Machine Learning in Robotics

International Journal of Trend in Scientific Research and Development (IJTSRD)
Volume 6 Issue 3, March-April 2022 Available Online: www.ijtsrd.com e-ISSN: 2456 – 6470
@ IJTSRD | Unique Paper ID – IJTSRD49925 | Volume – 6 | Issue – 3 | Mar-Apr 2022 Page 2146
Machine Learning in Robotics
Matthew N. O. Sadiku1
, Uwakwe C. Chukwu2
, Abayomi Ajayi-Majebi3
, Sarhan M. Musa1
1
Roy G. Perry College of Engineering, Prairie View A&M University, Prairie View, TX, USA
2
Department of Engineering Technology, South Carolina State University, Orangeburg, SC, USA
3
Department of Manufacturing Engineering, Central State University, Wilberforce, OH, USA
ABSTRACT
The field of robotics is heterogeneous and multi-disciplinary. The
community of robotics consists of robotics engineers, biologists,
computer scientists, mathematicians, to mention a few. The goal of
robotics is to create intelligent machines (called robots) that behave
and think like humans. Machine learning in robots makes them more
efficient with self-learning ability to recognize the new objects. The
rapidly advancing capabilities of machine learning are helping robots
learn specific tasks. With the current rise of humanoid robots in
robotics research, the need for machine learning in robotics has
grown significantly. The paper provides various uses of machine
learning in robotics.
KEYWORDS: robots, robotics, machine learning, deep learning,
artificial intelligence
How to cite this paper: Matthew N. O.
Sadiku | Uwakwe C. Chukwu | Abayomi
Ajayi-Majebi | Sarhan M. Musa
"Machine Learning in Robotics"
Published in
International Journal
of Trend in
Scientific Research
and Development
(ijtsrd), ISSN: 2456-
6470, Volume-6 |
Issue-3, April 2022,
pp.2146-2150, URL:
www.ijtsrd.com/papers/ijtsrd49925.pdf
Copyright © 2022 by author (s) and
International Journal of Trend in
Scientific Research and Development
Journal. This is an
Open Access article
distributed under the
terms of the Creative Commons
Attribution License (CC BY 4.0)
(http://creativecommons.org/licenses/by/4.0)
INTRODUCTION
A robot is a system that contains sensors, control
systems, power supplies, and software, all working
together to perform a task. An autonomous robot has
a basic body structure (the chassis), sensors, a central
control system (microprocessor), actuators (motors), a
power supply and a software to control its behavior. It
functions as an intelligent machine, meaning that, it
can be programmed to take actions or make choices
based on input from sensors. In jobs with repetitive
and monotonous functions, robots might even
completely replace humans. Robotics and
autonomous systems is regarded as the fourth
industrial revolution [1].
Robots are commonly used in dangerous
environments where humans cannot survive situations
such as defusing bombs, finding survivors in unstable
ruins, welding, and exploring mines and shipwrecks.
Today, robots perform important functions in homes,
hospitals, industries, military institutions, education,
entertainment, and outer space. They are designed to
perform a wide variety of programmed tasks. Robots
can support, assist, and extend the services of
healthcare professionals. Robots are rapidly changing
the face of manufacturing. They may help
manufacturers increase precision, repeatability, and
productivity [2].
Both robotics and machine learning (ML) are
important components of artificial intelligence (AI).
The history of ML is tightly interwoven with the
history of AI. AI and ML are establishing the door to
entirelynew automation possibilities. The relationship
between AI and machine learning is illustrated in
Figure 1 [3]. Earlier robots were developed without
any artificial intelligence (AI) to perform only
repetitive tasks. Today, AI in getting integrated into
robots to develop the advanced level of robotics that
can perform multiple tasks and also learn new things
with a better perception of the environment. Robots
are developed and trained through machine learning in
order to improve robotics performance.
WHAT IS A ROBOT?
The word “robot” was coined by Czech writer Karel
Čapek in his play in 1920. Isaac Asimov coined the
IJTSRD49925

International Journal of Trend in Scientific Research and Development @ www.ijtsrd.com eISSN: 2456-6470
term “robotics” in 1942 and came up with three rules
to guide the behavior of robots [4]:
1. Robots must never harm human beings;
2. Robots must follow instructions from humans
without violating rule 1,
3. Robots must protect themselves without violating
the other rules.
Robotics has advanced and taken many forms
including fixed robots, collaborative robots, mobile
robots, industrial robots, medical robots, police
robots, military robots, officer robots, service robots,
space robots, social robots, personal robots, and
rehabilitation robots [5,6]. Robots are becoming
increasingly prevalent in almost every industry, from
healthcare to manufacturing.
Although there are many types of robots designed for
different environments and for different
purposes/applications, they all share four basic
similarities [7]: (1) All robots have some form of
mechanical construction designed to achieve a
particular task; (2) They have electrical components
which power and control the machinery; (3) All
robots must be able to sense its surroundings; a robot
may have light sensors (eyes), touch and pressure
sensors (hands), chemical sensors (nose), hearing and
sonar sensors (ears), etc. (4) All robots contain some
level of computer programming code. Programs are
the core essence of a robot since they provide
intelligence. There are three different types of robotic
programs: remote control, artificial intelligence, and
hybrid. Some robots are programmed to faithfully
carry out specific actions over and over again
(repetitive actions) without variation and with a high
degree of accuracy. Some advantages and
disadvantages of robots are shown in Figure 2 [8].
Robotics evolved to allow human- or animal-like
functions to be carried out by machines, with greater
strength, speed, and accuracy, in a roughly human- or
animal-shaped machine. Early robots were simple
mechanical automated machines. Modern robots
employ microprocessors and computer technology.
They can be programmed and “taught” to perform
certain tasks. They are taking on more “human” traits
such as sensing, dexterity, remembering, and
trainability.
BACKGROUND ON MACHINE LEARNING
Machine Learning is a branch of artificial intelligence
that helps in building automated systems that can learn
by themselves. It may be regarded as the study of
computer algorithms that can improve through
experience and by the use of data. It is the process of
training an AI model to make it intelligent enough to
perform specific tasks. All machine-learning
technologies basically involve taking in data, train a
model on that data, and then use the derived model to
make predictions on new data. Training a model is a
learning process where the model is exposed to
unfamiliar data at each step and is asked to make
predictions. The training is terminated when the
model’s predictions reach a point at which the error
does not improve [9]. At that stage, the machine no
longer needs continuous time-intensive teaching from
humans. Learning would occur as a consequence of
continuous use. Machine learning thrives on abundant
data and leans toward pattern recognition. It is self-
learning in nature and becomes “smarter” over time.
The majority of machine learning scenarios generally
fall into three categories: supervised learning,
unsupervised learning, or reinforcement learning. The
supervised learning task is to predict some additional
aspect of an input object. It is usually characterized
by using explicit examples with labels that the system
should use to learn to predict labels of previously
unseen data. Supervised learning problems, such as
regression and classification are more common in
robotics applications. In unsupervised learning, the
model is provided with unlabeled data and asked to
segment the elements into groups. The goal of
unsupervised learning is to find useful representations
of data based on regularities of data without labels.
Reinforcement learning (RL) enables a robot to
autonomously discover an optimal behavior through
trial-and-error interactions with its environment. It
embraces the full complexity of these problems by
requiring both interactive, sequential prediction as in
imitation learning as well as complex reward
structures.
Machine learning has been used successfully in
numerous other applications, from drones to assistive
robots. It is currently identified as one of the major
research areas in robotics. The key application of
machine learning in robotics is perception of the
environment, i.e., enable the robot to react
appropriately to the input from cameras and sensors or
recognize objects in the surrounding environment.
Machine learning mainly trains a robot to recognize
the wide-ranging objects visible in different shapes,
sizes , and various scenarios. It helps robots interact
dynamically and avoid obstacles. It is capable of
improving the efficiency in various fields such as pick
and place operations, drone systems, manufacturing
assembly, and quality control. It can be used to
translate text into another language instantaneously. In
a manufacturing setting, machine learning is
preventing factory machine breakdowns.
APPLICATIONS OF ROBOTS
Robots can perform feats no current machines can
accomplish. Robotics systems are being applied to

real-world situations. Machine learning is widely
used in diverse fields such as investment, banking and
finance, information technology, media and
entertainment, gaming, healthcare, military, and
automotive industry. The breadth of current and
potential applications of machine learning is vast. The
powerful combination of robotics and machine
learning opens the door to completely new
possibilities. We are yet to reach the full potential of
robotics and machine learning, but the current
applications are promising. Some of these
applications are discussed as follows.
Industrial Robots: This kind of robots are
utilized in industries for diverse purposes.
Machine learning is increasing the capabilities of
industrial robot systems. It will have a
transformative impact on industrial robots. It will
continue to push the boundaries of what is
possible with industrial robotic automation for
decades to come. Industrial robots trained by
machine learning become more aware of people
and surroundings. They can assist companies to
get things done with fewer errors. Some industrial
robots combines computer vision, AI, and sensors.
There are numerous advantages for combining AI,
industrial robotics, and machine learning in one
system for manufacturing production. Such
combination reduces programming time and
increases productivity and uptime as a result of
predictive maintenance. A growing number of
businesses worldwide are waking up to the
potentially transformative capabilities of machine
learning on robots. Major companies such as
Google, Tesla, Mercedes Benz, Nissan, and Ford
that have invested heavily in machine learning.
Warehouse Robots: Warehouse needs manpower
to manage the huge amount of inventory. Robots
can be trained to handle such inventories,
reducing the human workforce in performing such
repetitive tasks. A typical warehouse robot is
shown in Figure 3 [10].
Assistive Robots: The healthcare industry is
taking advantage of machine learning
methodologies and applying them to robotics.
Assistive robot is a device that can process
information and implement actions that can help
people with disabilities. Movement robots give
you a therapeutic or diagnostic benefit. A robotic
arm is a recent example of machine learning-
based robotic assistive technologies.
Surgical Robots: Potential applications of
machine learning in the surgical field are many.
Automation works well for routine processes. To
mitigate avoidable errors, automation is now
being considered as a potential tool for improved
surgical workflow. It can potentially reduce the
length of surgical procedures, minimize surgeon
fatigue, and decrease hospital costs. This is
particularly important in remote operations or
telesurgery, where any lag between human
surgical commands and robot responses can cause
complications. Through training, robots will be
capable of doing life-saving surgical procedures
and different medical approaches. When surgeries
are performed by trained robots, they would
perform the surgery with the utmost precision.
Machine learning can be used to evaluate surgeon
performance in robot-assisted surgery. Figure 4
presents a typical use of surgical robot [11].
BENEFITS AND CHALLENGES
The advantages of robotics include heavy-duty jobs
with precision and repeatability.
Through machine learning, robotics applications may
have the identical potential as people to get smarter
through experience. For example, there is one doctor
specialist for every 10,189 individuals in India. This is
where machine learning comes in. Machine Learning
systems can be developed for healthcare services,
what can change the diagnosis and treatment of
diseases and guarantee that patients get the correct
treatment at the right time.
Robotics has faced numerous challenges, notably
questions regarding efficacy, safety, and cost-
effectiveness. Traditional robots have some
disadvantages due to their stiffness and rigidity.
Robotics challenges can inspire and motivate new
machine learning research as well as being an
interesting field of application of standard ML
techniques.
CONCLUSION
Progress in AI and machine learning robotics is
taking rapidly. Machine learning is poised to disrupt
almost every sector conceivable. It will push robots to
do physical work at their best. It will accelerate the
processing power of the automation system used in
various technologies. It will occupy an increasingly
pivotal position in the creation of the next generation
of commercial and industrial systems Researchers
around the world are using AI, ML, computer vision,
and other related technologies to develop robots that
mimic the human brain.
Becoming a machine learning professional is a good
idea. Machine Learning is one of the best lucrative
career of the 21st century due to plenty of available
job opportunities with ahigh-payingincome. Pursuing
a graduate degree in machine learning provides
students with the core a state-of-the-art opportunity

knowledge needed to succeed in the field of artificial
intelligence. It has been estimated that 58 million new
jobs will be available in the AI sector in 2022,
affording you with a wide variety of opportunities
[12]. More information about the use of machine
learning in robotics can be found in the books in
[13,14] and the following journals devoted to robot-
related issues:
Robotica
Robitics and Autonomous Systems
Advanced Robotics
Journal of Robotic Systems
Journal of Robotics
Journal of Robotic Surgery
Journal of Intelligent & Robotic Systems
Journal of Mechanisms and Robotics-
Transactions of the ASME
Industrial Robots
Intelligent Service Robotics
IEEE Journal on Robotics and Automation
IEEE Robotics & Automation Magazine
IEEE Robotics and Automation Letters
IEEE Transactions on Robotics
International Journal of Medical Robotics and
Computer Assisted Surgery
International Journal of Robotics Research
International Journal of Social Robotics
International Journal of Humanoid Robotics
REFERENCES
[1] F. Iida and C. Laschi, “Soft robotics:
Challenges and perspectives,” Procedia
Computer Science, vol. 7, 2011, pp. 99-102.
[2] M. N. O. Sadiku, Y. Wang, S. Cui, and S.M.
Musa, “Healthcare robotics: A primer,”
International Journal of Advanced Research in
Computer Science and Software Engineering,
vol. 8, no. 2, Feb. 2018, pp. 26-29.
[3] K. Kandra, “Artificial intelligence and machine
learning in robotics,” July 2021,
https://medium.com/mlearning-ai/artificial-
intelligence-and-machine-learning-in-robotics-
15948dcc2f56
[4] “Human–robot interaction,” Wikipedia, the free
encyclopedia
https://en.wikipedia.org/wiki/Human–
robot_interaction
[5] R. D. Davenport, “Robotics,” in W. C. Mann
(ed.), Smart Technology for Aging, Disability,
and Independence. John Wiley & Sons, 2005,
Chapter 3, pp. 67-109.
[6] M. N. O. Sadiku, S. Alam, and S.M. Musa,
“Intelligent robotics and applications,”
International Journal of Trends in Research
and Development, vol. 5, no. 1, January-
February 2018, pp. 101-103.
[7] “Robotics,” Wikipedia, the free encyclopedia
https://en.wikipedia.org/wiki/Robotics
[8] L. Calderone, “Robots in manufacturing
applications,” August 2016,
https://www.manufacturingtomorrow.com/artic
le/2016/07/robots-in-manufacturing-
applications/8333
[9] “How AI and machine learning can improve
robotics,”
https://www.shadowrobot.com/blog/how-ai-
and-machine-learning-can-improve-robotics/
[10] “What is warehouse robotics?”
https://6river.com/what-is-warehouse-
robotics/#:~:text=Warehouse%20robotics%20i
nclude%20automated%20systems,associates%2
0to%20the%20correct%20task.
[11] “Machine learning in surgical robotics – 4
Applications that matter,” https://emerj.com/ai-
sector-overviews/machine-learning-in-surgical-
robotics-4-applications/
[12] “Future scope of machine learning,”
https://intellipaat.com/blog/future-scope-of-
machine-learning/
[13] J. R. Peters, Machine Learning of Motor Skills
for Robotics. Universityof Southern California,
2007.
[14] M. Borgatti and K. Love, Introduction to
Squishy, Stretchy, and Flexible Robots. Make
Community, 2018.

Figure 1 The relationship between AI and machine learning [3].
Figure 2 Some advantages and disadvantages of robots [8].
Figure 3 Robotics in warehouse [10].
Figure 4 A typical use of surgical robot [11]

Machine Learning in Robotics

More Related Content

What's hot (20)

Similar to Machine Learning in Robotics (20)

More from ijtsrd (20)

Recently uploaded (20)

Machine Learning in Robotics