Robot Kinematics and Dynamics

Robotics. Robot Design and Modeling

CONTENTS
1. Objectives

2. Kinematics

3. Dynamics

4. Underactuated Robots

5. Redundant Robots

6. Summary

7. Bibliography
 Robot Kinematics and Dynamics | 3

Objectives • The direct kinematic problem: how to solve it?

To solve the problem of direct kinematics, elements
• To identify the mathematical foundations for of vector and matrix algebraic calculus are usually
kinematic and dynamic modeling of robots. used to represent and describe the location of an
• To understand the main characteristics of object in three-dimensional space with respect to a
underactuated and redundant robots. fixed reference system.
• To implement a framework for kinematic and In principle, to solve the direct kinematic
dynamic robot modeling. calculation, it is proposed to “find a homogeneous
transformation matrix T that relates the position

Kinematics and orientation of the robot tip with respect to the

fixed reference system located at the base of the
robot” [1].
In robotics, kinematics studies the movement
of robots with respect to a reference system. The direct kinematic problem is solved by means
This is why robot kinematics refers to the study of how of homogeneous transformation matrices. In this
they move and position themselves in space, describing sense, what is the procedure to be followed?
the movements and positions of the components of a Accordingly, the following is described [3]:
robot, as well as their joints, as a function of time and
a. Each link is matched to a reference system of
especially of the relationships established between the
solidarity
position and orientation of the robot’s end end [1]. Based
on the above, in practice there are two fundamental b. Allows to represent the relative translations and
problems in the resolution of problems of the kinematics rotations between different links.
of a robot. c. The matrix i-1 Ai represents the relative position
The first of these is known as the direct kinematic and orientation between the systems associated
problem and concerns the calculation of the position to two consecutive links of the robot.
and orientation of its robot end-effector frame (usually
the tool or part interacting with the environment) from
its joint coordinates θ of the joints and other robot
parameters. Direct kinematics is essential for tasks such 0
A3 = 0A1 1A22A3
as precise positioning and control of a robot in a given
environment.

On the other hand, inverse kinematics is used to T = 0A6 = 0A11A22A33A4 4A55A6

determine the values that the robot’s joint coordinates
must adopt in order to position and orient the robot’s tip In this order of ideas, it should be noted that the
according to a given spatial location [2] Fig. 1. matrix T is directly related to all the degrees of
freedom of the matrix A0n .
To solve this type of problem, kinematic methods
are used to locate the associated coordinate
systems in order to obtain the kinematic chain of
the robot. The Denavit-Hastenberg (D-H) modeling
is used for this type of problem.
The main characteristics of this method are related
to the possibility of systematizing the obtaining of the
matrices between the different links of the kinematic
Fig. 1. The two fundamental problems to be solved in the robot
chain. To do this, these coordinate systems associated
kinematics [2]​
with each link must be chosen according to specific
conditions.
Kinematics in robotics is applied in a wide variety of
applications, from industrial robots assembling products
on production lines to autonomous robots moving in
unknown environments, such as autonomous vehicles or
drones.

In this vein, understanding the kinematics of a robot is

essential to design, program and precisely control its
movements, allowing robots to perform specific tasks
efficiently and safely.
 Robot Kinematics and Dynamics | 4

Now, this method allows the realization of four

simple movements or transformations (rotations Dynamics
and translations), involving a D-H parameter ( qi , di , Although kinematics deals with the motion of robots,
ai , aj ) as shown in the following equation: robot dynamics focuses on knowing the relationship
between the robot’s motion and the forces involved in it.

i -1
A = T (z,qi) T([0,0,di ])T([ai ,0,0]) T(x,ai ) The dynamic model mathematically relates the elements
related to the forces acting on a body and its motion.
This model is usually solved when the structure has one
When performing the product between matrices we or two degrees of freedom (GDL), however, when the
obtain the following: number of degrees of freedom increases, the approach
and the obtaining of the dynamic model becomes more

[ ][ ][ ][ ]
cqi -sqi 0 0 1 0 0 0 1 0 0 0 1 0 0 0 complicated, so that many times it is not possible to
sqi cqi 0 0 0 1 0 0 0 1 0 0 0 cαi -sαi 0
i-1
A= 0 0 1 0 0 0 1 di 0 0 1 ai 0 sαi cαi 0 obtain a dynamic model that can be expressed in a
0 0 0 1 0 0 0 1 0 0 0 1 0 0 0 1
closed form.

Now, when the degree of complexity increases, the

dynamic model must be solved iteratively using

[ ]
cqi -cαi sqi sαi cqi αi cqi numerical procedures. As can be expected, obtaining a
i-1
A=
sqi cαicqi -sαicqi αisqi dynamic model of a robot is a complex aspect and often
0 sαi cαi di
0 0 0 1 tends to be overlooked.

In practice, it is essential to perform tasks such as

Here it is important to note that the Denavit - simulation of robot motion, design and evaluation of the
Hartenberg (D-H) method is characterized by the mechanical structure of a robot, sizing of actuators or
fact that it allows the passage from one link to the design and evaluation of dynamic control of the robot [2].
next by means of basic transformation matrices,
which depend exclusively on the constructive Obtaining the dynamic model of a robot is based on
characteristics of the robot [3]. the principles of the force balance approach set out
The basic transformations referred to are as follows in Newton’s second law, or also in its equivalent for
[1]: rotational movements, known as Euler’s law:
• Rotation θi around the Z-axisi -1 From the approach of the equilibrium of forces and
• Translation di around the Z axisj -1 torques that intervene on the operation of the robot, two
• Translation ai along Xi fundamental models can be obtained [2]:
• Rotation αi around the X axisi Fig. 2.

i-1
A = T (z, θi )T([0,0, di ])T([αi, 0,0])T (x, αi ) ΣF = m. a ΣΤ = Ιω + ωx (Ιω)

• The direct dynamic model: this model focuses

[ ]
cθi -cai sθi sαi sθi αi cθi on the temporal evolution of the robot’s joint
i-1
A=
sθi cαicθi -sαicθi αisθi coordinates as a function of the forces and torques
0 sαi cαi di
0 0 0 1 involved in the robot’s motion.
• The inverse dynamic model: which expresses the
forces and torques involved as a function of the
evolution of the articulated coordinates and their
derivatives.

Fig. 2. Denavit - Hartenberg method (D-H) [3]

 Robot Kinematics and Dynamics | 5

On the other hand, the equations of motion dynamics

for solving robot dynamics problems can be formulated Underactuated Robots
through several methods, among which the following Underactuated robots are a type of robots whose
can be mentioned [4]: mechanical structures have fewer degrees of freedom
• Lagrange equations of motion: The main (GDL) than necessary to fully control their limbs or
advantage of using this method is the possibility of moving parts.
eliminating the constraint forces from the dynamics This means that they have fewer actuators (such as
equations if the generalized coordinates are chosen motors or artificial muscles) than moving parts, resulting
independently. The elimination of this type of forces in some redundancy or lack of complete control over
is suitable for motion control and simulation and their motion [5].
can be recovered using Lagrange multipliers.
• The well-known Newton’s and Euler’s laws for Unlike traditional robots that use all motors to cancel
linear and rotational motions can be applied. These external forces to perform the programmed movements,
equations described together contain the applied underactuated robots are characterized by the fact that
forces and the constraint forces; thus, the constraint they do not rely solely on their motors to perform the
forces can be eliminated using the geometrical and activities, but allow external forces (such as gravity or
kinematic relations describing the nature of the inertia) to do part of the work.
constraints.
This results in some mathematical complexity of the
• Other methodologies applicable to solving robot modeling, since it is not possible to make simplifications
dynamics are D’ Alembert’s principle, Kane’s in which they are obtained as a result of linear equations,
equations of motion and those based on orthogonal so these external forces must be considered in the
complements of velocity constraint matrices. equations Fig. 3.
In the study of robotics dynamics it is important to
handle some important concepts and properties such as
the following:

• Center of mass: is a point in space that represents

the average location of all the mass of the object. It is
calculated by considering the distribution of masses
in the object and is used to describe the motion of
the object as if all its mass were concentrated at that
point.
• Moment of inertia: refers to the measure of
resistance that an object manifests when changing
its rotational speed around a specific axis, so the
greater the moment of inertia of an object relative
to an axis of rotation, the more difficult it will be to
accelerate or decelerate its rotation. The moment
of inertia depends on the mass distribution of the
object and the distance to the axis of rotation.
• Inertia tensioner: Unlike the moment of inertia,
which applies to rotation around a single axis, the
inertia tensor takes into account rotation in three
dimensions (around three orthogonal axes). It also
describes how the inertia is distributed in an object
in relation to the three principal axes of the body.
Fig. 3. Sub-actuated robot created by Boston Dynamics. Retrieved from:
[Link]
[Link]
 Robot Kinematics and Dynamics | 6

On the other hand, redundancy in underactuated robots

implies that there are more degrees of freedom in the Redundant Robots
system than actuators available to control them. This In robotics, as has been studied in this and other topics, it
can be advantageous in situations where adaptability is important to know the number of independent motions
and flexibility in object or task manipulation is required, or parameters that define the configuration of a robot,
which in turn allows the robot to adjust its motion known as degrees of freedom (GDL).
efficiently to adapt to changing conditions or to optimize
certain criteria, such as minimizing energy consumption In this order, for a robot to position and orient itself in a
or maximizing speed. three-dimensional space, it must have 6 GDL, whereby
three will be used to determine the position while the
• The main advantages that can be identified in this type remaining three will be used for orientation.
of robot are related to the following criteria:
• Lower energy consumption, due to fewer actuators Therefore, when the designed robot has more degrees
to carry out movements of freedom than minimally necessary to perform a given
task, these robots fall into the redundant category [6].
• Smaller motor sizes as torque and power
requirements are lower It should be clarified that in practice different
• They are usually compact (lighter) robots. classifications can be found depending on the number of
• They are more agile, have a higher load capacity degrees of freedom, or the number of joints, the following
and greater autonomy. being the most common:
• Can better adapt to dynamic and complex • Cyclic robots: these robots have more than 7 GDL
environments allowing you to put your tool in a given situation,
• They can perform a wide variety of tasks under different postures.
• Hyper-redundant robots: are those with a high
Underactuated robots are especially useful in applications
number of GDL, such as continuous robots or soft
that require interaction with the environment or
robots. ​[6]​.
adaptability, such as soft robotics and the manipulation
of delicate objects. They are also used in applications • Underactuated robots: are those that have fewer
such as biomedical robotics, mobile robotics and actuators than GDL, i.e., they are those that cannot
robotics for unstructured environments. control all their movements independently Fig. 4.

Examples of underactuated robots include snake robots,

tentacled robots, soft robots, and multi-limbed robotic
systems that can adapt to challenging environments
and perform tasks such as underwater exploration,
robot-assisted surgery, or search and rescue in disaster
situations.

Despite their advantages in terms of adaptability,

underactuated robots also face challenges, such as
difficulty in motion planning and the design of efficient
control algorithms. In addition, it should be noted that the
lack of actuators can limit the ability of these robots to
perform certain high-precision tasks.

In summary, underactuated robots are robotic systems

with fewer actuators than degrees of freedom,
giving them flexibility and adaptability in a variety of
applications. These robots often require more advanced
control and planning strategies to take full advantage of
their redundancy and achieve optimal performance in
specific tasks. Fig. 4. Classification of robots by number of degrees of freedom​[6]​.

Redundancy then, is related to the existence of multiple

solutions to perform a task, which allows the robot to
choose between different configurations to achieve the
same goal.
 Robot Kinematics and Dynamics | 7

This is an advantage of this type of redundant robots

that makes them more versatile and able to adapt to Summary
changing situations or diverse tasks. In that sense, the The motion and forces interacting in robotic systems
main advantages of redundancy are: are studied by kinematics and robot dynamics. This is
• Adaptability: redundant robots can modify their why kinematics deals with the description of position,
configuration to overcome obstacles or avoid velocity and acceleration in these systems, as well as
collisions, which is useful in complex work the relationship between the angles of the joints and the
environments. coordinates of the robot end.
• Precision: they can adjust their posture to reach Dynamics is responsible for analyzing the forces and
precise positions and orientations more easily. moments acting on each link, as well as the energy
• Better performance: by taking advantage of consumption and torque required by the joints.
redundancy, they can perform tasks more efficiently
In this sense, robots can be classified taking into
and smoothly, which is useful for delicate handling
account different factors, one of these is given by
tasks.
taking into account the number of degree of freedom.
• Avoiding singularities: Non-redundant robots often According to this classification, underactuated robots
have singular positions where certain joints become and redundant robots can be identified.
uncontrollable. Redundant robots can avoid these
singularities. An underactuated robot is one that has fewer degrees of
freedom than the workspace it accesses, i.e. it cannot control
Redundant robots are used in a variety of applications, all the variables of its motion.
including the following:
Finally, a redundant robot is one that has more degrees of
• Robotic surgery freedom than the workspace it accesses, i.e. it can adopt an
• Electronics manufacturing infinite number of configurations to reach the same position
• Welding and assembly tasks and orientation.
• Handling of fragile objects

Their versatility makes them suitable for jobs requiring Bibliography

precision and adaptability. Controlling redundant robots
can be more complex than non-redundant robots, so ​[1] Barriento, A. Peñín, F. Balaguer, C. Aracil, R.
techniques such as force control and task-based control Fundamentos de Robótica, Madrid: McGraw-Hill
are used to manage redundancy and achieve optimal Interamericana de España, 1997.
performance in a variety of situations. ​[2] Acosta S. Sigut, S. «Matemática y robótica,»
[Online]. Retrieved from: [Link]
Redundant robots are robotic systems with greater
es/sctm05/modulo2tf/4/[Link]. [Accessed:
freedom in performing a specific task, giving them
November 29, 2023].
advantages in terms of adaptability, precision
and obstacle avoidance. Their use is common in ​[3] Ramírez, L. Robótica. Modelado Cinemático de
applications where flexibility and precision are robots. 2021. [Online]. Retrieved from: https://
required, such as robot-assisted surgery or the [Link]/~hugo/robot/[Link]. [Accessed:
handling of delicate objects in manufacturing November 08, 2023].
environments.
​[4] Kumar, S. Introducción a la robótica, México D.F:
McGraw-Hill, 2010.
​[5] Megaz IEEE Ecuador, «Sistemas sub-actuados:
el futuro de la robótica,» 03 03 2022. [Online].
Retrieved from: [Link]
sistemas-sub-actuados/. [Accessed: November 29,
2023].
​[6] Martín, A. Terrile, S. Barrientos, A. del Cerro, J.
Robots Hiper-Redundantes: Clasificación, Estado
del Arte y Problemática. Revista Iberoamericana de
Automática e informática industrial, volumen 15,
pp. 351-362, 2018.