Volume-04 ISSN: 2455-3085 (Online)

Issue-06 RESEARCH REVIEW International Journal of Multidisciplinary

June-2019 www.rrjournals.com[UGC Listed Journal]

Artificial Intelligence in Automation

Sarmah Simanta Shekhar

Business Intelligence Architect, Alpha Clinical Systems Inc, NJ, USA

ARTICLE DETAILS ABSTRACT

Article History The development of Artificial Intelligence is speeding up rapidly and combination of
Published Online: 15 June 2019 Artificial Intelligence with automation has started to change the business landscape.
Companies and business are focusing on applying existing Artificial Intelligence with
Keywords automation processes to gain the new heights of efficiency and quality. The paper
Artificial Intelligence, machine learning, depicts about artificial intelligence and automation, and it tries to demonstrate the
automation, automation industry, AI audience how both Artificial intelligence and automation are related and how they can
companies. be more effective when they work together and can give competitive advantage.
*
Corresponding Author
Email: sarmah.simanta[at]gmail.com

1. Introduction automation in various industries. There are various types of

Artificial Intelligence (AI) is the science [1] that enables the automation, some of the popular ones are as follows-
computers and the machines to learn, judge and use own
reasons. As the technologies are becoming more complex, the 2.1 Numerical Control
Drills, 3D printing, glass cutting, etc. fall in this category
demand of Artificial Intelligence is growing because of its ability
where machines are programmed to execute repetitive tasks.
to solve complex problems with limited human resources and
expertise and within a limited time. AI adopts the abilities to
2.2 Computer-aided manufacturing (CAM)
equip the technical expertise and can amplify expertise to learn Computer software are used for this automation example of
and deploy new methods and applications. There is a big which are like Computer-aided design (CAD), Computer -aided
breakthrough in the field of image recognition using machine design and drafting etc.
learning along with the [2] advances in big data and GPU
(Graphic processing units) which evidently helped Artificial 2.3 Flexible manufacturing systems (FMS)
Intelligence grow faster. Artificial Intelligence (AI) system It is a sophisticated automation system where robots and
consists of an agent and its environment. An agent such as other advanced automation tools are used to provide flexibility
human or a robot identifies the environment through sensors and customization to the users.
and effectors. It uses a method called search and pattern
matching where the computer is instructed to search its 2.4 Industrial robot:
knowledgebase based on the match found and if specific Robots are being used for welding, assembly and handling
conditions are met to solve a problem. materials etc. where robots can be programmed and
manipulated in three or more axes.
Now a days, Artificial Intelligence is widely used in various
industries and business domain, starting from healthcare, 3. Differences between Artificial Intelligence (AI) and
finance, manufacturing to law, education, etc. With the help of Automation
machine learning, doctors can diagnose diseases faster than
Before diving into artificial intelligence in automation, it's
before. Applications such as chatbots is helping customers to
important to acknowledge that both these terms are used
schedule appointments and support patients through the billing
interchangeably in daily life. They are associated with physical
process. In education, AI can provide auto grading, help
or software robots along with other machines which allow us to
students with learning by providing support to their need and
work more effectively and efficiently. This can be either
help them to stay on track. In the law industry, AI has made it
mechanical tasks like piecing together something like a car of
easier for the lawyers to go through thousands of large legal
sending off a follow up mail just the day after you find that your
documents effectively and accurately which is a real tedious
customer hasn't yet completed her/his order. But what people
work in general. Industrial robots have made the manufacturing
fail to realize is that there are also big differences between
way easy and efficient than it used to be a few years back.
these two. These differences correspond to the complexity level
if both systems. Here's what those differences are:
2. Concept of Automation
 Terms of Difference: Automation,[9] basically means
The use of automation began [3] to increase in the last making a software or hardware which is capable of
decade with an aim to reduce manpower and time. Automation automatically doing things and that too without any
has introduced a system of computer and machines and form of human intervention. Artificial intelligence on the
replaced a system that was built by combining man and other hand is a science as well as engineering which is
machine. Highly intense and repetitive tasks have become involved in making machines which are intelligent. AI is
efficient and the product quality has also increased with the use about attempting to make machines mimic or even try
to supersede human intelligence and behavior.

RRIJM 2015, All Rights Reserved 14 | P a g e

 Volume-04, Issue-06,June-2019 RESEARCH REVIEW International Journal of Multidisciplinary

 Data: Automation may or may not be based on artificial  Machine Learning: It refers to the ability of a machine
intelligence. The whole practice of [4] automation has to learn using the data fed to it. This involves the
evolved into its current form between the first and third outcomes of environment variables and decisions to
industrial revolution. It involves production using improve itself. Using machine learning we will be able
automatic testing, mechanical labor, control systems, to improve the total efficiency of current solutions. Let's
computers and operating equipment’s. All the types of take an example to understand it better: if an issue
automation which has manifested all around us are comes under an intelligent automation system and to
bound using explicit programming and rules. To resolve it human intervention has needed them the
ensure that the same thing becomes an AI, [10] all that next time, the system will very automatically follow the
is needed to be done is to power it up using data. set of procedures which was used by a human. So
Huge quantities of data, like using neural networks, over time, the human effort will be reduced ad the
graphs and deep machine learning must be put in the system efficiency will improve.
software. Your coding levels will certainly decide just
how much you will be able to make your system 6. What Are the Important Applications of Artificial
stimulate like a human. But most likely, you will be Intelligence in Automation?
teaching the system all that you already know. In case
Artificial Intelligence can be used [8] in various ways in
of automatic, you will be able to easily know the output
Automation. From drones to self-driving cars, all are making use
using sensor readings. But in case of AI there is
of intelligent automation. Here are some of the major ways in
always a little bit of uncertainty, just like it's there with
which a business will be able to benefit from a combo of AI and
the human brain.
automation:
 Purpose: Automation can execute repetitive tasks.
 Preventing Fraud: It will be able to link the theft directly
This frees up valuable time [5] for people to take up
to the face of the perpetrator. There will be a camera
more important task which require rational judgment
that will be attached to the POS system and it will
and thought. This makes the whole thing more efficient
record all the types of transactions and then link them
and cost effective. Artificial Intelligence is designed to
directly to the face with the details that are already
not just seek patterns but also to learn from experience
present in the system. Say, for example, someone
so that they can self-select the appropriate responses
commits credit card fraud them it will become much
according to situations.
easier to put them behind bars. Moreover, an
4. Can the Two Work Together? intelligent system will also be able to prevent cyber-
attacks by swiftly identifying abnormal behavior from
The use of software to reduce human effort isn't new news the user. In such situations, the system automatically
for the business community. Artificial Intelligence [6] on top of it stops taking any requests and fires off an alert right to
has opened a whole new possibility. Automation provided for a the administrator.
very limited range of reducing human work. But by combining  Brand Management: The work of brand management
artificial intelligence with automation, one will be able to reduce can be made a lot easier with the help of automation.
not just human effort but also totally remove the need for such Brand Marketers struggle to fully understand the
intervention altogether. This kind of combination in artificial consumer's opinion about their own brand. So, with the
intelligence in automation is known as automation continuum (or help of automation, they will be able to automate the
intelligence Robotic Process Automation) analysis of all the present content across the internet.
This can be carried out daily and will be helpful in
5. What Are the Major Components of AI in Automation?
identifying critical issues. A good example of such an
An automation system functions using the three automation is the Watson Analytics for Social Media.
components of artificial intelligence. So, depending on the need, Here, the user defines a set of focus words and
they can be either combined or even used separately to allow context and understand what online users are
for a fully automated response: describing in a moment.
 Machine Vision: This refers to the potential [7] of any • Customer Service: A great example of this is chat bots.
program to understand what the visual input is. The They have become a lot popular and that too in a very
machine makes use of the training data (images) as a short period. It began with Apple's Siri but have been
type of foundation for the identification or classification adopted by most brands now. Chat bots are programs
mechanism. For example, face recognition system of which can understand the user's input on a contextual
iPhone X uses machine vison technology. basis and then respond to the queries accordingly.
 Natural Language Processing: Machine language These are used for automating customer service, sales
world on the visuals, Natural Language Processing and marketing messages. These bots create friction
(NLP) does the same to understand human voice and during app download process in the case of popular
text inputs. It's now possible for machines to platforms like Facebook, messenger and such. They
understand what the context behind the feel human too and can reduce the burden that falls on
communication is being carried out and then take customer help desks.
actions based on the kind of prebuilt data and  Software Testing and Development: Automation in
contextual variables which are at play. Some examples software testing is a very rapidly evolving field. So, with
of this are Apple's Siri, Amazon Alexa and such. a whole range of tools which are made available, it is

RRIJM 2015, All Rights Reserved 15 | Page

 Volume-04, Issue-06,June-2019 RESEARCH REVIEW International Journal of Multidisciplinary

very possible that the work involved in is fully in recidivism. Such steps will help put the power back
automated in the future. Some examples of popular in the hands of the consumers.
testing tools are ReTest, Applitools, SauceLabs and  Convergence with Other Technologies: 2019 is clearly
more. With the help of these tools, developers can the year when AI combined up with other technologies
focus on doing core testing and leave out the worries like that of ERP solution, analytics and such will be the
of bug fixing in the intelligent systems. Development secret for the success of enterprises. Convergence, by
automation is a long way to go for sure but there are making use of data from various sensors will be able to
tools which can relieve developers from being give better data driven decisions which could improve
engaged in menial tasks. the quality and performance.
 Human Resource Management: Recruiters struggle a  Leadership Takes Ownership: AI rollout to just one
lot in sorting through the CVs which get submitted with specific team won't really gain traction as you will have
them. With the help of automation, they will be able to to involve all the segments of an organization. But
identify the potential candidates and manage old data. more than that, tech rollout, especially ones which can
Most of the CVs are already received using some change the entire dynamics of an enterprise needs to
automated applicant tracking system. These solutions be properly led from the top. Along with this,
upload the job application materials straight to their leadership should make sure that making AI is a
database when the users apply for a role. priority in the enterprise and offer retraining programs.
 Reduces Cost: As one can [11] imagine, the cost of Employees should learn the skills so that they can run
training a human in any routine task will recurring. You the AI programs efficiently in the organization. So, an
will have to deal with employee turnover, give time for investment will have to be made in both the software
gradual experience and skill development as well as and employees if the real reward is to be seen.
incur vocational costs. On the other hand, a machine  AI for Decision-Making: As AI begins to make its way
must be trained only once and then it will improve over through the various business and organizations and
time and that too with absolutely no cost involved in we can see significant openness [12] to AI regarding
repeat training. automated decision making. For example- AI can
 Improved Efficiency: People can be efficient at a job or make decisions about making a refund to a customer
they can surely grow to be so after some time but no or accepting an applicant’s mortgage application by
matter how efficient people are, they will still make evaluating all the conditions. So, the load of off
mistakes. But an Automation solution is way more employees and the robot instead will take up the task.
foolproof and it will not indulge in many errors. With But of course, it must be made clear to consumers
time it will also be able to learn from the outputs and what is the algorithm which the AI is making use of.
so its efficiency will also improve.
AI and automation are clearly the future and failing to
7. How can AI Help in Improving the Human Condition? incorporate that in your enterprise will simply mean sleeping
 Transparent AI leads to improve human trust: Most your way through a miraculous digital transformation. Because
companies do make use of AI whether for marketing, of that happens then your company will lose the race and will
hiring or such. The issue which we are facing now is surely go extinct.
the absolute lack of consumer trust. If you are not
already using AI within the company, then this might 8. Conclusion
not be something which has been thought about. But Artificial intelligence in automation has contributed to the
this needs to be focused on as AI continues to grow businesses by reducing operational cost and vocational costs. It
rapidly. There are many companies which make use of has introduced a new level of accuracy and due to Artificial
AI and algorithms to make decisions for the customers. Intelligence’s learning ability, efficiency increases over time.
But they lead the customers believe that it's the human Even though, there is a good advancement in the field of
making the decisions. This isn't the best policy Automation and Artificial Intelligence, both Artificial Intelligence
obviously because customers will want to know how and machine learning are yet to be optimized. Companies have
their information is being used, the decisions are being realized that the key to the business success is subjected to
made and if there is any bias in making those machine learning, artificial intelligence and automation. Soon,
decisions. Along with this, in legal fields you could be the companies will be fully equipped with these start systems
asked to share the algorithms which helped determine and would completely change the traditional systems with by
yielding significant benefits.

References

1. Tzafestas, Spyros, and Henk Verbruggen. "Artificial Impact on Automation and Autonomy," in IEEE Transactions
intelligence in industrial decision making, control, and on Automation Science and Engineering, vol. 14, no. 2, pp.
automation: an introduction." Artificial Intelligence in Industrial 407-409, April 2017.doi: 10.1109/TASE.2016.2640778
Decision Making, Control and Automation. Springer, 3. Nau D.S. (2009) Artificial Intelligence and Automation. In: Nof
Dordrecht, 1995. 1-39. S. (eds) Springer Handbook of Automation. Springer
2. K. C. Morris, C. Schlenoff and V. Srinivasan, "Guest Editorial Handbooks. Springer, Berlin, Heidelberg
A Remarkable Resurgence of Artificial Intelligence and Its

RRIJM 2015, All Rights Reserved 16 | Page

Volume-04, Issue-06,June-2019 RESEARCH REVIEW International Journal of Multidisciplinary

4. R. Boire, "Artificial intelligence (AI), automation, and its impact www.processexcellencenetwork.com/rpa-artificial-

on data science," 2017 IEEE International Conference on Big intelligence/articles/a-guide-to-robotic-process-automation-
Data (Big Data), Boston, MA, 2017, pp. 3571-3574.doi: rpa?utm_source=google&utm_medium=cpc&utm_content=ag
10.1109/BigData.2017.8258349 uidetoRPA&gclid=EAIaIQobChMIuNn0lM-
5. Autor, David H. (2015) “Why Are There Still So Many Jobs? Y4gIVA5CPCh0MaA6zEAAYAiAAEgIuRPD_BwE.
The History and Future of Workplace Automation,” Journal of 9. Evans, Grayson. "Solving home automation problems using
Economic Perspectives, 29(3):3–30. artificial intelligence techniques." IEEE transactions on
6. 1054082769. “What Is The Real Difference Between consumer electronics 37.3 (1991): 395-400.
Automation And AI?” Becoming Human: Artificial Intelligence 10. Samuel, A. L. (1962). Artificial Intelligence: A Frontier of
Magazine, Becoming Human: Artificial Intelligence Magazine, Automation. The ANNALS of the American Academy of
9 Aug. 2018, becominghuman.ai/what-is-the-real-difference- Political and Social Science, 340(1), 10–20.
between-automation-and-ai-366513e0c910. 11. Arcuri, Andrea. "On the automation of fixing software bugs."
7. “A Guide to Robotic Process Automation (RPA).” Google, Companion of the 30th international conference on Software
Google,https://www.processexcellencenetwork.com/rpa- engineering. ACM, 2008.
artificial-intelligence/articles/a-guide-to-robotic-process- 12. King, Ross D., et al. "The automation of science." Science
automation-rpa? 324.5923 (2009): 85-89.
8. Hawkins, Ian. “A Guide to Robotic Process Automation
(RPA).” Process Excellence Network, 6 Mar. 2019,

RRIJM 2015, All Rights Reserved 17 | Page