New Paradigm Of Industry 4.0 Internet Of Things, Big Data Cyber Physical Systems (Srikanta Patnaik)（Z-Library）

Page 1

Studies in Big Data 64 Srikanta Patnaik Editor New Paradigm of Industry 4.0 Internet of Things, Big Data & Cyber Physical Systems

Page 2

Studies in Big Data Volume 64 Series Editor Janusz Kacprzyk, Polish Academy of Sciences, Warsaw, Poland

Page 3

The series “Studies in Big Data” (SBD) publishes new developments and advances in the various areas of Big Data- quickly and with a high quality. The intent is to cover the theory, research, development, and applications of Big Data, as embedded in the fields of engineering, computer science, physics, economics and life sciences. The books of the series refer to the analysis and understanding of large, complex, and/or distributed data sets generated from recent digital sources coming from sensors or other physical instruments as well as simulations, crowd sourcing, social networks or other internet transactions, such as emails or video click streams and other. The series contains monographs, lecture notes and edited volumes in Big Data spanning the areas of computational intelligence including neural networks, evolutionary computation, soft computing, fuzzy systems, as well as artificial intelligence, data mining, modern statistics and Operations research, as well as self-organizing systems. Of particular value to both the contributors and the readership are the short publication timeframe and the world-wide distribution, which enable both wide and rapid dissemination of research output. ** Indexing: The books of this series are submitted to ISI Web of Science, DBLP, Ulrichs, MathSciNet, Current Mathematical Publications, Mathematical Reviews, Zentralblatt Math: MetaPress and Springerlink. More information about this series at http://www.springer.com/series/11970

Page 4

Srikanta Patnaik Editor New Paradigm of Industry 4.0 Internet of Things, Big Data & Cyber Physical Systems 123

Page 5

Editor Srikanta Patnaik Department of Computer Science and Engineering SOA University Bhubaneswar, Odisha, India ISSN 2197-6503 ISSN 2197-6511 (electronic) Studies in Big Data ISBN 978-3-030-25777-4 ISBN 978-3-030-25778-1 (eBook) https://doi.org/10.1007/978-3-030-25778-1 © Springer Nature Switzerland AG 2020 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Page 6

Editorial The Fourth Industrial Revolution, also known as Industry 4.0, has gained attention from researchers, academicians, and industry personnel across the world over the past few years. The first three revolutions being (i) the invention of steam engine; (ii) the invention of electricity and adoption of mass production by division of labor; and (iii) the combining development of computers with information tech- nology (IT) to automate things have brought a lot of changes to industrial devel- opments. Now, with the advent of Internet of things (IoT) and cyber-physical systems (CPS), both industries and governments have started investing huge amount in related projects to develop Internet of everything in all sectors. During this digital transformation phase, novel industrial technologies have made it easier to gather data from machines and analyze them for enhancing efficiency, flexibility, and speed for producing high-quality goods at optimal cost, thus increasing pro- ductivity and growth of companies. Some of the key technologies forming the building blocks and driving the research of Industry 4.0 include: big data and analytics, cloud computing, vertical and horizontal integration, cyber-physical systems, autonomous robots, Internet of things, cyber-security, additive manufacturing, augmented reality, and simulation of concepts and models. While adoption of big data analytics enables collection and evaluation of data from heterogeneous sources such as various industrial equip- ments, processes, and systems, these outcomes can be utilized to generate insights for making several real-time decisions. Cloud computing enables data sharing across the globe among different production sites beyond the boundaries, thus opening more doors for data-driven services for various connected systems. Vertical and horizontal integration of data-driven networks and services in Industry 4.0 will lead to evolution of fully automated value chains. Deploying autonomous robots will enhance production capability and efficiency to a great extent at reduced costs. Moreover, these robots are capable of interacting with each other as well as human beings and can learn things eventually. Again industrial IoT enriches field devices with embedded computing that allows devices to interact and communicate with each other as well as centralized and decentralized controllers for real-time response and decision making. Also, the intensified network connectivity and v

Page 7

increased use of communication protocols to enhance accessibility in Industry 4.0 make the system quite vulnerable and exposed to several types of threats. This induces adoption of cyber-security processes to provide reliable and secured communication channels and essential authenticity management systems. Further, additive manufacturing methods help in producing customized products according to requirements. And augmented reality supports providing real-time information through smart phones or devices to make decisions regarding selection of objects in warehouses or other instructions regarding repairing of machines, etc. Finally, simulation of concepts and models or prototypes of machines, products, and processes with different parameter settings will allow fine-tune and optimize the prototypes in the virtual world before implementing them physically and will also improve quality. Although there are so many positive aspects to adopt Industry 4.0, still there are many challenges that companies and firms face at a stretch. However, being a developing research area, the current literature about Industry 4.0 still lacks systematic coverage of different aspects of this new wave of revo- lution. Although there are lots of articles and research papers available all over, still the unavailability of the state of the art of the current progresses in the field lacks to provide a complete picture of the current developments as well as to what extent researchers have been progressed. This book attempts to focus on current and future research directions of Industry 4.0. It addresses a wide range of application fields of Industry 4.0 and various issues being faced by the companies while adopting Industry 4.0. It also tries to establish some standard concepts which can further form the base for future models and applications. This book again provides a platform to the authors to study underlying processes and develop tools and solutions for effective management of these processes. Finally with selective chapters, this book aims to systematically address the gap between the academic investigation of different aspects of Industry 4.0 and actual feasibility and chal- lenges being faced by companies while adopting Industry 4.0. The volume consists of eight selective chapters to fulfill the purpose of the book and provide a broad coverage of the topic. Chapter “Management of V.U.C.A. (Volatility, Uncertainty, Complexity and Ambiguity) Using Machine Learning Techniques in Industry 4.0 Paradigm” provides an introduction to Industry 4.0 and identifies various risk factors involved in the digitization process of machineries as volatility, uncertainty, complexity, and ambiguity (VUCA). The authors define the terms and showcase their systematic relationships associated with Industry 4.0 practices. They also provide some future research directions in the same context. Chapter “Role of Industry 4.0 in Performance Improvement of Furniture Cluster” investigates critical success factors (CSFs) and studies their impact on channelizing resources and improving performance of furniture clusters by utilizing data effectively for decision making. Chapter “Imparting Hands-on Industry 4.0 Education at Low Cost Using Open Source Tools and Python Eco-system” examines various areas of the Industry 4.0 paradigm and discusses how open-source tools like Python and their supporting libraries can be utilized to provide education through hands-on sessions. Further, chapter “Decision Support Framework for Smart Implementation of Green Supply Chain Management vi Editorial

Page 8

Practices” focuses on the development of a decision support framework for adopting green supply chain management practices in an attempt to reduce carbon footprints. The authors present a hybrid decision-making approach that adopts multicriteria evaluation approach to process orders dynamically and cope with market variations. Also, chapter “Decision Support System for Supply Chain Performance Measurement: Case of Textile Industry” proposes a decision support framework that identifies the key performance indicators that contribute toward performance measurement of supply chain management systems in textile industry. Chapter “A Review Study of Condition Monitoring and Maintenance Approaches for Diagnosis Corrosive Sulphur Deposition in Oil-Filled Electrical Transformers” studies corrosive sulfur depositions in oil-filled electrical trans- formers and establishes an effective plan for maintenance. The authors monitor relevant conditions to reduce transformer failures by diagnosing the corrosion caused inside the transformer. The authors further suggest a dynamic and cost-effective condition-based maintenance plan for detecting failures at early stages, so as to avoid any hazardous effect. Chapter “Principal Components Based Multivariate Statistical Process Monitoring of Machining Process Using Machine Vision Approach” attempts to integrate a machine-vision-based multivariate sta- tistical process monitoring technique with principal component analysis to monitor machining process and provide automated industry-ready solution. Lastly, chapter “Green IS—Exploring Environmental Sensitive IS Through the Lens of Enterprise Architecture” focuses on green information system (green IS) and explores enter- prise architecture and several related aspects and challenges in the context of supporting organizations to attain sustainability goals for Industry 4.0. Dr. Srikanta Patnaik Director, International Relation & Publication Professor, Department of Computer Science and Engineering SOA University Bhubaneswar, India Editorial vii

Page 9

Acknowledgements I am thankful to Editor-in-Chief of the Springer Book series on “Studies in Big Data” Prof. Janusz Kacprzyk for his support to bring out this volume. I would like to extend our heartfelt thanks to Dr. Thomas Ditzinger, Executive Editor, and his Springer publishing team for his encouragement and support. I am personally thankful to all the authors of this volume for their valued contribution in this upcoming research area. I am sure that the readers shall get immense benefit and knowledge from this volume entitled New Paradigm of Industry 4.0: Internet of Things, Big Data & Cyber Physical Systems. ix

Page 10

Contents Management of V.U.C.A. (Volatility, Uncertainty, Complexity and Ambiguity) Using Machine Learning Techniques in Industry 4.0 Paradigm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Bhagyashree Mohanta, Pragyan Nanda and Srikanta Patnaik Role of Industry 4.0 in Performance Improvement of Furniture Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 R. B. Chadge, R. L. Shrivastava, J. P. Giri and T. N. Desai Imparting Hands-on Industry 4.0 Education at Low Cost Using Open Source Tools and Python Eco-System . . . . . . . . . . . . . . . . . 37 J. Dasgupta Decision Support Framework for Smart Implementation of Green Supply Chain Management Practices . . . . . . . . . . . . . . . . . . . 49 Arvind Jayant and Neeru Decision Support System for Supply Chain Performance Measurement: Case of Textile Industry . . . . . . . . . . . . . . . . . . . . . . . . . 99 Pranav G. Charkha and Santosh B. Jaju A Review Study of Condition Monitoring and Maintenance Approaches for Diagnosis Corrosive Sulphur Deposition in Oil-Filled Electrical Transformers . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 Ramsey Jadim, Anders Ingwald and Basim Al-Najjar Principal Components Based Multivariate Statistical Process Monitoring of Machining Process Using Machine Vision Approach . . . . 145 Ketaki N. Joshi, Bhushan T. Patil and Hitendra B. Vaishnav Green IS—Exploring Environmental Sensitive IS Through the Lens of Enterprise Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Somnath Debnath xi

Page 11

Management of V.U.C.A. (Volatility, Uncertainty, Complexity and Ambiguity) Using Machine Learning Techniques in Industry 4.0 Paradigm Bhagyashree Mohanta, Pragyan Nanda and Srikanta Patnaik Abstract With the fast advancement in technology and induction of information technology and internet into various aspects of organizations, there has been both large scale of both quantitative as well as qualitative changes in industries. These revolutionary changes being done to attain digital transformation for organizational improvements, has been termed as Fourth Industrial Revolution also known as Indus- try 4.0. However, transforming the traditional processes and machineries for digi- tization involves lots of risk factors such as volatility, ambiguity, complexity and uncertainty. The amalgamations of all risk factors can be abbreviated as V.U.C.A. where V stands for volatility, U stands for uncertainty, C stands for complexity and A stands for ambiguity. It’s just like the cancerous cells that are present in each organization, if ignored at an early stage then it can lead to the deterioration of the organization.Hence for identifying anddefining the contingencies,V.U.C.A. canplay a highly pioneer role and thereby avoiding catastrophic results and cascaded issues in an organization. One effective way to manage V.U.C.A. is integrating machine learning (ML) techniques into Industry 4.0 applications. ML acts as a guide to iden- tifying, getting prepared for, and responding to events in each category. In this paper, therefore, we have reviewed the V.U.C.A. terminologies and its importance associ- ated with Industry 4.0 practices; the various effects of V.U.C.A. which are presented in a systematic overview about Industry 4.0 along with its peripherals, challenges, and finally identifying some future research directions. It also includes the different ML techniques with their applications towards each factor. B. Mohanta (B) · S. Patnaik Department of Computer Science and Engineering, Faculty of Engineering and Technology, SOA (Deemed to be) University, Bhubaneswar, Odisha, India e-mail: bhagyashree.mohanta09@gmail.com S. Patnaik e-mail: patnaik_srikanta@yahoo.co.in P. Nanda Department of Information Technology, IIMT, Bhubaneswar, Odisha, India e-mail: n.pragyan@gmail.com © Springer Nature Switzerland AG 2020 S. Patnaik (ed.), New Paradigm of Industry 4.0, Studies in Big Data 64, https://doi.org/10.1007/978-3-030-25778-1_1 1

Page 12

2 B. Mohanta et al. Keywords Ambiguity · Autonomy · Big data · Complexity · Cyber physical systems · Decentralization · Industry 4.0 · Internet of Things · Machine learning (ML) · Uncertainty · Volatility · V.U.C.A. 1 Introduction In past, the introduction of advanced technologies like mechanization, computeri- zation, automation, and digitization in industries lead to industrial revolutions. The main objective of the fourth industrial revolution in any sector is to serve risk free, effortless operations and on-time delivery of services as set previously before pro- duction. Cyber physical systems and Internet of Thing (IoT) concepts are the basic backbone of industry 4.0. More efficient the technology, there is more the chances of huge data generation. So, big data concept with cloud data storage concept are collaborated into industry 4.0 to effectively handle this issue. The present scenario of industrialization can be considered as fourth industrial revolution or Industry 4.0. It comprises of upgraded trends of technologies by keeping the compatibility intact to integrate interactive intelligent systems with the concept of big data. Now a days, almost every organization is in rush for digitization which has lead them to face dif- ferent challenges and struggle. Besides it’s not about digitization only but it’s about different aspects of the risk factors also. With so many challenges, forecasting the future of any organization is really very difficult. Hence, considering today’s busi- ness world with all the challenges concluding the scenario to be a V.U.C.A. world is highly recommendable (Kinsinger and Walch 2012). Every industry has to tackle with these four factors, which are volatility (rapid rise and fall in market), uncertainty (unpredictable market conditions), complexity (difficulties in problem understand- ing) and ambiguity (confusion in market situations) (Bennett and Lemoine 2014a). All the aforesaid four factors are inter-related to each otherwhich can definitely affect the growth of the business for industries. Hence the decision-making power of an organization definitely relies on these V.U.C.A. terminologies and concepts (Bennett and Lemoine 2014b). Basically, these are considered by the higher authorities of any organizations to deal with the possible adverse situations arising in future. 1.1 Industry 4.0 Paradigm Over a period of hundreds of years, the search of innovative solutions for providing competitive advantages in emerging markets has led to various revolutions in indus- trial developments. In current technological development era, with the advancement in digitization of organizations, Industry 4.0 has evolved as a consequence of fourth industrial revolution. Before this fourth revolution in industries, there are three phases of revolutions.

Page 13

Management of V.U.C.A. (Volatility, Uncertainty, Complexity … 3 The phase-I is considered as Industry 1.0 or first industrial revolution, wherewater and steam machines were developed to facilitate the workers efficiency in the era of 1800s. The true mechanization concept was introduced in this period. With the increase in production capabilities decentralized decision making was needed which was previously handled by the owners of their own. The phase-II or second industrial revolution or Industry 2.0 concept was intro- duced in 1910 by Patrick Geddesandwas firstly being used in 1951 by the economists Erick Zimmerman. In this phase of electrical machines were the primary contributors in industrial revolution as they are much more efficient as these takes least time to complete the job and easy to operate and to maintain. The assembly line was firstly introduced and mass production using assembly line systems were became an easy practice in this era of revolution. The phase-III or third industrial revolution or Industry 3.0 was introduced in around 1970. In this era computer aided machines are introduced. Automation is the key concept of this revolution, still it required human intervention. The term Industry 4.0 was coined in 2011 in Germany, for a high-tech project strategy by the government of Germany to promote digitization of industrial manu- facturing system. Industry 4.0 involves automation of data exchange process among manufacturing systems by integrating IoT with cognitive and cloud computing, thus known as Cyber Physical Systems (CPS). Industry 4.0 holds within itself innumer- able paradigms and technologies including collaborative product development, cloud computing, Internet of Things (IoT), Enterprise Resource Planning (ERP), Informa- tion and Communication Technology (ICT), Radio Frequency Identification (RFID) etc. (Brettel et al. 2014; Gruber 2013; Hermann et al. 2016; Ivanov et al. 2016). In a way Industry 3.0 laid the foundation for Industry 4.0, since the realization of the former focused on automation of individual processes and machines while Industry 3.0 involves digitization of almost all physical systems along with their integration with each other to form a digital ecosystem across the value chain. The four phases of industrial revolution has been shown in Fig. 1. The rapid growth of industrialization, induction of IoT concepts with the benefits of cyber physical systems lead to huge amount of data flow in the system. For easy maintenance and processing of these data, big data concept came into demand. 1.1.1 Industrial IoT IoT is a concept that is introduced by British technology pioneer Kevin Ashton in the year 1999. In this smart world almost, everything is connected to each other through highly intelligent interactive systems, so as to form a smart automated world. The concept behind the term “Smart” is, to handle organizational challenges efficiently by effectively utilizing the sensitive information with advance communication and internet technologies. This conceptualizes an ideal concept of smart technologies, termed as “Smart Industries”. The IoT devices interact with other connected devices and act accordingly they get information from one another. This advances the ways of organizational approach that their businesses, industries and markets think and

Page 14

4 B. Mohanta et al. Fig. 1 Four phases of industrial revolution behave and gives them the tools to improve their business strategies. It enables the devices to monitor their overall business processes flow, improve the customer experience, enhance employee productivity, integrate and adapt business models, integrate and adapt business models save time and money, make better business decisions and generate more revenue. The four phases of an IoT system eases and advances the process flow of entire system. In the first phase, generated data are sensed through sensors and actuators; then collected and transferred the sensed data to nearest base station for further analysis and processing. In the second phase, data collected through sensors are aggregated and processed; the data conversion such as analog to digital and vice versa is also carried out. The third phase comprises the edge IT processing systems that may be located in remote offices or other edge locations for pre-processing of the data before it is transferred to the data center or cloud platform. In last phase the data generated are stored, maintained and in-depth processing for future analysis is established. The automated computerized mechanization with highly active sensory devices is the basic building blocks of Industrial IoT (IIoT). IIoT with big data concept jointly known as cyber physical system which led to fourth industrial revolution. It also accelerated the interaction and coordination processes of cyber-physical systems. Here self-governing machines are well equipped with wireless sensors, which can bemonitored remotely through any devices like computers, laptops even frommobile phones. The smart machines are operated through smart technology concepts with least human intervention. The increase of smart industries concept will increase the product customization and flexibility in working process which will impact the consumer-supplier dependencies on these smart self-governing processes. In the near future Industry 4.0 is going to set a new standard in manufacturing industries.

Page 15

Management of V.U.C.A. (Volatility, Uncertainty, Complexity … 5 The rapid increase in digital market and with the increase in consumer’s demand, directly affect the production process. The induction of CPS in manufacturing indus- tries acts as a boon to the production process. In IIoT themed industries all the machineries are automated and interacted throughwireless sensor networks. The sen- sors attached to each machine automatically sense the requirement of the production process and perform its assigned job in time andwith higher accuracy without human intervention. Industries with IoT concepts operate more efficiently that understand customers’ demand to deliver enhanced customer service, improve decision-making and increase the profitability of the business. But it’s not correct to assume that IIoT devices are the replacements of human beings. Rather it enhances the efficiencies of the workers and upgrading their skills. 1.1.2 Big Data In today’s digital world, there is hardly any aspect of business environment that has not been touched by digitization. The induction of new digital technologies to tradi- tional business world has lead to use of online channels and fast algorithms by large number of organizations. This digitalization in industries leads to huge dataflow in the systems. During this changing time, data keeping, processing and maintenance became amajor concern due to large amount of structured and unstructured data gen- eration from different sources. Big data is a concept which can efficiently analyze and process these voluminous and wide verities of data. Literally, Big Data means massive collection of data containing massive information. Although initially Big Data was defined using four dimensions by IBM data scientists (Gartner and Beyer 2011). Now, this concept is mainly defined by 5 dimensions that can be considered as 5Vs of data (Mohanta et al. 2018). 1st one is the Volume: refers to huge amount of data being generated. 2nd one is the Variety: refers to the different kind of structured and unstructured data, generated from different data sources. 3rd one is the Velocity: the rate at which the data are generated and analyzed for further processing. 4th one is the Veracity: refers to data authenticity and the 5th one is the Value: refers to the age of generated data in order to improve the accuracy of the information generated with proper analysis. This resulting information being specific about processes, man- power, sales and marketing, pricing and customer requirements, can be used to make crucial decisions in unfavorable conditions leading to increased returns in different forms. Thus, Big Data is going to fuel the realization of Industry 4.0. In current scenario, Big Data is laying the foundation for digital transformation of all sorts of businesses ranging from small to large scale businesses, thus playing a major role in the development of Industry 4.0. As discussed previously, realization of smart factories with inter-connected machines and infrastructures communicating with each other via global network is the prime goal of Industry 4.0. These smart factories are supposed to make intelligent products which while being used, collect, analyze and transmit huge amount of real-time data. The information, thus gener- ated provides improved insights to provide smart processes which in-turn provides intelligent interactive services to internet-based end-customers.

Page 16

6 B. Mohanta et al. Using Big Data in various industrial functions and applications will provide cost- effective and fault-free process execution by reducing, further enhancing the perfor- mance and quality levels to meet requirements. Again, proper collection and analysis of Big Data will improve competitive productivity in several sectors such as product fabrication, predictive manufacturing, supply chain management, logistics and risk management etc. (Chen et al. 2012; Reichert 2014). As the rise of industrialization may arise the issue of proper data management due to huge data flow and decision making process of any organization due to pos- sible occurrences of V.U.C.A. factors. ML techniques are in recent trends that can efficiently handle the above mentioned issues. They can effectively grasp unknown, volatile, uncertain and ambiguous data without prior training. 1.2 V.U.C.A. Factors V.U.C.A. is a concept which is derived from the military education of US ArmyWar College to manage the volatility, uncertainty, complexity and ambiguity factors of wars. The abbreviation was introduced back in 1991 which changed the nature and condition of wars (U.S. ArmyHeritage and Education Center 2018). Unlike the war’s unpredictable conditions, the nature of business environment is also unpredictable anddynamic. It is a trendymanagerial termstands for volatility, uncertainty, complex- ity and ambiguity which can possibly affect the industrial survival. As Darwin said “Survival of the fittest”, same principle applies for the ever-changing business world that is subjected to vary unexpectedly and challenges the smart intelligent machines which can successfully reorganize these factors and beat these challenges, will only survive (Bennett and Lemoine 2014b). These challenges address the expected future conditions and getting ready to tackle these situations to survive in the competi- tive business world and in highly dynamic industrialization. The volatility refers to highly changeable business environment which lead to uncertainty that refers to the unpredictable business performance followed by complexity in decisionmaking pro- cess (Prem et al. 2016). The volatile, uncertain and complex business environment generates ambiguity via changeable, unpredictable market situations (Bennett and Lemoine 2014a). In Fig. 2, organizational risks are categorized based upon four basic parameters that are, volatility, uncertainty, complexity and ambiguity, which are acronym as V.U.C.A. a. Volatility: It is a term that indicates extreme and rapid fluctuations in business environment. The pace, the volume and the magnitude of change can define as the degree of turbulence it creates, in the business or industrial environments. Volatility is a concept that is created inside an organization; hence, the possible causes of change are known. E.g. price volatility might be an effect of supply- chain risk. Lack of availability of commodity in contrast to its demand flow,

Page 17

Management of V.U.C.A. (Volatility, Uncertainty, Complexity … 7 Fig. 2 V.U.C.A. term categorization can cause price risks. This might result in fall in stock value which directly hit increase in cost. b. Uncertainty: It is a concept that comes from outside the organization; hence, the cause and effect of risks are unknown. The lack of knowledge about the situations causes uncertainty in any field which results an unpredictable future and affects the long-term grown of that organization. Some of the factors that may cause uncertainty in any organization are never-ending customer needs and changes in customer’s tastes and preferences; technological changes; introduc- tion of new trade policies andmultiple barriers to trade; launching of new product as a substitute of currently used product in market etc. c. Complexity: With the rapid industrialization, complexity arises due to the inter- connected parts, networks and procedures within the organization; the external business environment which might even be unidentifiable and contradicting with each other and lead to complexity in decision-making.More the inter-relatedness in an organization, more difficult to understand the cause of problem statement of risks. The complexity may arise due to outsourcing activities (some of them are human resource management, facilities management, supply chain management, accounting, customer support and service, marketing, computer aided design, research, content writing, engineering, diagnostic services etc.) and induction of new supply chain with the introduction of new product range in production. E.g. a successful consumer goods company that contains different brands, products, efficient network of global distribution may lead to complexity in understanding the cause of future risks. d. Ambiguity: If the problem statement lacks clarity, confidence in probability assessments and the diversity of potential results in which the outcome can- not be clearly described; then it is termed as ambiguity in business environment. E.g. when a new product or plan or technology is introduced in the market, then the diversity in customer’s expectations and behaviours may cause ambiguity in decision making to an organization. By conceptualizing these factors today’s industries can be able to outperform in their respective fields. The rapid growth of industrializations and induction of new technologies in business environment has led to high risk factors in every

Page 18

8 B. Mohanta et al. organization. It’s not sufficient if an organization only grows, smartness is needed to challenge the possible V.U.C.A. factors. Today’s techno park revolution is the possible solution to these challenges as it introduces automation, cyber-physical sys- tems, intelligent interactive systems (IISs) and Internet of Things (IoT) concept that includes big data analysis and cognitive computing in different industries through different machine learning algorithms. Machines are trained with different machine learning techniques in accordance to the V.U.C.A. parameters. The smart business concept enhances the intelligent interactive interconnections between machines, devices, sensors and people to communicate smartly through IoT systems. The cyber-physical systems provide technical assistance to humans to tackle the adverse conditions ofV.U.C.A. factors of business risks. Information trans- parency and decentralized decisionmaking power are the twomajor characteristics of cyber-physical systems that automated the process flow in an organization. Today’s machines are trained by expert machine learning techniques (rules); that includes software packages so that proper decision can be taken. Markets like ecommerce, stocks which are highly competitive, volatile etc.; and ever changing dynamic learn- ing of machines by using machine learning techniques, post data analysis through V.U.C.A. will be highly profitable. The continuous attempt to survive in today’s competitive era has led to the rise of innovative solutions in the form of technologies and services. In other words, we can say innovative solutions in the form of smart products and services are driving the global competitive markets which lead to the fourth industrial revolution named as Industry 4.0. Big Data fuels the realization of Industry 4.0 and dealing with Big Data poses several crucial challenges due to its inherent complexities. 1.3 Applied Machine Learning ML comes under data science which can be considered as an umbrella term cover- ing all aspect of data processing like data analysis, preparation and useful decision making in an organization. It is a branch of artificial intelligence, gives machines the ability to learn by their own without human intervention. Machines can adapt automation without being explicitly programmed and act according to that. Online shopping’s recommender system is a real time live application of ML used by almost everyone. Figure 3 depicts the generalizedmachine learning process flow. The process starts from dataset identification and proper analysis of the identified dataset. Based on the V.U.C.A. factors, types of machine learning algorithms are adopted and an analytical model is formed to train the identified test dataset. Lastly the model is executed to get the desired results. a. Data collection: This is the first and most important step of ML work flow. Here, relevant data are collected according to the problem statement. The data collected should be accurate one as per the problem requirement to maintain the integrity

Page 19

Management of V.U.C.A. (Volatility, Uncertainty, Complexity … 9 Fig. 3 ML process flow

Page 20

10 B. Mohanta et al. of entire process of ML model and determine the predictive model’s accuracy to solve the problem statement. If irrelevant data are gathered for the process, then further processing of those data is worthless. The gathered data are considered as the training data. E.g. if the problem statement is to identify an image type, then the data collection needs to be of image type, corresponding to the problem requirement. b. Data pre-processing: In this step, feature extraction engineering is carried out to make the collected incomplete, inconsistent or erroneous data to a feasible one so as to well fit the ML model. After removing the errors like inconsistency, redundancy and missing data from the dataset, features are extracted by feature engineering process and those features are used for model training purpose. E.g. If the problem statement is that to identify a person’s face, then the features that are extracted for training and evaluation is shape of the image or outline of the image like width and breadth of nose, eyes, ears, mouth and face etc. c. Model building: This is the step where appropriate ML techniques are chosen to evaluate the desired result. Different ML techniques are there like supervised, unsupervised and reinforced learning, which are further categorized to many different techniques such as logistic regression, Support Vector Machine (SVM), NaïveBayes, Linear regression, k-Means, Dimensionality Reduction process etc. Some of the techniques are well suited for image data for image processing, some are used for signal process like voice, music, text data etc. d. Model training and testing: After model was chosen, the pre-processed dataset is divided into two parts as training and testing dataset. Usually for smaller dataset train/test dataset ratio is 70/30 or 80/20. With the increase in train dataset size, test dataset should decrease in size i.e. the train/test dataset ratio should be 99/1 for large dataset size. In model training, ML algorithm takes training dataset and learned specific features that will minimizes the error during model evaluation. Once the model is trained, it is then tested on the training dataset which are never been used for training purpose to evaluate the efficiency of the model chosen for prediction and how accurately the model behaves in the real world. It only tries to performon the test dataset by using the knowledge gained from themodel training process. It uses some evaluation metrics such as F1 Score, Mean Absolute Error (MAE) and Mean Squared Error (MSE) to evaluate the test dataset. e. Performance evaluation: In this step, the performance of tested model can be improved on both train/test dataset by using various methods such as cross- validation, hyper-parameter tuning or by trying out multiple machine learning algorithms and using the one which performs the best or even better, by using assembling methods which combine the results from multiple algorithms. f. Model execution: This is the step where the value or output of the ML model is predicted or realized. The model is finally used to predict the desired result with higher accuracy. E.g. in an image classification, if the problem statement is to predict a person’s face, here the model can able to predict or detect the face of the person efficiently with high accuracy.