DEGREE PROJECT IN INDUSTRIAL ENGINEERING AND MANAGEMENT,
SECOND CYCLE, 30 CREDITS STOCKHOLM, SWEDEN 2020
Effects of Digitalization in Steel Industry
Economic Impacts & Investment Model JENNY CHENG
JOSEFIN WESTMAN
Effects of Digitalization in Steel Industry
Economic Impacts & Investment Model
by
Jenny Cheng Josefin Westman
Effekter av digitalisering i stålindustrin
Ekonomisk påverkan & investeringsmodell
av
Jenny Cheng Josefin Westman
Examensarbete TRITA-ITM-EX 2020:280 KTH Industriell teknik och management
Industriell ekonomi och organisation SE-100 44 STOCKHOLM
Master of Science Thesis TRITA-ITM-EX 2020:280 Effects of Digitalization in Steel Industry
Jenny Cheng Josefin Westman
Approved
2020-06-12
Examiner
Hans Lööf
Supervisor
Gustav Martinsson
Commissioner
SSAB
Contact person
Abstract
The awareness and interest for digitalization have increased tremendously during recent years.
However, many companies are struggling to identify the economic benefits and often face long payback time and large initial investment costs. This study aims to assess the potential economic effects from digitalization projects in the steel production industry. The study begins by elucidating central concept like, digitization, digitalization, digital transform and the link between digitalization and automation. Furthermore, the study identifies effects of digitization at production level from an internal efficiency perspective, based on existing literature. On this basis, an investment tool for digitization projects has been developed, consisting of three different analyzes; a level of automation analysis, a quantitative analysis and a qualitative analysis.
To continue, the investment model has been applied to a potential investment of a smart automatic crane. The results from all three analyses provided positive results and incentives to initiate the project. As a result of poor data collection and rigid data, only one effect could be accounted for in the quantitative analysis, which generated a net present value of nearly 12 MSEK over a ten- year period. The most critical parameter proved to be the timing of initiating the project.
Examensarbete TRITA-ITM-EX 2020:280
Effekter av digitalisering i stålindustrin
Jenny Cheng Josefin Westman
Godkänt
2020-06-12
Examinator
Hans Lööf
Handledare
Gustav Martinsson
Uppdragsgivare
SSAB
Kontaktperson
Sammanfattning
Medvetenheten och intresset för digitalisering har ökat enormt under de senaste åren. Många företag kämpar emellertid med att identifiera de ekonomiska fördelarna och möter ofta långa återbetalningstider och stora initiala investeringskostnader. Denna studie syftar till att utvärdera de potentiella ekonomiska effekterna av digitaliseringsprojekt i stålproduktionsindustrin. Studien börjar med att redogöra för vad digitalisering är samt kopplingen mellan digitalisering och automation. Vidare identifierar studien effekter av digitalisering på produktionsnivå ur ett internt effektivitetsperspektiv baserat på befintlig litteratur. Baserat på detta har ett investeringsverktyg för digitaliseringsprojekt utvecklats, bestående av tre olika analyser; en automationsgradsanalys, en kvantitativ analys och en kvalitativ analys.
Investeringsmodellen har dessutom tillämpats på en potentiell investering i form av en smart automatkran. Resultaten från samtliga tre analyser var positiva och utgjorde incitament till att initiera projektet. Som ett resultat av bristande datainsamling och ostrukturerade data kunde kostnadsbesparingen från endast en effekt redovisas i den kvantitativa analysen, vilken genererade ett nuvärde på nästan 12 MSEK under en tioårsperiod. Den mest kritiska parametern visade sig vara tidpunkten för att implementera projektet.
Nyckelord: digitalisering, automation, stålproduktion, automationsgrad, ”discounted cash flow”, multikriterieanalys
Foreword
This Master Thesis report was conducted by Jenny Cheng and Josefin Westman at the Royal Institute of Technology (KTH) at the department of Industrial Engineering and Management, Stockholm, Sweden. The authors are both majoring in Industrial Engineering and Management but with different masters respectively; financial mathematics and sustainable power production. The idea was to combine the diversified competencies and create an outlet for both management and finance. Furthermore, this Master Thesis work was carried out in collaboration with a European special steels company over a five-month period during spring 2020.
Acknowledgments
Firstly, we would like to thank our supervisor at KTH, Gustav Martinsson, Associate Professor in Financial Economics, for always being accessible when we have been in need of support and feedback; both regarding advise on formalities but also in logical reasoning and ensuring the academical level of the work.
Secondly, we would also like to express our gratitude to everyone at the commission company who has been involved in this project, one way or another. Especially, we want to thank our supervisor; thank you for taking your time to have continuous meetings with us and providing useful data. It has been a pleasure to get to know you and the company, and this work would never have been completed without you.
Last but not least, we want to send a great thank you to our friends and classmates at KTH who supported us not only throughout this period, but all five years at KTH, making every day a bit more enjoyable.
We are incredibly grateful for everything you have contributed to enable or facilitate this
List of Abbreviations
AHP Analytical Hierarchy Process AI Artificial Intelligence
CF Cash Flow
DCF Discounted Cash Flow FTE Full-time Equivalent H2M Human-to-Machine IoS Internet of Services IoT Internet of Things
ICT Information and Communication Technologies IRR Internal Rate of Return
IT Information Technology KET Key Enabling Technologies KPI Key Performance Indicator LoA Level of Automation MCA Multicriteria Analysis M2H Machine-to-Human M2M Machine-to-Machine NPV Net Present Value OAT One-at-the-time
OED Oxford English Dictionary
PB Payback Period
ROI Return on Investment RRR Required Rate of Return
SME Small and medium-size enterprise SoPI Square of Possible Improvements TTM Time-to-market
WACC Weighted Average Cost of Capital
Table of Contents
1 Introduction ... 1
1.1 Background ... 1
1.2 Problematization ... 3
1.3 Purpose and Research Questions ... 3
1.4 Delimitations ... 4
1.5 Outline of Thesis ... 5
2 Method ... 7
2.1 Research Design ... 7
2.2 Research Method ... 7
2.3 Data Collection ... 8
3 Literature Review ... 9
3.1 Digital Definitions ... 9
3.1.1 Digitization ... 9
3.1.2 Digitalization ... 10
3.1.3 Digital Transformation ... 10
3.2 Automation and Digitalization ... 10
3.3 History of Industrial Revolution ... 12
3.3.1 Industry 4.0 ... 14
3.4 Steel Industry ... 14
3.4.1 Production Process ... 16
3.3.2 Current State ... 17
3.5 Assessment Methods ... 19
3.5.1 LoA Framework ... 19
3.5.2 Discounted Cash Flow ... 22
3.5.3 Multicriteria Analysis ... 25
4 Effects of Digitalization ... 28
4.1 Approach ... 28
4.2 Quantitative Effects ... 31
4.2.1 Quantified Quantitative Impacts ... 36
4.3 Qualitative Effects ... 37
5 Investment Model ... 40
5.1 Conceptual Overview ... 40
5.2 LoA Analysis ... 41
6.1 Project “Smart Crane” ... 58
6.2 LoA Analysis ... 58
6.3 Quantitative Analysis ... 59
6.4 Qualitative Analysis ... 65
7 Results ... 67
7.1 LoA Analysis ... 67
7.2 Quantitative Analysis ... 69
7.2.1 Sensitivity Analysis ... 72
7.3 Qualitative Analysis ... 73
7.3.1 Sensitivity Analysis ... 74
8 Analysis of Results ... 75
9 Discussion ... 77
9.1 Discussion of Method ... 77
9.2 Reliability & Validity ... 77
9.3 Generalizability ... 78
10 Conclusion ... 79
10.1 Answer of Research Question 1 ... 79
10.2 Answer of Research Question 2 ... 79
10.3 Answer of Research Question 3 ... 80
10.4 General Conclusion ... 80
10.5 Recommendation & Future Research ... 81
References ... 82
Appendix A – Investment Model ... 86
List of Figures
Figure 1 Research Process ... 8
Figure 2 History of Industrialization ... 13
Figure 3 Industry Chart ... 15
Figure 4 Steel Production Process ... 17
Figure 5 Mechanical-Information-LoA Diagram Showing SoPI ... 22
Figure 6 Levels of Digitalization ... 29
Figure 7 Viewpoints for Analyzing Digitalization Impact ... 30
Figure 8 Classification of Effects ... 31
Figure 9 Summary of Quantitative Effects ... 35
Figure 10 Summary of Qualitative Effects ... 39
Figure 11 Conceptual Overview of Investment Model ... 41
Figure 12 Overview of Initial Investment Data ... 60
Figure 13 Overview of Maintenance Savings ... 61
Figure 14 Overview of Productivity Savings ... 62
Figure 15 Overview of Personnel Savings ... 63
Figure 16 Overview of Quality Savings ... 63
Figure 17 Overview of Downtime Savings ... 64
Figure 18 LoA Chart over Investment Potential ... 68
Figure 19 SoPI Results ... 69
Figure 20 Saving Potential ... 71
Figure 21 Savings Per Factor ... 71
Figure 22 Savings Pie Chart ... 71
Figure 23 Quantitative Sensitivity Analysis Result ... 72
Figure 24 Discount Rate Tornado Diagram ... 73
Figure 24 Qualitative Sensitivity Analysis Results ... 74
List of Tables Table 1 Levels of Automation Reference Scale ... 20
Table 2 Summary of Quantified Quantitative Impacts ... 37
1 Introduction
This chapter provides the background information about the research area and aims to increase the understanding of the problem. The purpose of the study is explained, and research questions defined, followed by a presentation of delimitations and the outline of the thesis.
1.1 Background
Rapid changes in the digital technology is revolutionizing the industries and the society (Snabe Hagemann & Weinelt, 2016). The impact of digitalization is major, and many companies believe it is vital to follow the digitalization trend in order stay competitive in terms of effectiveness, growth and prosperity (Vernersson et al., 2015). There are several consequences, but also possibilities, followed by the industrial digital transformation.
Today we are currently entering a new technological paradigm, the next industrial revolution, Industry 4.0, where we transform towards an industrial internet with smart devices, higher flexibility and larger applications (Vernersson et al., 2015).
The steel industry is no exception and is undergoing tremendous digital transformations today, even though it seems like the steel industry in many aspects lag behind other industries when it comes to digitalization. The steel industry alone accounted for 3.8% of the annual global GDP in 2017 and contributed to over 6 million employments the same year (Oxford Economics, 2019). The industry is both capital and human capital intensive, resulting in certain rigidity. Therefore, it seems only reasonable that transformations within steel industry would require more time. On the other hand, large corporations hold some benefits over small and medium-size enterprises (SMEs), where they can utilize scale advantages and afford knowledgeable IT specialist to accelerate the transformation. In order to reach higher production efficiency, more competitive products and better business models, Key Enabling Technologies (KET) such as; Artificial Intelligence (AI), Internet of Things (IoT), Internet of Services (IoS), Mechatronics and Advanced Robotics, Cloud Computing, Cybersecurity, Additive Manufacturing and Digital Twin has been or will be used. These KETs together build the foundation of digitalization, which in turn is the core of Industry 4.0 and has become more popular than ever. (Murri et al., 2019)
The importance of digitalization and Industry 4.0 are well known and the technological shift in the industries is inevitable. Bill Ruh, Chief Executive Officer, GE Digital, USA believes it is a now or never chance to act (Snabe Hagemann & Weinelt, 2016), but the question is what the benefits from these actions are. It is rather easy to find both articles and other studies dealing with the subject digitalization. However, it is difficult to find studies that examine the economic impacts of digitalization and more specifically the economic impacts of digitalization in steel industry. A big contributing factor to this fact is that it is hard to identify the economic impacts from digitalization projects. Projects are often very costly and require large capital investments while it is expected to meet short payback requirements set by stakeholders. (Murri et al., 2019) According to the European Steel Skills Agenda, the steel industry faces several barriers; difficulty in integrating new technologies and processes among site workers, a strong age gap between current employees and prospective employees creates knowledge transfer issues and lack of investment in training and education from steelmaking companies as well as an insufficient amount of in-house training provided by companies (Henriette et al., 2015).
Digital transforms affect the entire organization including the business model, operational process and both internal and external stakeholders (Stolterman & Fors, 2004). Even though the challenges are many and it is shown that technical barriers are less crucial than organizational issues (Branca et al., 2020), digitalization is still something highly valued.
Companies must try to find ways to quantify the benefits of these kind of projects, but if it cannot be done, the companies should ask themselves what they lose by not adopting to the new technological shift rather than what they gain (Bossen & Ingemansson, 2016).
To conclude, production managers often foresees high potentials with new digital solutions, while management is struggling to identify potential profit, preventing rapid digitalization progress. Therefore, the desire for economic reason behind digitalization is undeniably great in most industries. Increased popularity and utilization of digital technologies leads to an incentive for several well-known journals and consultancy
1.2 Problematization
There is no doubt the majority has a strong belief that digitalization has a net positive effect on the entire organization. The unlimited number of reports, case studies and articles addressing positive impacts of digitalization creates a thrive for companies to follow the trend. However, papers dealing with quantifying economic aspects of digitalization are scarce. Furthermore, studies on current state of digitalization in steel industry in particular are limited as well. Therefore, researchers and steel companies find it difficult to quantify the actual effects of digitalization.
Furthermore, the notation digitalization is widely used in everyday language, contributing to a confusion regarding what it actually comprises. Therefore, it is important to factorize, concretize and specify the definition of digitalization in order to estimate the potential economic impacts. The quantification of these impacts is obstructed by the uncertainty of possible aggregated effects enabled by extension projects as well as the difficulty to identify synergies from future integration of subprojects. As we currently are in the middle of the digital transformation, the opportunity to compare potential outcomes with historical data is highly limited and further increasing the level of difficulty.
At last, it is proven that digitalization projects have both long pay back times and contribute to many soft term consequences, implying even higher uncertainty in calculations. For all reasons stated above, it seems difficult to quantify obvious impacts and to address less prominent varying soft term factors. This leads to financial uncertainties and difficulties to justify the implementation of these projects.
1.3 Purpose and Research Questions
The overall purpose of this paper is to partly solve few of the obstacles digitalization brings, described in the background and problematization sections. This study aims to identify potential impacts of digitalization within the special steels industry, in order to address relevant saving opportunities and finally draw strategic conclusions. We aspire to develop an investment model where the relationship between future digitalization projects in a delimited steel manufacturing process and different cost saving factors will be carefully examined through the lens of economic KPIs and other qualitative metrics. The
intention of the model is to be used as a tool to help steel companies make well-grounded digitalization investment decisions, taking not only the most obvious but all possible effects into account.
With the problematization and purpose as a foundation, the following research questions will be considered:
Q1: What are the potential impacts of digitalization in a delimited steel production section?
Q2: How can potential impacts from digitalization projects be quantified?
Q3: What potential cost savings can be expected from digitalization projects?
1.4 Delimitations
This report mainly focuses on digitalization projects at Process level which will be studied from an Internal Efficiency perspective, based on the two frameworks developed by Tihinen et al. (2017). Digitalization is implemented at Process level when it facilitates the adoption of digital tools and streamlining processes by reducing manual steps. Process level is thereby directly connected to the production department of a firm. When digitalization is studied from an Internal Efficiency perspective, it is analyzed with regards to how it improves the ways of working through digital means and by re-planning of internal processes, see section 4.1 for further explanations. Digital implementations at any other levels will not be considered, and projects will mainly be evaluated from this certain perspective.
The investment model developed in this paper have been designed for valuation of potential digitalization projects in a delimited production process within the special steels industry. Projects that change or affect the organization in its whole and projects only utilizing digital technology without generating a higher level of digitalization are not
1.5 Outline of Thesis
This thesis consists of ten chapters, which are briefly presented below.
1. Introduction: This chapter gives the background information to the research area and creates an understanding of the problem. The purpose of the study is explained, and research questions defined followed by a presentation of delimitations and the outline of the thesis.
2. Method: This chapter presents the methodological approach and method chosen. A conceptual visualization of the research process is given in order to make sense of the logics and connections of different parts. An exposition of how data is collected and utilized is provided as well.
3. Literature Review: This chapter consists of a literature review comprising relevant knowledge for the subject of the thesis. Necessary concepts are defined and the background to digitalization and its origin is given. Furthermore, basics of the steel industry are explained and useful frameworks for the investment analysis are presented.
4. Effects: This chapter explains the approach from which effects of digitalization are identified and describes the underlying frameworks. Potential effects are identified in the existing literature based on the identified approach.
5. Investment Model: This chapter contains a presentation of how the investment model is developed based on three analyses; LoA Analysis, Quantitative Analysis and Qualitative Analysis. The model is built based on findings from the literature review together with insights from the case study company, a European special steels producer.
6. Application of Investment Model: This chapter is directly referring to the case study conducted at a European special steels company, aiming to answer the research questions of this study. One specific potential investment is considered, and all data presented in this section is collected at the case study company. Moreover, a detailed explanation on how it is supposed to be used is given.
7. Results: This chapter provides a presentation of the results from all three different analyses of the investment model. The main results are shown in terms of NPV, IRR, ROI, PB, SoPI and qualitative indexes. Results from sensitivity analyses are also presented.
8. Analysis of Results: This chapter is an overall analysis of the results in chapter 7, with theoretical findings in the literature review as a starting point. Results are being triangulated in order to broaden the understanding of their implications.
9. Discussion: This chapter contains a discussion and argumentation of the research method used in this thesis. Furthermore, the reliability, validity and generalizability of the model, as well as the thesis in general is discussed.
10. Conclusion: This chapter answers the stated research questions of the thesis and explains how answers were arrived at. It also provides a summary of the main findings on a higher level as well as a recommendation for producing companies and suggestions for future research.
2 Method
This chapter presents the methodological approach and method chosen. A conceptual visualization of the research process is given in order to make sense of the logics and connections of different parts. An exposition of how data is collected and utilized is provided as well.
2.1 Research Design
This research is primarily descriptive in essence, as it attempts to “determine, describe or identify what is” rather than why something is or how it came to be (Ethridge, 2004). We aim to collect data and information that enables a better and more complete description about the impacts of digitalization projects. Descriptive research is effective for analyzing non-quantified topics and issues, and it also gives opportunity for integrating qualitative and quantitative methods of data collection, where case-studies are one commonly used data collection method. Furthermore, a deductive approach is taken for conducting this descriptive research, meaning that reasoning goes from the general to the particular. Using a deductive approach is advantageous for explaining causal relationships between concepts and variables as well as for measuring concepts quantitively. Due to the nature of the chosen field of study, this approach is considered suitable for appropriately address the stated purpose and research questions. (Ethridge, 2004; Fox, 2007)
2.2 Research Method
The research method can be seen as a systematic roadmap to how research is planned to be conducted. The project will be conducted based on a mixed method, where the aim is to combine a qualitative single case study with quantitative findings in the literature to fulfill the stated purpose. Data will be collected using both existing literature as well as the single case study to iteratively develop a quantitative investment model for evaluation of digitalization projects.
The first part of the study consists of a qualitative pre-study where information and data will be collected by conducting a literature review. This literature review will consist of three main parts; defining relevant concepts and their origins, explaining the steel industry and identifying successfully used frameworks for evaluation of projects. Areas of our
particular interest are for instance digitalization, digitization, digital transforms, automation and steel industry production processes. In addition, existing literature will be examined in order to identify potential effects of digitalization.
The aim of the single case study is twofold; firstly, one aim is to identify additional factors to include in the model that were not covered by the literature, by observing the production line. Secondly, primary data will be provided by the company in the case study, which will be used for verification of the investment model and for applying the model on a real case in a specific subprocess in production.
The merged data collection from the literature review and case study will form the foundation of our study and the quantitative investment model. After identifying crucial economic consequences of digitalization, the investment model will be built in the software Excel. The outcome of the model when applied in the case study situation shall be carefully examined, and a sensitivity analysis will be established. Results will be compared with the literature and analyzed so that useful insight and conclusions can be drawn. Study of literature and model construction will be an iterative process where all our findings should be anchored in the literature and not only based on hypotheses from the case study company. A conceptual overview of the research process can be found in figure 1 below.
Figure 1 Research Process
2.3 Data Collection
The collection of data will be derived from two channels; secondary data from existing
3 Literature Review
This chapter consists of a literature review comprising relevant knowledge for the subject of the thesis. Necessary concepts are defined and a background to digitalization and its origin is given. Furthermore, basics of the steel industry are explained and useful frameworks for the investment analysis are presented.
3.1 Digital Definitions
Digitization, digitalization and digital transformation are closely related concepts and often interchanged in a way that shortchange the power and importance of digital transformation.
The definition of these digital concepts is scattered and diffuse. These words are wrongfully used as synonyms in everyday language and depending on whom you ask the answer of the definitions will vary. Most people are confident when speaking about digitization and digitalization since the notations are frequently used in both the academic world and everyday life. However, the close association is triggering confusion and not even the researchers agree upon a standardized definition. Thus, the unclear definition could be a smaller contributing factor to why many companies struggle to see the potential and benefits the transformations really brings. The truth is that neither of the three terms are synonyms, but indeed very closely related.
3.1.1 Digitization
Most people agree upon the definition of digitization established by the Oxford English Dictionary (OED) and the straightforward definition is “…the conversion of analogue data (esp. in later use images, video, and text) into digital form”. (Oxford English Dictionary, 2016). The process of digitizing could for an example be the conversion of handwritten papers to digital documents or conversion of LP and VHS to Spotify and Netflix. In other words, digitization could also be defined as “the ability to turn existing products or services into digital variants, and thus offer advantages over tangible products” (Stolterman & Fors, 2004). The last definition is closer related to digitalization since the conversion of a good or service to a digital variant may be argued to change the whole business model for some companies e.g. Netflix, HBO etc. However, the aim of a digitization project is rarely to change the value proposition or the business model in order to create new revenue streams and it does not include the organizational transformation needed to adopt to the new digitized solution.
3.1.2 Digitalization
The OED states that digitalization is “the adoption or increase in use of digital or computer technology by an organization, industry, country, etc.” (Oxford English Dictionary, 2016).
Digitalization it is not only the digital technology in itself, where information is represented in bits, it is “the use of digital technologies in order to change a business model and to provide new revenue and value producing opportunities.” (Bloomberg, 2018). The core of digitalization also includes the digital skills and reorganization needed to implement a new digital solution. Digitization is a prerequisite for digitalization and plays a key role in such processes. For an example, the conversion from manual manufacturing to smart manufacturing is a digitalization process where the employees need to change from working with physical equipment to managing a computer program and handle new problems like cybersecurity and transparency.
3.1.3 Digital Transformation
Digital transformation is far beyond digitization and digitalization. According to Stolterman and Fors (2004) digital transformation is “…the changes that the digital technology causes or influences in all aspects of human life.” (Björkdahl et al., 2018).
Another literature states that digital transformation refers to “the customer-driven strategic business transformation that requires cross-cutting organizational change as well as the implementation of digital technologies” and cannot be implemented as a project. A digital transformation often includes several digitalization projects at the same time. (Bloomberg, 2018) The organization should thrive to restructure the whole organization in order to more effectively benefit from data, create new values and finally acquire some of the economic value that it has created (Fasth et al., 2008). Only when the norm is adjusted to the new digital technologies and work ethics, the transformation is considered complete.
3.2 Automation and Digitalization
technology by which a process or procedure is accomplished without human assistance”.
This definition allows not only machines and computers to be a part of automation, but also communication systems and other digital systems that help reduce the need of human assistance in a process. Consequently, digitalization is an important tool in order to increase the level of automation in production systems
Due to the definition, automation is not only about transforming manual processes to automatic ones but also about transforming them into completely autonomous systems with no need of human assistance, which is what defines a 100 % automatic system or process. However, the main purpose with automation is to achieve increased system efficiency, in that regard 100 % automation is not always the best solution. The aim is to target most appropriate level of automation in each manufacturing situation, rather than the highest level possible, as a certain mix of machines and human interaction may be the more efficient solution. (Ten & St, 2015; Tihinen et al., 2017) It may sound surprising that the level of automation can be “too high”. In fact, excessive levels of automation may result in weak system performance, (Endsley and Kiris 1995; Parasuraman et al. 2000) as a result of too complex processes. Complex processes are often more vulnerable to disturbances, which might decrease the overall production efficiency (Ylipää 2000). It may also be that production tasks are too unstructured to be fully automized. On the other hand, if the level of automation is too low production efficiency is not maximized. A low level of automation could also cause working injuries and sick leaves.
An arrangement where devices and components communicate through a continuous flow of information is commonly called Machine-to-machine (M2M) interaction, which is appropriate when tasks benefit from automation. Furthermore, in cases where higher levels of automations are inappropriate and human interaction is preferable, the arrangement is called Human-to-Machine (H2M) collaboration. In addition, research efforts are invested in so-called Machine-to-Human (M2H) communication or “collaborative robotics”. Here, complex and unstructed manufacturing tasks are performed in collaboration between advanced specially designed robots and humans. The goal with these highly advanced technologies is to enable automation for tasks that earlier was preferred to be performed totally manual. (Rojko, 2017)
A commonly used framework for measuring the level of automation is the LoA framework, which is described in more detail in section 3.5.1. The LoA framework evaluates the level of automation based on two grounds; one mechanical and one informational, where the informational part is closely related to digitalization.
3.3 History of Industrial Revolution
To create a better understanding of the concept of digitalization and its impacts it is important to derive all the way back to its origin. The source of the contemporary concept can be derived to centuries ago and started with the first industrial revolution. Some basic components of digital transformation are machinery, electricity, automation and knowledge. The process from manual manufacturing by manpower to smart mass production executed by smart machines, operating using own mental power is over two and a half decade long. The industrial revolution did not only change how companies produced goods, how people lived and how people defined political issues, it basically changed the whole world. (Rojko, 2017)
The definition of industrial revolution can be divided into two parts. First, industrial revolution incudes a large collection of transformations with origin in radical technological innovations. Second, it infers organizational reforms changing manufacturing industries, leading to widely established innovations changing the economy at large. (Gassmann et al., 2014)
The first industrial revolution developed in Britain during late 17th century, followed by western Europe and United States. Eventually, places such as Russia, Japan and southern Europe unfolded the concept of industrialization. It is indeed difficult to determine an exact year when the different industrial waves bursted out, since industrialization occurred during different times at various places. What could be done is to identify when the concepts developed and started to become more widespread and in the figure 2, an
power and weaving looms driven by power. The steam engine was constructed to extract energy from heated coal in order to create steam and the power looms did no long need human assistance as the foot pedals were replaced. The revolution enabled more efficient manufacturing, but also brought groups of people together and created sense of solidarity.
The steam power discovery was followed by electricity and factory production in late 18th century, which was the key invention of the second revolution. (Henriette et al., 2015)
The third industrial revolution, also the so-called digital revolution took place a century after the second and most producing companies could now benefit from mass production, line production and the importance of automation became more essential. (Tihinen et al., 2017) During this paradigm Information Technology (IT) started booming and analogue technology was transformed to digital. Central innovations as integrated circuit chips, computers, microprocessors, cellular phones and internet transformed the traditional production and created a foundation for future digitalization. (Rojko, 2017) Industry 3.0 allows flexible production, higher variety of products and programmable machines, however flexible production in terms of quantity was still a limitation. (Rojko, 2017) Today the western countries just entered the Fourth Industrial Revolution that originally emerged in Germany and was provoked by the fast growth of Information and Communication Technologies (ICT). Central to this era is smart automation of cyber- physical systems leading to decentralization within the organization and more advanced data connection systems, which in turn enables higher flexibility within mass custom production and in production quantity.
Figure 2 History of Industrialization
3.3.1 Industry 4.0
Industry 4.0 differ considerably from previous industrial happenings in the history. It is not just another disruptive technology or yet another industrial revolution. The fourth industrial revolution is a thrive to change into something unknown and implies using Industry 4.0 strategy to sustain competitive in the market. The revolution was announced prior to its implementation and not after it was fully established, which is one main difference to previous industrial revolutions. (Rojko, 2017).
As mentioned, the fourth industrial revolution was triggered by the digitalization upswing and development of ICT, but also saturation of the market, which forced the emergence of new solutions. Production cost have been diminished by lean production and concepts of just-in-time production and even more by outsourcing production to developing countries offering lower work cost. (Björkdahl et al., 2018) The new paradigm with robotic, digital and automatic technologies allows lower production cost in developed countries such as Sweden and not only in low cost countries. (Rojko, 2017) The main idea is to seize the potential of new technological concepts such as internet, IoT, integration of technical and business processes, digital mapping and smart manufacturing, to minimize costs. (Bossen
& Ingemansson, 2016)
However, there are difficulties to identify potential impacts of Industry 4.0 and the implementation of new technology in the early process. The benefits from industrialization and digitalization may be recognized centuries after its implementation and some intermediate steps in the process are required in order to enable later innovations. It is possible that some steps in the transformation process are nonprofitable at first, even if the whole solution in the end is a positive investment. A bottle neck in industrial transforms are to identify the financial gains and the economic impacts, since it takes time to realize profits from over-time projects contributing to many soft term consequences.
3.4 Steel Industry
the largest actor on the market alone stood for more than 50 % of the total steel output. A common global challenge is the large CO2 emissions the production entails. On average every ton of produced steel yields 1.83 tons of CO2 emissions. (World Steel Association, 2019). The steel industry is a subindustry of the manufacturing industry which in turn is a subindustry of a larger process industry, as shown in figure 3. The manufacturing industry covers all manufacturers producing products by converting raw materials or commodities, often in large scale, for example textiles, machines, equipment etc. While processing is a broader term and could be defined as series of mechanical or chemical operations to change or preserve something. Food is for example, processed and not manufactured.
Figure 3 Industry Chart
Steel in particular is manufactured using an alloy of iron and carbon, which sometimes also includes other alloying elements in order to obtain different characteristics. It is used in buildings, infrastructures, automobiles, machines etc. Some advantages of steel are, it is possible to mold it plastically in both cold and hot conditions, harden it in multiple ways, use alloying elements in order to change the properties of the steel and recycle most of the materials. There exist three typical variations of steel; carbon steel, low alloy steel and high alloy steel. Each type of steel holds different characteristics and are used for different purposes. Furthermore, the variation of steels can be categorized as either commercial steels with plain carbon and no alloys or special steels produced for special purposes with different alloys.
3.4.1 Production Process
Today there mainly exists two different ways to produce steel and the process varies by the raw materials and the furnaces process. The traditional way to produce steel is to use iron ore and a blast furnace. However, today’s technology also allows us to reuse scrap steel. When using scrap steel in the production process, electric arc furnaces are used instead, where electricity is forced through an arc enforcing desired result and temperature.
Both methods can be described by the modern steel making process, which can be divided into six steps and in a primary and secondary steel making phase. Please find illustration of both methods in figure 4 below.
The first step is iron making where iron ore is reduced using coke and coal in a blast furnace with high temperature, this way molten iron is produced. At this stage there are still many impurities in the molten iron, so a smaller amount of scrap steel is infused. In the primary steel making phase, oxygen is forced into an LD-converter, causing a temperature rise to 1700 Celsius degrees (World Coal Association, 2019), which reduce the carbon impurities by 90 % and the molten iron is transformed to molten steel. (Melfab Engineering, 2017) This process in particular gives rise to a high amount of carbon dioxide emissions. (SSAB, 2020) When only using scrap, the two first stages will be reduced by an electric arc furnace, since the scrap steel already holds some of the desired characteristics. Following step is the secondary steel making where more specific properties of the steel is determined, in a so-called ladle, by de-oxidation, alloy addition (boron, chromium, molybdenum etc.) and other operations ensuring the exact quality. (Wikipedia, 2020) Next in the casting, the molten steel is tapped into cooling molds, drawn out and finally cut into desired length, before completely cooled. When it is fully cooled it is transported for primary forging, where the casts are formed in a hot rolling process. Here, small defects can be corrected, and the optimal quality is ensured. Sometimes a secondary forming is necessary and operations like coating, thermal treating, pressing etc. is performed in order to get the correct shape and finish.
Figure 4 Steel Production Process
In this paper, the main focus will lie on potential digitalization projects in the last steps of the steelmaking process, i.e. continuous casting, rolling and coating.
3.3.2 Current State
Even though the steel industry, as a part of the process industry, lies far behind the automotive and traditional manufacturing industry when it comes to digitalization, they see high potentials with future transformation projects (Björkdahl et al., 2018). The process industry has in general more strict manufacturing processes and products with less flexibility- Therefore, the current focus is to digitalize the value chain rather than the product itself. Research believe that more focus on surveillance, control and optimization of value chain can result in higher resource efficiency in energy, environment, transport and raw material management. The Swedish steel industry is currently focusing on higher value-added products (Björkdahl et al., 2018) where they compete with production efficiency and capacity. Thus, the greatest driving force in the steel industry is internal cost saving and the goal is to reach a more even production flow with higher automation levels through digitalization. Even though the investments are extensive, many companies have a positive believe that these investments are profitable and look forward to implementing concepts of smart manufacturing such as auto corrections and Machine to Machine communication (M2M) (Murri et al., 2019).
Today the steel industry is in general very energy intensive. However, the European steel industry is characterized by modern energy and emission efficient plants and make fast
progress towards a carbon dioxide free production (Bossen & Ingemansson, 2016). With Big Data analysis the steel industry can expect a more energy efficient production with only small efforts (Björkdahl et al., 2018). Currently, most actors on the steel market have a connected melting process where they can collect measure points such as temperature.
Some also take measurements for quality and productivity related factors in order to understand the relation between the production process and material characteristics, and thereby developing products with higher quality. Another company highlights the importance of the interface between the raw data and the user and most companies collect large amounts of data but does not utilize it in a user-friendly way. One example of such user-friendly interface is a mobile app that shows the current states of different furnaces.
(Murri et al., 2019)
Downstream production areas such as rolling and coating are the processes most affected by digitalization and Industry 4.0 (Neef et al. 2018). The technical barriers are considered less problematic than the organizational issues. As a conclusion, the main challenges are legacy equipment, long payback time, data security and uncertainty about impacts on jobs.
Another challenge is the aging of workforce where many of the existing employees possesses great industry knowledge, but on the other hand lack digital knowledge like programming skills. (Gassmann et al., 2014) The resistance to change, learning and collaborate is giving the companies a hard time to get through the digital transformation without replacing parts of the staff. In a modern rolling production, using cameras and other digital solutions as decision support, the employees are younger and hold both computer and multilanguage skills. Meanwhile the traditional rolling production facility consist of higher average age of employees where every individual possesses skills that are harder to pass forward onto new employees.
3.5 Assessment Methods
Most companies have a large number of potential projects competing to be implemented.
Project proposals usually grows from multiple levels; top management, head of departments and people working on the floor all possesses creative ideas about how to improve the business. In order to make well-grounded decisions about which of all projects to initiate, they need to be evaluated on a structured basis. As this report aims to take both quantitative and qualitative effects into account, the investment model consequently needs to consist of two main analyses; one quantitative and one qualitative analysis.
The quantitative analysis includes aspects that could be described in monetary terms while the qualitative analysis includes more vague aspects that are more difficult or even impossible to explain in monetary terms. In general, a variety of both monetary and nonmonetary objectives may influence a decision, which is the reason why qualitative analyses are usually developed side by side with economic costs and benefits analysis to include both aspects. As this thesis consider digitalization projects specifically, it is in addition interesting to evaluate the change in level of automation, see section 3.2 for explanation of how automation and digitalization are related. Theories building the foundation of the quantitative and qualitative analysis as well as how level of automation can be measured, will be explained in the sections below. These theories form the basis for the investment model developed in this thesis.
3.5.1 LoA Framework
One common framework for evaluating the level of automation in manufacturing processes is the Level of Automation (LoA) framework that was developed in the DYNAMO project between 2004 and 2007, carried out in association with Chalmers University of Technology, Jönköping School of Engineering, and IVF Industrial Research Corporation. The LoA framework is a tool to measure and get an overview of the level of automation and current information flows in production systems. It is built on a concept assuming that tasks in manufacturing include both mechanical and cognitive activities. The mechanical activity refers to the physical part of the task and are represented by the Mechanical LoA, while the cognitive activity refers to the data and information exchange which is represented by the Information LoA. The reference scale for different LoAs is ranging from 1 – 7, corresponding to different levels of automation ranging from totally
manual to totally automatic. An overview of the LoA reference scale is shown in table 1 below. (Granell et al., 2007)
Table 1 Levels of Automation Reference Scale
To enable better understanding of the different levels in the reference scale, a short explanation of each level will be given. Starting with Mechanical LoA, level 1 suggests for tasks to be “totally manual”, meaning that it is performed entirely by man-force. For instance, this level could apply to manual lifts in production. The second level, level 2, refers to “static hand tool” which for example could be about using a screwdriver to tighten a screw. Level 3, “flexible hand tool” would instead be the level of automation if a wrench was used for this matter, as it can be set in different ways and thereby perform a variety of operations. Next level, level 4, says “automatic hand tool” and if following the same example as for previous levels, this level suggests using an electric screwdriver to complete the task. Another example of level 4 would be usage of a crane. For level 1 – 3, the work has been performed manually by man-force but with more or less helpful and flexible tools.
From level 4, tasks are supported by some sort of automation, meaning the task no longer
or thicknesses. To reach level 7, a totally automatic machine is used to perform a task, that automatically adjusts its settings depending on the situation. AI, M2M and big data are inevitable technologies that need to be considered, in order to reach LOA above 6.
Continuing with Information LoA, the first level “totally manual” applies to when the person performing a task finds their own way of working without any informational exchange. In other words, when there are no instructions available for how a task should be performed. One example of this level is when the quality of a painted sheet of steel is inspected with a person’s eyes only, without any specified routines for how it should be assessed. Moving on, level 2 is when information is used in a decision giving matter, where the person performing a task receives suggestions on the order of actions. The informational exchange focuses more on mediating what should be done rather than how.
One example of this level is when employees conduct their work based on a working order that suggests them what to do. The third level, “teaching”, is when the worker receives instructions for how a task should be performed, for example by checklists or manuals.
Next level, level 4, is explained as “questioning” and can be considered the first level of human-machine interaction. This level applies to when a system or machine generates questions in order to ensure that correct settings are selected. For instance, it could be that an employee changing the settings in order to produce another product type, whereupon the machine asks “do you really want to change from X to Y?” before resetting production settings from X to Y. Level 5 refers to “supervising”, referring to all kinds of alarm systems and other control systems that calls for workers attention if an abnormal situation arises.
The sixth level is when the technology is “interventional” and takes its own command if necessary. An example could be using sensors for automatic control and adjustment of a task. The highest level, level 7, is reached when a system is totally automatic with no need of human interaction.
On a higher level, there are two main steps when using the LoA framework. The first step is to measure Mechanical LoA and Information LoA for different tasks in production in order to define the current state of automation. The measurement process for determining levels of Mechanical LoA and Information LoA for a specific task in production will not be considered in this paper. Please find the report “Measuring and analysing Levels of Automation in an assembly system” by Fasth et al. (2008), which gives a more detailed explanation about how measurements should be done.
The second step is to assess relevant minimum and maximum values of LoA for each operation. By determining relevant values, an area of automation potential can be defined, to which observed LoAs from on-site measurements should be compared (Björkdahl et al., 2018). Example of such area could be found in the Mechanical-Information-LoA diagram in figure 5, where the vertical and horizontal lines correspond to relevant minimum and maximum values for Mechanical and Information LoA respectively and the black spot represents the observed value. The defined area forms a square, which are called “Square of Possible Improvements” (SoPI) and sets the boundaries for possible automation improvements, with regards to a company’s requirements. SoPI can indicate how to take advantage of the automation potential and help assessing the current state with regards to its future potential.
Figure 5 Mechanical-Information-LoA Diagram Showing SoPI
3.5.2 Discounted Cash Flow
An economic valuation of an investment is the analytical process of determining its current
The discounted cash flow (DCF) method is a commonly used valuation method used for valuating a company, a project or an asset, that is suitable for both financial investments as well as for industry investments. This method takes the time-value of money in account and is therefore appropriate for any situation where money is spent in the present with expectations of receiving money in the future. The valuation is based on finding the present value of the expected future cash flows of an investment, which is done by using a discount rate. When conducting a DCF analysis the investor must estimate future cash flows and an appropriate discount rate. Please find equation (1) for DCF calculations where DCF = discounted cash flow, CF = cash flow and r = discount rate. (Chen 2020a)
!"# = "#!
(1 + ()!+ "#"
(1 + ()" + ⋯ + "##
(1 + ()# (1)
The value of an appropriate discount rate can vary depending on the situation but needs to be sufficient enough to cover the required rate of return of an investment, when taking risk and time-value of money in consideration. One discount rate that is commonly used by companies is the weighted average cost of capital (WACC). WACC is the overall required return of a firm, calculated by its cost of capital proportionally weighted between the two categories equity and debt. However, any discount rate could be used in the DCF analysis, as long as it is an appropriate reflection of the required rate of return (RRR). (Chen 2020a)
Based on the DCF method, different perspectives can be used for comparing investments with each other as well as for deciding which ones to pursue. Some commonly used analyzes are net present value, internal rate of return, payback period and return on investment which will all be given further explanations in the below sections.
Net Present Value
The general perception is that assessing an investment based on its net present value (NPV) is very effective when it comes to evaluating projects as it takes the time-value of money as well as risk in consideration NPV is calculated by summarizing the discounted future cash flows, which is the present value of future cash flows, and subtracting the initial investment cost. If the present value of cash flows is equal to or exceeds the initial investment cost, the investment should be considered. In other words, a positive NPV
