𝐵 of link 𝑖)

𝑎i = displacement 𝐴𝐵 between joint axes (always well defined)

with sign
𝛼𝑖 = twist angle between joint axes
(pos/neg)!
— projected on a plane 𝜋 orthogonal to the link axis

10
 Denavit-Hartenberg (DH) Layout
Spatial relation between link axes

axis of joint 𝑖
link 𝑖 − 1

link 𝑖

𝐷
axis of link 𝑖
axis of link 𝑖 − 1 𝐶

𝜎
𝜃𝑖

𝑑i = displacement 𝐶𝐷 (a variable if joint 𝑖 is prismatic)

𝜃𝑖 = angle between link axes (a variable if joint 𝑖 is revolute) with sign
— projected on a plane 𝜎 orthogonal to the joint axis (pos/neg)!

11
 Denavit-Hartenberg (DH) Layout
Denavit-Hartenberg (DH) frame
joint axis 𝑖
joint axis
𝑖−1
link 𝑖 − 1
link 𝑖 joint axis
𝑖+1
link 𝑖 − 1 is moved by
joint 𝑖 − 1 frame R𝐹𝑖 is
attached to link 𝑖
𝛼𝑖
𝑧𝑖
𝑦𝑖
𝑑𝑖 𝑧𝑖−1 𝑎𝑖
𝑦𝑖−1 𝑂𝑖 𝑥𝑖
𝑥𝑖−1
𝜃𝑖
𝑂𝑖−1
axis of link 𝑖
axis of joint 𝑖 around as common normal to
which link 𝑖 rotates or joint axes 𝑖 and 𝑖 + 1
along which link 𝑖 slides
12
 Denavit-Hartenberg (DH) Layout
Definition of DH parameters
joint axis
joint axis joint axis 𝑖 link 𝑖 𝑖 + 1
𝑖 − 1 link 𝑖 − 1

𝛼𝑖
𝐷
•
𝑧𝑖
𝑧𝑖−1 𝑎i
𝑑𝑖
𝑥𝑖−1 𝑂𝑖 𝑥𝑖
𝜃𝑖

• unit vector 𝑧𝑖 along axis of joint 𝑖 + 1

• unit vector 𝑥𝑖 along the common normal to joint 𝑖 and 𝑖 + 1 axes (𝑖 → 𝑖 + 1)
• 𝑎i = displacement 𝐷𝑂𝑖 , + if oriented as 𝑥i, always constant (= ‘length’ of link 𝑖)
• 𝑑i = displacement 𝑂𝑖 −1 𝐷, + if oriented as 𝑧 𝑖 −1 . , variable if joint 𝑖 is PRISMATIC
• 𝛼𝑖 = twist angle from 𝑧 𝑖 −1 to 𝑧 𝑖 around 𝑥 𝑖 , + if counterclockwise, always constant
• 𝜃𝑖 = angle from 𝑥 𝑖 −1 to 𝑥 𝑖 around 𝑧 𝑖 −1 , + if counterclockwise, variable if joint 𝑖 is REVOLUTE
13
 Denavit-Hartenberg (DH) Layout
DH layout made simple (a popular 3-minute illustration)

https://www.youtube.com/watch?v=rA9tm0gTln8
• note: the author of this video uses 𝑟 in place of 𝑎, and does not add subscripts!

14
 Ambiguities in Defining DH Frames
• frame 0 : origin and 𝑥0 axis are arbitrary
• frame 𝑛 : 𝑧𝑛 axis is not specified
• however, 𝑥𝑛 must intersect and be chosen orthogonal to 𝑧𝑛−1

• positive direction of 𝑧𝑖−1 (up/down on axis of joint 𝑖) is arbitrary

• choose one, and try to ‘avoid flipping over’ to the next one

• positive direction of 𝑥𝑖 (back/forth on axis of link 𝑖) is arbitrary

• if successive joint axes are incident, we often take 𝑥𝑖 = 𝑧𝑖−1 × 𝑧𝑖

• when natural, follow the direction ‘from base to tip’

• if 𝑧𝑖 and 𝑧𝑖−1 are parallel (common normal not uniquely defined)

• 𝑂𝑖 chosen arbitrarily along 𝑧𝑖 , still trying to ‘zero out’ parameters

• if 𝑧𝑖−1 and 𝑧𝑖 are coincident, normal 𝑥𝑖 axis can be chosen at will

• this case occurs only if the two joints are of different kind (𝑃/𝑅 or 𝑅/𝑃)

• again, try using ‘simple values’ (e.g., 0 or ±𝜋/2) for constant angles

15
 Homogeneous Transformation
Homogeneous transformation (between successive DH frames)
(two steps from i-1 to frame i)

• Step 1: roto-translation (screw motion) around and along 𝑧𝑖−1

cos 𝜃𝑖 − sin 𝜃𝑖 0 0 1 0 0 0 cos 𝜃𝑖 − sin 𝜃𝑖 0 0
𝑖−1 𝐴 ′ sin 𝜃𝑖 cos 𝜃𝑖 0 0 0 1 0 0 sin 𝜃𝑖 cos 𝜃𝑖 0 0
𝑖 𝑞i = =
0 0 1 0 0 0 1 𝑑i 0 0 1 𝑑𝑖
0 0 0 1 0 0 0 1 0 0 0 1
the product of these two matrices commutes as both around same axis!

rotational joint ⇒ 𝑞𝑖 = 𝜃𝑖 prismatic joint ⇒ 𝑞𝑖 = 𝑑i

• Step 2: roto-translation (screw motion) around and along 𝑥𝑖

1 0 0 𝑎i 1 0 0 0 1 0 0 𝑎𝑖
always a
𝑖−1 𝐴 = 0 1 0 0 0 cos 𝛼𝑖 − sin 𝛼𝑖 0 0 cos 𝛼𝑖 − sin 𝛼𝑖 0
𝑖 = constant matrix
0 0 1 0 0 sin 𝛼𝑖 cos 𝛼𝑖 0 0 sin 𝛼𝑖 cos 𝛼𝑖 0
0 0 0 1 0 0 0 1 0 0 0 1
the product of these two matrices commutes as both around same axis!
16
 Homogeneous Transformation
Denavit-Hartenberg matrix

J. Denavit and R.S. Hartenberg, “A kinematic notation for lower-pair mechanisms based on matrices,”
Trans. ASME J. Applied Mechanics, 23: 215–221, 1955

cos 𝜃𝑖 − cos 𝛼𝑖 sin 𝜃𝑖 sin 𝛼𝑖 sin 𝜃𝑖 𝑎𝑖 cos 𝜃𝑖

𝑖−1 𝐴 𝑖−1 𝐴 ′ 𝑖−1 sin 𝜃𝑖 cos 𝛼𝑖 cos 𝜃𝑖 − sin 𝛼𝑖 cos 𝜃𝑖 𝑎𝑖 sin 𝜃𝑖
𝑖 𝑞𝑖 = 𝑖 𝑞𝑖 . 𝐴𝑖 =
0 sin 𝛼𝑖 cos 𝛼𝑖 𝑑𝑖
0 0 0 1

compact notation: 𝑐 = 𝑐os, 𝑠 = sin

super-compact notation (if feasible): 𝑐𝑖 = cos 𝑞i , 𝑠i = sin 𝑞𝑖
𝑐𝑖𝑗 = cos 𝑞𝑖 + 𝑞𝑗 , 𝑠𝑖𝑗 = sin 𝑞𝑖 + 𝑞𝑗
…
17
 Direct Kinematics
Direct kinematics of robot manipulators

𝑦𝐸 slide 𝑠
description ‘internal’ 𝑅𝐹𝐸
to the robot using • 𝑞 = (𝑞1 , … , 𝑞𝑛 )
• product of DH matrices 𝑧𝐸 approach 𝑎
0𝐴1 𝑞1 1 𝐴2 𝑞2 ⋯𝑛−1 𝐴𝑛 𝑞𝑛 𝑥𝐸 normal 𝑛
𝑧0
𝑅𝐹0
description ‘external’ to
𝑦0 the robot using

𝑅𝐹𝑤 𝑥0 𝑤𝑅
𝐸 = n s r
𝑤𝑅 𝑤𝑝
𝑤𝑇 = 𝑤 𝑇 0𝐴 1
𝑞1 𝐴2 𝑞2 ⋯ 𝑛−1 𝐴𝑛 𝑛
𝑞𝑛 𝑇𝐸 𝑤𝑇 𝐸 𝐸
𝐸 0 1 • 𝐸 = 𝑇
𝑟 = 𝑓𝑟 (𝑞) 0 1
• 𝑟 = 𝑟1 , … , 𝑟𝑚
alternative representations of the direct kinematics

18
Direct Kinematics (SCARA)
DH assignment for a SCARA robot

𝑞4

𝑞1 𝑞2
𝑞3

Sankyo SCARA 8438 Sankyo SCARA SR 8447

19
 Direct Kinematics (SCARA)
Step 1: joint axes
j𝑜𝑖𝑛𝑡 𝑎𝑥𝑖𝑠 3,4
j𝑜𝑖𝑛𝑡 𝑎𝑥𝑖𝑠 1 j𝑜𝑖𝑛𝑡 𝑎𝑥𝑖𝑠 2

all parallel
(or coincident)

J3 prismatic
twist angles
J2 elbow ≡
𝛼𝑖 = 0 or 𝜋 J4 revolute
J1 shoulder

20
 Direct Kinematics (SCARA)
Step 2: link axes
j𝑜𝑖𝑛𝑡 𝑎𝑥𝑖𝑠 3,4
j𝑜𝑖𝑛𝑡 𝑎𝑥𝑖𝑠 1 j𝑜𝑖𝑛𝑡 𝑎𝑥𝑖𝑠 2

the vertical ‘heights’ of

the link axes
are arbitrary
(for the time being)

𝑎1 𝑎2 𝑎3 = 0

21
 Direct Kinematics (SCARA)
Step 3: frames
j𝑜𝑖𝑛𝑡 𝑎𝑥𝑖𝑠 3,4
j𝑜𝑖𝑛𝑡 𝑎𝑥𝑖𝑠 1 j𝑜𝑖𝑛𝑡 𝑎𝑥𝑖𝑠 2

axes 𝒚𝒊 for 𝑖 > 0 𝑧1

are not shown 𝑧2 = 𝑧3
(nor needed; they form 𝑥1
𝑥2
right-handed frames)
𝑥3
𝑥4
𝑧0
𝑧4 = 𝒂 axis
(approach)
𝑦0
𝑥0

22
 Direct Kinematics (SCARA)
Step 4: DH table of parameters
j𝑜𝑖𝑛𝑡 𝑎𝑥𝑖𝑠 3,4
j𝑜𝑖𝑛𝑡 𝑎𝑥𝑖𝑠 1 j𝑜𝑖𝑛𝑡 𝑎𝑥𝑖𝑠 2
𝑖 𝛼𝑖 𝑎𝑖 𝑑𝑖 𝜃𝑖
1 0 𝑎1 𝑑1 𝑞1

2 0 𝑎2 0 𝑞2

𝑧1 3 0 0 𝑞3 0
𝑧2 = 𝑧3
𝑎1 𝑥1
4 𝜋 0 𝑑4 𝑞4
𝑥2
𝑎2
𝑞3
𝑑4 note that
𝑑1 𝑥4 𝑥3 • 𝑑1 and 𝑑4 could be set = 0
𝑧0 𝑧4 • 𝑑4 < 0 (opposite to 𝒛3 )
•also, 𝑞3 < 0 in this configuration
𝑥0 𝑦0

23
 Direct Kinematics (SCARA)
Step 5: DH transformation matrices

𝑐𝜃1 −𝑠𝜃1 0 𝑎1 𝑐𝜃1 1 0 0 0

0 𝑠𝜃1 𝑐𝜃1 0 𝑎1 𝑠𝜃1 0 1 0 0
𝐴1 𝑞1 = 2
𝐴3 𝑞3 =
0 0 1 𝑑1 0 0 1 𝑑3
0 0 0 1 0 0 0 1

𝑐𝜃2 −𝑠𝜃2 0 𝑎2 𝑐𝜃2 𝑐𝜃4 𝑠𝜃4 0 0

1𝐴 𝑠𝜃2 𝑐𝜃2 0 𝑎2 𝑠𝜃2 𝑠𝜃4 −𝑐𝜃4 0 0
2 𝑞2 = 3𝐴
0 0 1 0 4 𝑞4 =
0 0 −1 𝑑4
0 0 0 1
0 0 0 1

𝑞 = 𝑞1 , 𝑞2 , 𝑞3 , 𝑞4 = 𝜃1 , 𝜃2 , 𝑑3 , 𝜃4

24
 Direct Kinematics (SCARA)
Step 6a: direct kinematics (homogeneous matrix 𝒘
𝑻𝑬 as
p r o du ct o f t h e 𝒊−𝟏 𝑨 𝒊 ( 𝒒 𝒊 ) ′ 𝒔 )

𝑐12 −𝑠12 0 𝑎1 𝑐1 + 𝑎2 𝑐12 𝑐12 −𝑠12 0 𝑎1 𝑐1 + 𝑎2 𝑐12

0 𝑠12 𝑐12 0 𝑎1 𝑠1 + 𝑎2 𝑠12 0 𝑠12 𝑐12 0 𝑎1 𝑠1 + 𝑎2 𝑠12
𝐴2 𝑞1 , 𝑞2 = 𝐴3 𝑞1 , 𝑞2 , 𝑞3 =
0 0 1 𝑑1 0 0 1 𝑑1 + 𝑞3
0 0 0 1 0 0 0 1

𝑤𝑇 =4 𝑇𝐸 = 𝐼
0
𝑅 𝑞1 , 𝑞2 , 𝑞4 = 𝑛 𝑠 𝑎 𝑝 = 𝑝 𝑞1 , 𝑞2 , 𝑞3
𝑐124 𝑠124 0 𝑎1 𝑐1 + 𝑎2 𝑐12
𝑤𝑇 0𝐴 𝑠124 −𝑐124 0 𝑎1 𝑠1 + 𝑎2 𝑠12
𝐸 = 4 𝑞1 , 𝑞2 , 𝑞3 , 𝑞4 =
0 0 −1 𝑑1 + 𝑞3 + 𝑑4
0 0 0 1

25
 Direct Kinematics (SCARA)
Step 6b: direct kinematics (as task vector 𝒓)

𝑐124 𝑠124 0 𝑎1 𝑐1 + 𝑎2 𝑐12

𝑤 0 𝑠124 −𝑐124 0 𝑎1 𝑠1 + 𝑎2 𝑠12
𝑇𝐸 = 𝐴4 𝑞1 , 𝑞2 , 𝑞3 , 𝑞4 =
0 0 −1 𝑑1 + 𝑞3 + 𝑑4
0 0 0 1

extract 𝛼𝑧 4
take 𝑝 ∈ ℝ as such from
from 𝑅(𝑞1 , 𝑞2 , 𝑞4 ) 𝑝(𝑞1 , 𝑞2 , 𝑞3 )

𝑝𝑥 𝑎1 𝑐1 + 𝑎2 𝑐12
𝑝𝑦 𝑎1 𝑠1 + 𝑎2 𝑠12 4
𝑟 = 𝑝 = 𝑓𝑟 (𝑞) = ∈ℝ
𝑧 𝑑1 + 𝑞3 + 𝑑4
𝛼𝑧 𝑞1 + 𝑞2 + 𝑞4

26
 Direct Kinematics (2R PR)
Direct kinematics of 2R planar robot (using DH frame assignment…)
𝑦0 𝑦2 𝑥2 𝑐𝜃 1 −𝑠𝜃 1 0 𝑙1 𝑐𝜃 1
𝑃 𝜙 0 𝑠𝜃1 𝑐𝜃1 0 𝑙1 𝑠𝜃1
𝐴1 (𝜃1 ) =
𝑝𝑦 • 0 0 1 0
𝑦1 𝑎 2 = 𝑙 2
0 0 0 1
𝑙𝑖𝑛𝑘 2 𝜃2 = 𝑞2 𝑐𝜃2 −𝑠𝜃2 0 𝑙2 𝑐𝜃2
𝑥1 1 𝑠𝜃 𝑐𝜃 0 𝑙 𝑠𝜃
𝑎1 = 𝑙1 𝐴2 (𝜃2 ) = 2 2 2 2
𝑗𝑜𝑖𝑛𝑡 2 0 0 1 0
𝑙𝑖𝑛𝑘 1 𝑝𝑥 𝑥0
𝑗𝑜𝑖𝑛𝑡 1 𝜃1 = 𝑞1 0 0 0 1
𝑐12 −𝑠12 0 𝑙1 𝑐1 + 𝑙2 𝑐12
𝑠12 𝑐12 0 𝑙1 𝑠1 + 𝑙2 𝑠12
𝑙𝑖𝑛𝑘 0 𝐴2 (𝑞) =
0 0 1 0
0 0 0 1
𝑧0 , 𝑧1 , 𝑧2 outgoing from plane
𝑝𝑥 0 𝑙1 𝑐1 + 𝑙2 𝑐12
𝑖 𝛼𝑖 𝑎𝑖 𝑑𝑖 𝜃𝑖 𝑝𝑦
𝑝 0 𝑙1 𝑆1 + 𝑙2 𝑆12
= = 0 𝐴2 (𝑞) =
1 0 𝑙𝑖 0 𝑞𝑖 1 0 0 0
1 1 1
2 0 𝑙2 0 𝑞2
𝜙 = 𝑞1 + 𝑞2 (extracted from 0 𝑅2 (𝑞))

27
 Direct Kinematics (2R PR)
Direct kinematics of 2R planar robot (TCP location on the robot end effector)

𝑦0 𝑦2 𝑥2 𝑖 𝛼𝑖 𝑎𝑖 𝑑𝑖 𝜃𝑖
𝑃 𝑦𝐸 1 0 𝑙𝑖 0 𝑞𝑖
• •
𝑦1 𝑙2 𝑇𝐶𝑃 𝑧𝐸 2 0 𝑙2 0 𝑞2
𝑞2
𝑥1
𝑙1 𝑐12 −𝑠12 0 𝑙1 𝑐1 + 𝑙2 𝑐12
𝑞1 𝑥0 0 𝑠12 𝑐12 0 𝑙1 𝑠1 + 𝑙2 𝑠12
𝐴2 (𝑞) =
0 0 1 0
0 0 0 1
Tool Center Point 𝑇𝐶𝑃 and associated end-effector frame R𝐹𝐸

2 0 2
0 1 0 𝑇𝐶𝑃𝑥 𝑇𝐶𝑃𝑥 𝑞 𝑇𝐶𝑃𝑥 0
2 0 0 2 0
2
𝑇𝐸 = 0 0 −1 𝑇𝐶𝑃𝑦 𝑇𝐶𝑃(𝑞) = 𝑇𝐶𝑃𝑦 𝑞 = 0 𝐴2 𝑞 𝑇𝐶𝑃𝑦 = 0 𝐴 𝑞 = 0 𝐴2 (𝑞)2 𝑇𝐸
2
−1 0 0 0 1 0 0 0
0 0 0 1 1 1 1

28
 Direct Kinematics (Stanford)
Stanford manipulator
• 6-dof: 2R-1P-3R (spherical wrist)

• robot with shoulder offset

• ‘one possible’ DH assignment of frames is
shown
• determine the associated
• table of DH parameters
• homogeneous transformation matrices
• direct kinematics
• write a program for computing the direct
kinematics
• numerically (Matlab), given a 𝒒
• symbolically (Mathematica, Maple,
Symbolic Manipulation Toolbox of
Matlab, …)

29
 Direct Kinematics (Stanford)
DH table for Stanford manipulator
• 6-dof: 2R-1P-3R (spherical wrist)

𝑖 𝛼𝑖 𝑎𝑖 𝑑𝑖 𝜃𝑖
1 −𝜋/2 0 𝑑1 > 0 𝑞1 = 0

2 𝜋/2 0 𝑑2 > 0 𝑞2 = 0

3 0 0 𝑞3 > 0 −𝜋/2

4 −𝜋/2 0 0 𝑞4 = 0

joint variables are in red, while their values 5 𝜋/2 0 0 𝑞5 = −𝜋/2

in the robot configuration shown are in blue
6 0 0 𝑑6 > 0 𝑞6 = 0

30
Robot Introduction for Projects
The robot model is Mintasca’s GLUON-6L3

GLUON-6L3
Specification Maximum range of 425
motion (mm)
Degree of Freedom 6
Workspace Joint Limits Joint 1 (°) -140~140
Joint 2 (°) -90~90
Joint 3 (°) -140~140
Joint 4 (°) -140~140
Joint 5 (°) -140~140
Joint 6 (°) -360~360
Maximum Joint Speed Joint 1 (°/𝒔) 302
Joint 2 (°/𝒔) 302
Joint 3 (°/𝒔) 302
Joint 4 (°/𝒔) 302
Joint 5 (°/𝒔) 302
Joint 6 (°/𝒔) 302
Working Environment Voltage (V) 42
Watt (W) Around 120
Working Heat (°C) 10 - 50
31
 Preparation

Note that the frame i, is located on joint i+1 The robot arm model diagram

32
 Create Standard DH Table
Certain Cases

𝑧5 • For Frame 0, only the direction of axis 𝑧0 is

arbitrarily chosen as upwards; then 𝑂0 and 𝑥0
can be arbitrarily chosen.
𝑧3
• For Frame n, since there is no Joint n+ 1, 𝑧𝑛 is
𝑧4
not uniquely defined while 𝑥𝑛 has to be normal
to axis 𝑧𝑛−1 . Typically, Joint n is revolute, and
𝑧2 The method thus 𝑧𝑛 can be aligned with the direction of
𝑧𝑛−1 . The common normal between two lines
• Choose axis 𝑧𝑖 along the axis of Joint i + 1. is the line containing the minimum distance
segment between the two lines.

• When two consecutive axes are parallel, the

𝑧1 common normal between them is not uniquely
defined.

𝑧0 • When two consecutive axes intersect, the

positive direction of 𝑥𝑖 is arbitrary.

• When Joint i is prismatic, only the direction of

𝑧𝑖−1 is specified.

33
 Create Standard DH Table
Certain Cases

𝑧5 • For Frame 0, only the direction of axis 𝑧0 is

arbitrarily chosen as upwards; then 𝑂0 and 𝑥0
𝑂5 can be arbitrarily chosen.
𝑧3
𝑂3
𝑂4 • For Frame n, since there is no Joint n+ 1, 𝑧𝑛 is
𝑧4
The method not uniquely defined while 𝑥𝑛 has to be normal
to axis 𝑧𝑛−1 . Typically, Joint n is revolute, and
𝑧2 • Choose axis 𝑧𝑖 along the axis of Joint i + 1. thus 𝑧𝑛 can be aligned with the direction of
𝑧𝑛−1 . The common normal between two lines
𝑂2 • Find the origin 𝑂𝑖 at the intersection of axis 𝑧𝑖 is the line containing the minimum distance
with the common normal to axes 𝑧𝑖−1 and 𝑧𝑖 . segment between the two lines.

• When two consecutive axes are parallel, the

common normal between them is not uniquely
𝑂1 defined.
𝑂0
𝑧0 • When two consecutive axes intersect, the
positive direction of 𝑥𝑖 is arbitrary.

• When Joint i is prismatic, only the direction of

𝑧𝑖−1 is specified.

34
 Create Standard DH Table
Certain Cases

𝑥5 • For Frame 0, only the direction of axis 𝑧0 is

arbitrarily chosen as upwards; then 𝑂0 and 𝑥0
𝑥3 can be arbitrarily chosen.
𝑧3 𝑥4
The method
𝑂3
𝑂4 • For Frame n, since there is no Joint n+ 1, 𝑧𝑛 is
𝑧4
not uniquely defined while 𝑥𝑛 has to be normal
• Choose axis 𝑧𝑖 along the axis of Joint i + 1.
to axis 𝑧𝑛−1 . Typically, Joint n is revolute, and
𝑥2 𝑧2 thus 𝑧𝑛 can be aligned with the direction of
• Find the origin 𝑂𝑖 at the intersection of axis 𝑧𝑖
𝑧𝑛−1 . The common normal between two lines
with the common normal to axes 𝑧𝑖−1 and 𝑧𝑖 .
is the line containing the minimum distance
Additionally, locate 𝑂𝑖′ at the intersection of
segment between the two lines.
the common normal with axis 𝑧𝑖−1 .
𝑥1 • When two consecutive axes are parallel, the
• Choose axis 𝑥𝑖 along the common normal to
common normal between them is not uniquely
𝑂1 axis 𝑧𝑖−1 and 𝑧𝑖 with positive direction pointing
defined.
𝑂0 𝑥0 towards Joint i + 1 from joint i.
𝑧0 • When two consecutive axes intersect, the
positive direction of 𝑥𝑖 is arbitrary.

• When Joint i is prismatic, only the direction of

𝑧𝑖−1 is specified.

35
 Create Standard DH Table
Certain Cases

𝑥5 • For Frame 0, only the direction of axis 𝑧0 is

arbitrarily chosen as upwards; then 𝑂0 and 𝑥0
𝑥3 The method can be arbitrarily chosen.
𝑧3 𝑥4
𝑂3
𝑂4 • Choose axis 𝑧𝑖 along the axis of Joint i + 1. • For Frame n, since there is no Joint n+ 1, 𝑧𝑛 is
𝑧4
not uniquely defined while 𝑥𝑛 has to be normal
• Find the origin 𝑂𝑖 at the intersection of axis 𝑧𝑖 to axis 𝑧𝑛−1 . Typically, Joint n is revolute, and
𝑥2 𝑧2 with the common normal to axes 𝑧𝑖−1 and 𝑧𝑖 . thus 𝑧𝑛 can be aligned with the direction of
Additionally, locate 𝑂𝑖′ at the intersection of 𝑧𝑛−1 . The common normal between two lines
the common normal with axis 𝑧𝑖−1 . is the line containing the minimum distance
segment between the two lines.
• Choose axis 𝑥𝑖 along the common normal to
𝑥1 axis 𝑧𝑖−1 and 𝑧𝑖 with positive direction pointing • When two consecutive axes are parallel, the
towards Joint i + 1 from joint i. common normal between them is not uniquely
𝑂1 defined.
𝑂0 𝑥0 • Axis 𝑧𝑖 will be fixed as to complete a right-
𝑧0 handed frame. • When two consecutive axes intersect, the
positive direction of 𝑥𝑖 is arbitrary.

• When Joint i is prismatic, only the direction of

𝑧𝑖−1 is specified.

36
 Create Standard DH Table
Constructing the kinematics graph

𝑥5 The kinematics graph of the Gluon EduBot arm should look like this. The
parameters of the DH parameters are as follows:
𝑥3
𝑧3 𝑥4 • 𝑑𝑖 , coordinate of 𝑂𝑖′ along 𝑧𝑖−1
𝑂4 • 𝜃𝑖 , angle between axes 𝑥𝑖−1 and 𝑥𝑖 about axis 𝑧𝑖−1 to be taken positive
𝑂3 𝑧4
when rotation is made counter-clockwise
• 𝑎𝑖 , distance between 𝑂𝑖 and 𝑂𝑖′
𝑥2 𝑧2 • 𝛼𝑖 , angle between axes 𝑧𝑖−1 and 𝑧𝑖 about axis 𝑥𝑖 to be taken positive
when rotation is made counter-clockwise
The DH table

𝑥1
Joints 𝒅 (m) 𝜽 (𝒓𝒂𝒅) 𝒂 (m) 𝜶(𝒓𝒂𝒅)
1 0.1015 𝜋/2 0 𝜋/2
𝑂1
𝑂0 𝑥0 2 0.0 𝜋/2 0.173 0
𝑧0 3 0.0 0 0.173 𝜋
4 -0.07920 𝜋/2 0 𝜋/2
5 0.07920 0 0 𝜋/2
6 0 0 0 0

37
 Direct Kinematics
To visualize the forward kinematics, input a simple array of 𝜃 values with a
known pose value in space.

𝜋 𝜋
𝜃= 0, 0, , 0, 0,
2 2 This function then gives a homogeneous transformation matrix of:

This 𝜃 value shows a 90° rotation on joint 3

and 6, hence the Edubot arm should bend
into the paper with half of its height aligned
with negative y-axis

38
 Direct Kinematics
Run the homogeneous transformation matrix through a defined
function to calculate the pose value

0.08011
−0.2545
0.2795
𝜋
Pose Value = −
2
0
𝜋
−
2

39
 FK Demonstration
To further visualize and understand the forward kinematics derived from the DH
parameters, a simulation can be run to check the values of 𝜃

40
 Modified DH
Do not be confused!

• Both are used in textbooks.

• We use Standard DH in ENGG5402 to keep the consistence.
41
 Modified DH
• a modified version introduced in J. Craig’s book “Introduction to Robotics” (1986) and
aligned for the indexing by Khalil and Kleinfinger (ICRA, 1986)
• has 𝑧𝑖 axis on joint 𝑖
• 𝑎𝑖 &𝛼𝑖 =distance & twist angle from 𝑧𝑖−1 to 𝑧𝑖 , measured along & about 𝑥𝑖−1
• 𝑑𝑖 &𝜃𝑖 = distance & angle from 𝑥𝑖−1 to 𝑥𝑖 , measured along & about 𝑧𝑖
• source of much confusion... if you are not aware of it (or don’t mention it!)
• convenient with link flexibility: a rigid frame at the base, another at the tip...

𝑥2
𝑐𝜃𝑖 −𝑐𝛼𝑖 𝑠𝜃𝑖 𝑠𝛼𝑖 𝑠𝜃𝑖 𝑎𝑖 𝑐𝜃𝑖 𝑦2
classical planar 2R
𝑖−1 𝑠𝜃𝑖 𝑐𝛼𝑖 𝑐𝜃𝑖 −𝑠𝛼𝑖 𝑐𝜃𝑖 𝑎𝑖 𝑠𝜃𝑖 𝑦1
(or distal) 𝐴𝑖 = example
0 𝑠𝛼𝑖 𝑐𝛼𝑖 𝑑𝑖 𝑦0
0 0 0 1
𝑥1
𝑥0
𝑐𝜃𝑖 −𝑠𝜃𝑖 0 𝑎𝑖 𝑦2
𝑖−1 𝑚𝑜𝑑 𝑐𝛼𝑖 𝑠𝜃𝑖 𝑐𝛼𝑖 𝑐𝜃𝑖 −𝑠𝛼𝑖 −𝑑𝑖 𝑠𝛼𝑖 𝑦0 𝑥2
modified 𝐴𝑖 = 𝑦1
(or proximal) 𝑠𝛼𝑖 𝑠𝜃𝑖 𝑠𝛼𝑖 𝑐𝜃𝑖 𝑐𝛼𝑖 𝑑𝑖 𝑐𝛼𝑖 𝑥1
0 0 0 1 𝑥0
modified DH tends to place frames ‘at
the base’ of each link
42
Q&A