International Journal of Trend in Scientific Research and Development (IJTSRD)

Volume 6 Issue 2, January-February 2022 Available Online: www.ijtsrd.com e-ISSN: 2456 – 6470

Soft Computing in Robotics

Matthew N. O. Sadiku , Uwakwe C. Chukwu2, Abayomi Ajayi-Majebi3, Sarhan M. Musa1
1

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 How to cite this paper: Matthew N. O.

Soft computing is considered as one of the emerging areas of Sadiku | Uwakwe C. Chukwu | Abayomi
research in all fields of engineering and sciences. Soft computing Ajayi-Majebi | Sarhan M. Musa "Soft
algorithms have gained much popularity to solve engineering Computing in Robotics" Published in
International
applications in recent years. These algorithms have gained the
Journal of Trend in
attention of researchers in solving problems in robotics, which is
Scientific Research
becoming inescapable in daily life and attracting extensive and Development
applications. The advantages of robots can only be harnessed with (ijtsrd), ISSN:
their mass introduction in the real world. In this paper, an 2456-6470,
introduction to soft computing techniques and their application in Volume-6 | Issue-2, IJTSRD49294
robotics is provided. February 2022,
pp.864-869, URL:
KEYWORDS: soft computing, hard computing, computer science, www.ijtsrd.com/papers/ijtsrd49294.pdf
robots, robotics, soft robots
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
As computing is increasingly integrated into all Soft computing approaches are more preferable over
aspects of societies and as robotic machines play an conventional, hard methods, for solving problems that
ever increasing role in carrying out many tasks that are difficult to describe by analytical or mathematical
were previously only possible by skilled human labor, models. Autonomous robotics is such a domain in
there is a growing public awareness on how these which knowledge about the environment is inherently
developments are influencing almost all areas of our imprecise, unpredictable, and incomplete. Therefore,
lives, our societies, and our economies. The human- soft computing techniques (such as fuzzy logic,
machine integration has led to the development of neural networks, and evolutionary algorithms) are
intelligent systems, such as robots [1]. After decades beneficial for robotics [3].
of research, it seems that robots will finally leave the
The term “soft computing” was coined by Lofti A.
cages of industrial environment and start working
Zadeh in 1991. Since then, the area has experienced
with humans. An example of human-machine
rapid development. Soft Computing became a
integration is shown in Figure 1 [2].
discipline within computer science in the early 1990s.
Robotics have experienced a rapid growth over the The terms “machine intelligence” and “computational
last decades. Today, several applications benefit from intelligence” have been used to have close meaning as
robots. These include manufacturing, industry, health soft computing.
monitoring, efficient logistics, disaster management,
The principal premise of soft computing (SC) is that
rehabilitation, agriculture, aerospace,
we live in a world that is imprecise and uncertain.
communications systems, automation, transport
Soft computing refers to the use of “inexact”
systems, and military.
solutions to computationally hard tasks [4].

@ IJTSRD | Unique Paper ID – IJTSRD49294 | Volume – 6 | Issue – 2 | Jan-Feb 2022 Page 864
 International Journal of Trend in Scientific Research and Development @ www.ijtsrd.com eISSN: 2456-6470
OVERVIEW OF SOFT COMPUTING topics in robotics. There are some tasks that
Soft computing (SC) is a branch of computer science traditional, rigid robots cannot do. Soft-bodied
that resembles the processes of the human brain. It robots, on the other hand, may be able to interact
may also be regarded as a newly emerging with people more safely. Robots inspired by
multidisciplinary field. Its main objective is to biology are becoming common. Biological
develop intelligent machines in order to solve real- inspiration does not suggest that we attempt to
world problems. It differs from the conventional hard copy nature. Rather, the objective is to understand
computing as it can handle uncertainty, imprecision the principles underlying the behavior of animals
easily. While conventional hard computing is based and humans and transfer the behavior int the
on crisp values and binary numbers, SC uses soft development of robots. As shown in Figure 3, a
values and fuzzy sets. bio-inspired “soft” robot is capable of exhibiting
substantial behavioral diversity [7]. Soft material
Soft computing refers to a collection of computational
robots are elastically soft, versatile, and
techniques in computer science, artificial intelligence,
biologically inspired machines. They are
and machine learning. The techniques deal with
primarily composed of easily deformable matter
development of approximate models in finding
such as fluids, gels, and elastomers. Abilities such
solutions to real world problems. They aim to exploit
as squeezing, stretching, climbing, growing, and
the tolerance of imprecision and uncertainty to
morphing would not be possible with an approach
achieve tractability, robustness, and low solution cost.
based only on rigid links. A soft robot must adapt
Its principle components include:
its shape and locomotion strategy for a broad
Expert systems
range of tasks, obstacles, and environmental
Neural networks,
conditions. This emerging class of robots could
Machine learning
revolutionize the role of robotics in healthcare,
Probabilistic reasoning
field exploration, and cooperative human
Evolutionary algorithms
assistance. For example, some attributes of soft
Artificial neural networks
robotic systems make them well-suited for use in
Fuzzy logic
human wearables. The proliferation of soft
Swarm intelligence
robotics research worldwide has brought
Interactive computational models
substantial achievements in terms of principles,
These computation methods or technologies provide models, technologies, techniques, and prototypes
information processing capabilities to solve complex of soft robots. Most of the interest in soft robotics
practical problems. Some of these techniques are has been limited to robotic grippers, actuators,
illustrated in Figure 2 [5]. Integrating these and flexible computational mechanisms. Soft
complementary methodologies results in more computing techniques have been applied to
powerful hybrid methods than using a single method robotic grasping [8].
exclusively.
Swarm Robots: Swarm robotics refers to the
APPLICATIONS OF SC IN ROBOTICS development of tools and techniques to ease the
Soft computing is used for solving real-life problems coordination of multiple small-sized robots
and can be applied in different fields such as towards the accomplishment of difficult tasks.
education, healthcare, business, industry, engineering, The desired collective behavior of the robots is
power systems, transportation, communication usually based on swarm intelligence, with the
systems, wireless communications, data mining, underlying strategy inspired from biological
home appliances, robotics, etc. [6]. Common systems such as ants colonies, bird flocks, and
applications of soft computing in robotics include the fish schools. Swarm robotics is a relatively new
following: paradigm used to control the operation of a large
Soft Robots: In the past, most robots have rigid group of physically limited robots. The success of
bodies for different reasons, such as mobile robot, swarm robotics applications may be attributed to
humanoid and manipulator. A rigid robot has efficient use of smart sensing, communication,
some inherent drawbacks, such as noncompliance and organization functionalities, which allow for
causing low adaptability and unsafety and collaborative information sensing, operation, and
unfriendliness to users. Soft robots provide safe knowledge inference from the environment.
interactions and compliant behaviors. They also Swarm robotics have become a major research
are energy-saving as it is normally lighter than catalyst of the computational intelligence
rigid robot. Soft robotics has exploded in community. The main advantages of using swarm
popularity in recent years. It is one of the hottest robotic systems are: (1) Robustness: due to the

@ IJTSRD | Unique Paper ID – IJTSRD49294 | Volume – 6 | Issue – 2 | Jan-Feb 2022 Page 865
 International Journal of Trend in Scientific Research and Development @ www.ijtsrd.com eISSN: 2456-6470
distributed nature of the swarm, the failure of a Cooperative Robots: Cooperation is just
single robot does not compromise the integrity working together. This can occur even with a
and operation of the remaining robots, (2) group of robots collaborate or communicate.
Scalability: the addition of new robots to the Cooperative robots can perform tasks that are
swarm can be made incrementally (without either difficult or impossible to be accomplished
reconfiguring the entire population), (3) by a single-arm robot. Such tasks can be
Parallelization: complex control is achieved manipulating flexible objects and grasping the
through simple yet concurrently held interactions objects. Cooperating robots may also increase the
between the members of the swarm [9,10]. productivity in a manufacturing environment.
Cooperative robotic systems have also been used
Robot Vision: Although there are many sensors
in aerospace applications [14].
used by robots, one of the most used types is
vision sensors. Vision enables autonomous Other applications of soft computing in robotics
systems to complete complex tasks where include social robots, robot arm manipulators, web
environment information is needed. Robot vision robot detection, elastic robotic arm, demining robots,
is important because it is needed in performing and flying robots.
more complex tasks such as grasping, mapping,
BENEFITS
navigation, tracking, active sensing, and pattern It is difficult or if not impossible to design robots for
recognition. In each of these tasks, the robot
complex systems such as nonlinear, time-varying
needs to process large amounts of data at a fast
systems using conventional approaches. Soft
rate in order to satisfy real time operation
computing is a collection of intelligent techniques
constraints. Soft computing has been widely used working in a complementary way to build robust
in robotics and vision applications. Soft systems at affordable cost. The techniques contribute
computing algorithms, such as PSO and fuzzy to one of the long term goal in robotics, to solve the
logic, can be used to solve important computer problems that are unpredictable and imprecise namely
vision problems, like image tracking, plane in unstructured real-world environments. The major
detection, and pattern recognition [11]. benefits of soft computing are: (i) its rich knowledge
Mobile Robot: Mobile robots are intelligent representation (both at signal and pattern level), (ii)
robots that are capable of navigating intelligently its flexible knowledge acquisition process (including
anywhere using sensor-actuator control machine learning and learning from human experts)
techniques. They can move, sense, and react in a and (iii) its flexible knowledge processing [11].
given environment and are able to perform tasks
Robots have the potential for replacing humans in
and navigate without human intervention. Mobile
hazardous working environments, such as the cleanup
robots perform many tasks such as rescue
of toxic waste, nuclear power plant decommissioning,
operation, patrolling, disaster relief, planetary
mining, space exploration, search and rescue
exploration, and material handling. They are
missions, security, and surveillance. They are also
increasingly used in many fields such as industry,
used to replace humans in performing repetitive type
space, defense, and transportation. A mobile robot
of tasks, such as automated manufacturing, industrial
operating in an unstructured environment must be
maintenance, spray painting, and arc welding. These
able to deal with dynamic changes of the
advantages let us to build intelligent systems with a
environment. Thus, navigation is important in
high machine intelligence quotient at low cost. Due to
mobile robotics. There are two types of
these advantages, robots inspired by nature are
navigation: global navigation and local
becoming common in robotics labs. The increase in
navigation. In the global navigation, the prior
adoption of robots is causing the development of new
knowledge of the environment should be
battery technologies that are affordable, safe and
available. In the local navigation, the robot can
longer-lasting.
decide or control its motion and orientation
autonomously using equipped sensors. Several CHALLENGES
soft computing techniques (such as fuzzy logic, Despite decades after first robot was created, the
genetic algorithm, neural networks, and particle robotics industry faces many challenges. Several
swarm optimization (PSO)) have been used to challenges still need to be addressed in spite of the
solve the robot navigation and obstacle avoidance progress made during the recent past. In many robotic
problem in the various environments. For applications, it is difficult if not impossible to obtain a
example, neural network is one of the important precise mathematical model of the robot’s interaction
technique for the mobile robot navigation [12,13]. with its environment. The lack of precise and

@ IJTSRD | Unique Paper ID – IJTSRD49294 | Volume – 6 | Issue – 2 | Jan-Feb 2022 Page 866
 International Journal of Trend in Scientific Research and Development @ www.ijtsrd.com eISSN: 2456-6470
complete knowledge about the environment limits the [3] A. Saxena and A. Saxena, “Review of soft
applicability of conventional control system design to computing techniques used in robotics
the domain of autonomous robotics. It is unrealistic to application,” International Journal of
assume that any learning algorithm is able to learn a Information and Computation Technology, vol.
complex robotic task, in reasonable learning time 3, no. 3, 2013, pp. 101-106.
starting from scratch [15]. [4] M. N. O. Sadiku, Y. Wang, S. Cui, S. M. Musa,
There is still room for further research for the “Soft computing: An introduction,”
advancement of soft robotics Many soft robotic International Journal of Advanced Research in
systems still use rigid components that limit their Computer Science and Software Engineering,
range of capabilities. A grand challenge for soft vol. 8, no. 6, June 2018, pp. 63-65.
robotics is the development of an autonomous and [5] “Difference between AI and soft computing,”
intelligent robotic system fabricated entirely out of March 2020,
soft materials. The development of “smart” soft https://www.geeksforgeeks.org/difference-
materials is critical for engineering the next between-ai-and-soft-computing/
generation of fully soft robots that can easily adapt
and conform to their environment. Soft robots have [6] M. Dahiya, “Applications of soft computing in
been limited due to their lack of good sensing. Robot various areas,” International Journal of
ethics is a major challenge [16]. Engineering Sciences & Research Technology,
vol. 6, no. 5, May 2017, pp. 712-716.
CONCLUSION
Soft computing deals with the design of intelligent [7] R. Pfeifer, M. Lungarella, and F. Iida, “The
and robust systems, which exploit the tolerance for challenges ahead for bio-inspired ‘soft’
imprecision inherent in many real world problems. robotics,” Communications of the ACM, vol.55,
Robotics is such a domain in which soft computing no. 11, November 2012, pp. 76- 87.
techniques are greatly beneficial for robotics. Robots [8] K. Wu and G. Zheng, “Simulation and control
are generally built by associating a mechanical co-design methodology for soft robotics,”
structure, actuators, and their control system. As Proceedings of the 39th Chinese Control
engineers and scientists look to the future of robotics, Conference, Shenyang, China, July 2020, pp.
they are looking for the best ways to create a robot 3910-3914.
that is entirely soft, has no hard components, such as
batteries and circuit boards, and is not tethered to a [9] E. Osaba et al., “Soft computing for swarm
nearby circuit system. More information about soft robotics: New trends and applications.,”
computing in robotics can be found in the books in Journal of Computational Science, vol 39,
[17-28] and the following related journals: January 2020.
Soft Computing [10] “Special issue on soft computing applied to
Soft Robotics (SoRo) swarm robotics,” Applied Soft Computing,
August 2015.
Journal of Soft Computing Paradigm
Applied Soft Computing Journal [11] V. Ayala-Ramirez et al., “Soft computing
applications in robotic vision systems,” June
Artificial Intelligence & Robotics Development
2007,
Journal
https://www.intechopen.com/chapters/292
International Journal on Soft Computing
[12] A. Pandey, S. Pandey, and D. R. Parhi, “Mobile
International Journal of Artificial Intelligence and
robot navigation and obstacle avoidance
Soft Computing
techniques: A review,” International Robotics
REFERENCES & Automation Journal, vol. 2, no. 3, 2017, pp.
[1] “Future computing and robotics: A report from 96-105.
the HBP foresight bab,”
[13] M. Algabri et al., “Comparative study of soft
https://www.kcl.ac.uk/ghsm/assets/Foresight%2
computing techniques for mobile robot
0LabFuture%20of%20Computing%20and%20
navigation in an unknown environment,”
Robotics.pdf
Computers in Human Behavior, vol. 50,
[2] B. Williams, “An introduction to robotics,” September 2015, pp. 42-56.
https://www.ohio.edu/mechanical-
[14] W. Gueaieb, “Soft computing based approaches
faculty/williams/html/PDF/IntroRob.pdf
for the robust control of cooperative

@ IJTSRD | Unique Paper ID – IJTSRD49294 | Volume – 6 | Issue – 2 | Jan-Feb 2022 Page 867
 International Journal of Trend in Scientific Research and Development @ www.ijtsrd.com eISSN: 2456-6470
manipulator systems,” Doctoral Dissertation, [22] O. Castillo and W. Pedrycz (eds.), Soft
University of Waterloo, 2001. Computing for Intelligent Control and Mobile
[15] F. Hoffmann, “An overview on soft computing Robotics. Springer Science & Business Media,
2010.
in behavior based robotics,”
https://citeseerx.ist.psu.edu/viewdoc/download? [23] T. Kitamura, What Should be Computed to
doi=10.1.1.59.6163&rep=rep1&type=pdf Understand and Model Brain Function?: From
[16] S. Crowe, “10 Biggest challenges in robotics,” Robotics, Soft Computing, Biology And
Neuroscience to Cognitive Philosophy. World
February, 2018,
Scientific, 2001.
https://www.therobotreport.com/10-biggest-
challenges-in-robotics/ [24] A. Zilouchian and M. Jamshidi (eds.),
Intelligent Control Systems Using Soft
[17] S. Tzafestas (ed.), Soft Computing in Systems
Computing Methodologies. Boca Raton, FL:
and Control Technology. World Scientific
CRC Press, 2001.
Publishing Company, 1999.
[18] J. Fodor and J. Kacprzyk (eds.), Aspects of Soft [25] O. Castillo, J. Kacprzyk, and W. Pedrycz, Soft
Computing for Intelligent Control and Mobile
Computing, Intelligent Robotics and Control.
Robotics. Springer 2011.
Springer, 2009.
[26] Y. T. Kim, I. Kobayashi, E. Kim (eds.), Soft
[19] R. Rajendra and D. K. Pratihar, Modeling and
Computing in Advanced Robotics. Springer
Simulations of Robotic Systems using Soft
International Publishing, 2014.
Computing: Robotics and Intelligent Systems.
LAP LAMBERT Academic Publishing, 2012. [27] T. Kitamura, Fuzzy Logic Systems Institute Soft
Computing Ser.: What Should Be Computed to
[20] A. Shukla, R. Tiwari, and R. Kala, Real Life
Understand and Model Brain Function? : From
Applications of Soft Computing. Boca Raton,
Robotics, Soft Computing, Biology and
FL: CRC Press, 2019.
Neuroscience to Cognitive. World Scientific,
[21] J. Fodor and R. Fullér (eds.), Advances in Soft 2001.
Computing, Intelligent Robotics and Control.
Springer Science & Business Media, 2014. [28] L. C. Jain and T. Fukuda (eds.), Soft Computing
for Intelligent Robotic Systems. Springer, 1998.

Figure 1 A typical example of human-machine integration [2].

@ IJTSRD | Unique Paper ID – IJTSRD49294 | Volume – 6 | Issue – 2 | Jan-Feb 2022 Page 868
 International Journal of Trend in Scientific Research and Development @ www.ijtsrd.com eISSN: 2456-6470

Figure 2 Soft computing approaches [5].

Figure 3 A bio-inspired “soft” robot can exhibit substantial behavioral diversity [7]

@ IJTSRD | Unique Paper ID – IJTSRD49294 | Volume – 6 | Issue – 2 | Jan-Feb 2022 Page 869