1
Development and
Analysis of Automated
Data Collection
Solutions for Tool
Maintenance in CNC
Machines
A Case Study in a Swedish SME
PAPER WITHIN Final Project Work in Production Systems AUTHOR: Ioana Andreea Cocis and Cezar Adrian Stefan TUTOR: Kerstin Johansen
2
This exam work has been carried out at the School of Engineering in Jönköping in the subject area Production system with a specialization in production development and management. The work is a part of the Master of Science program.
The authors take full responsibility for opinions, conclusions and findings presented.
Examiner: Milad Ashour Pour Supervisor: Kerstin Johansen Scope: 30 credits (second cycle) Date: 2021-06-01
3
Abstract
Since small-to-medium enterprises (SMEs) represent a backbone for sustainable economic growth, the need for increasing the degree of automation in production plants is pushing today’s engineers to find customized and cost-efficient solutions. In the Industry 4.0 context, SMEs which are aspiring towards quality, competitiveness and overall customer satisfaction require to work with a systematic approach for implementing cyber physical systems (CPS) that are functioning autonomously and independently from human interaction. An essential area within manufacturing is represented by the maintenance of CNC machine’s tools that are directly responsible for achieving high-quality products. Since corrective maintenance is associated with uncertainty, unforeseen costs and increased downtime, the focus is diverted towards uprising trends of preventive and predictive maintenance policies. The purpose of this thesis is to increase knowledge in the field of new maintenance practices in the context of Industry 4.0 and to provide a framework for achieving customizable automated solutions adapted for the needs and requirements of each individual SME. The work has used a guideline which is following to transform the needs of SMEs into technical specifications which are used for designing and generating conceptual solutions whose effect is assessed from technical, economic, socio-environmental, operational and schedule perspective. For this matter, a case study has been conducted in a Swedish SME where there is a need for developing an automated solution for data collection in order to enable preventive and predictive maintenance. For achieving this purpose and contribute to filling the gap in academic research on this matter, a framework for development and analysis of automated data collections is created based on the systematic practices of concept development methodology to which a supplementary feasibility analysis is added for increasing the understanding of their impact. Furthermore, this framework is conceived to offer a long-term holistic view of the future changes resulted by the implementation of automated solutions.
Keywords: Preventive Maintenance, Predictive Maintenance, Industry 4.0, Small-to-medium
4
Acknowledgements
We would first like to thank our supervisor from School of Engineering, Kerstin Johansen, for the opportunities she provided for us and the close collaboration we have had throughout the entire thesis work. The guidance and feedback you offered have helped us finalize the project and reach our goals.
Furthermore, we would like to thank the supervisor from the case company for the openness and willingness to work with us. The fact that you granted us with all the necessary information, and your active involvement throughout the project has boosted its quality even more. We also want to thank the other stakeholders from the supplying companies that took part and provided feedback on every step of the project, guiding us through a very close collaboration.
Lastly, we want to present our gratitude towards our families that have supported us through the last two years of studies, even more during this project. The support you have showed has helped us into overcoming all the challenges we have encountered along the way.
5
Table of Content
1. Introduction ... 10
1.1BACKGROUND ... 10
1.2CASE COMPANY BACKGROUND ... 10
1.3PROBLEM STATEMENT ... 11 1.4PURPOSE AND AIM ... 11 1.5RESEARCH QUESTIONS ... 11 1.6DELIMITATIONS... 12 1.7OUTLINE ... 12
2. Methodology ... 13
2.1RESEARCH METHODOLOGY:ACASE STUDY ... 13
2.2DATA COLLECTION ... 13
2.2.1 Qualitative Data ... 14
2.2.2 Quantitative Data ... 15
2.3DATA ANALYSIS ... 16
2.3.1 Feasibility Analysis ... 17
2.3.2 Risk Assessment Analysis ... 17
2.3.3 Cost Analysis ... 18
2.3.4 Discrete Event Simulation (DES) ... 20
2.4CONCEPT DEVELOPMENT ... 22 2.4.1 Concept Generation ... 22 2.4.2 Concept Screening ... 22 2.4.3 Concept Scoring ... 23 2.4.4 Concept Testing ... 23 2.5RESEARCH QUALITY ... 24
3. Theoretical Framework ... 25
3.1MAINTENANCE ... 25 3.1.1 Corrective Maintenance ... 26 3.1.2 Preventive Maintenance ... 26 3.1.3 Predictive Maintenance ... 266
3.2PRODUCTION LAYOUT ... 27
3.3INDUSTRY 4.0PRINCIPLES ... 27
3.4INDUSTRIAL INTERNET OF THINGS (IIOT) ... 28
3.5CYBER PHYSICAL SYSTEMS (CPS) ... 29
3.6INDUSTRIAL ROBOTS AND END-EFFECTORS ... 29
3.7WIRELESS TECHNOLOGIES ... 30
3.8LEAN MANUFACTURING AND TOTAL PRODUCTIVE MAINTENANCE ... 30
4. Current State ... 34
4.1SUPPLY CHAIN IMPLICATIONS ... 34
4.2PRODUCTION LAYOUT AT COMPANY A ... 35
4.3CURRENT MAINTENANCE POLICY ... 36
5. Concept Development ... 37
5.1IDENTIFY CUSTOMER NEEDS ... 37
5.2ESTABLISHMENT OF TARGET SPECIFICATIONS ... 38
5.3CONCEPT GENERATION ... 40
5.3.1 Benchmarking ... 40
5.3.2 External Search ... 41
5.3.3 Concept Variants Generation ... 41
6. Concept Selection ... 45
6.1WIRELESS PROTOCOLS COMPARISON ... 45
6.2CONCEPT SCREENING ... 46
6.3CONCEPT SCORING ... 48
7. Concept Testing ... 49
7.1IMPROVEMENT AND COMBINATION OF SELECTED CONCEPTS ... 49
7.2FEASIBILITY FRAMEWORK FOR AUTOMATED MAINTENANCE SOLUTIONS ... 50
7.2.1 Discrete Event Simulation (DES) ... 51
7.2.2 Risk Assessment ... 54
7.2.3 Cost Analysis ... 58
7
8.1CONCEPT DEVELOPMENT FRAMEWORK FOR INDUSTRIAL SOLUTIONS IN SMES . 62
8.2FEASIBILITY OF AUTOMATED SOLUTIONS IN SMES ... 63
8.3SUPPLY CHAIN INFLUENCE OVER DECISION-MAKING IN SMES ... 65
9. Conclusions ... 67
10. Future work ... 68
11. References ... 69
12. Appendices ... 74
APPENDIX 1-SEMI-STRUCTURED INTERVIEW QUESTIONS AT COMPANY A ... 74
APPENDIX 2-SEMI-STRUCTURED INTERVIEW QUESTIONS AT COMPANY C ... 76
APPENDIX 3-MAZAK 6200-IIYCNCTURNING CENTER SPECIFICATIONS ... 78
APPENDIX 4-KUKA KR16ARTICULATED ARM IRSPECIFICATIONS ... 79
8
List of Figures
Figure 1 - Risk Assessment Matrix Design ... 18
Figure 2 - Concept Development Flow Inspired by (Ulrich & Eppinger, 2012) ... 22
Figure 3 - Theoretical Framework Tree ... 25
Figure 4 - The Eight Pillars of Total Productive Maintenance Adopted From (Smith & Hawkins, 2004) ... 31
Figure 5 - The Maintenance Arch (Gateway to Integrated Maintenance) (Smith & Hawkins, 2004) ... 32
Figure 6 - Supply Chain of Preventive Maintenance Technology ... 35
Figure 7 - Production Layout at Company A ... 35
Figure 8 - Identify Customer Needs ... 37
Figure 9 - Establishment of Target Specifications ... 38
Figure 10 - Concept Generation ... 40
Figure 11 - Conceptual Solutions ... 43
Figure 12 - Feasibility Framework for Automated Maintenance Solutions Inspired by (Ssegawa & Muzinda, 2021) ... 50
Figure 13 - No. of Produced Parts DES Results ... 54
Figure 14 - Optimized Maintenance Interval (Smith & Hawkins, 2004) ... 58
Figure 15 - Working with Change in the Future ... 65
List of Tables
Table 1 - Methods for Collecting Data Based on (Saunders et al., 2016; Ulrich & Eppinger, 2012) ... 14Table 2 - Literature Search Process ... 15
Table 3 - Conceptual Combination Among LM Tools and Industry 4.0 Technologies, (Valamede & Akkari, 2020) ... 33
Table 4 - Identified Stakeholders Needs ... 37
Table 5 - Need - Metric Matrix ... 39
Table 6 - Target Specifications for Concept Generation ... 39
Table 7 - Benchmarking ... 40
Table 8 - Wireless Solution Concepts ... 42
Table 9 - Protocols Comparison ... 46
Table 10 - Concept Screening Matrix for the Selected Concepts ... 47
Table 11 - Concept Scoring for the Generated Concepts with the Highest Ranks ... 48
Table 12 - Constant Input Data Table 13 - Variable Input Data ... 52
Table 14 - Scenario 0: Results for Simulation of Material Flow with Manual Maintenance Data Gathering ... 53
Table 15 - Scenario 1: Results for Simulation of Material Flow with Automated Maintenance Data Gathering Assisted by IR ... 53
Table 16 - Scenario 2: Results for Simulation of Material Flow with Advanced Maintenance Data Gathering Technology Without Assistance Required ... 53
Table 17 - Risk Assessment Analysis ... 55
Table 18 - Risk Assessment Matrix Concept Sigma ... 57
Table 19 - Risk Assessment Matrix Concept Zeta ... 57
Table 20 - Cost-Benefit Analysis Framework for the Selected Concepts ... 59
9
Abbreviations
BLE Bluetooth Low Energy
CBA Cost-Benefit Analysis
CMS Cellular Manufacturing System
CNC Computer Numerical Control
CPS Cyber Physical System
DES Discrete Event Simulation
FACTS Factory Analyses in ConcepTual phase using Simulation
HR Human Resources
ICS Industrial Control System
IIoT Industrial Internet of Things
IoT Internet of Things
IR Industrial Robot
ISM Industrial, Scientific, Medical
LM Lean Manufacturing
MTBF Mean Time Between Failures
OEE Overall Equipment Effectiveness
PER Packet Error Rate
ROI Return Of Investment
SCADA Supervisory Control And Data Acquisition
SME Small-to-Medium Enterprise
SoC System on a Chip
TPM Total Productive Maintenance
10
1. Introduction
The first chapter is dedicated for constructing the background in which the research study has been carried in. Here, there is also presented the case company’s background together with the problem statement and the purpose that this study is aiming to fulfill. Based on the purpose, the research questions are presented as well as the delimitations of the study. Lastly, the outline of the entire thesis is presented.
1.1 Background
Nowadays there is an increase in expression of interest from the behalf of small to medium enterprises (SMEs) as technology is gaining ground fast. The SMEs have to adapt to these changing times as otherwise the competitive advantage would be lost as long as other participants in the industry are adopting new techniques and state-of-the-art philosophies. In our times we are looking towards Industry 4.0 phenomena as more and more companies are striving towards adjusting their way of working to be in line with the fourth industrial revolution implications (Villa & Taurino, 2019). The evolution of the concept has in view a sustainable growth as in the actual context there is aimed towards reducing the carbon footprint while developing new technologies involved in this matter (Kamble et al., 2018). In the domain of manufacturing industry there is the most discernable proof of ambition to Industry 4.0 adherence, hence the competition is fierce. While larger enterprises have on one hand the advantage of resources available to invest in projects regarding Industry 4.0 compared to SMEs, on the other hand the SMEs have the advantage of an easier organization.
One important aspect in manufacturing industry is the production efficiency. This efficiency can be highly improved when a manufacturing company is striving towards downtime reduction tending to zero. While the complete elimination of downtime may not be achievable, there is the opportunity of reducing the downtime as much as possible through good maintenance practices. The maintenance habit that proves to be the most in agreement with the Industry 4.0 philosophy is the predictive maintenance which makes use of the available technology, especially sensors in order to create a tailormade efficient maintenance system (Ferreiro et al., 2016). This strategy enables the determination of tools that need to be changed in production before any breakdown is encountered and affects the production system. As there are sensors used in this system created for maintenance, there must be some sort of reader in order to gather the data, analyze and interpret it. This process can be performed manually by an operator, or it can be further on automated for reaching even higher ceilings of performance.
1.2 Case Company Background
The case company, also referred as Company A, is a family-owned small-to-medium company with more than seven decades of activity in the manufacturing industry, located in the geographical region of Småland, Sweden. The company has started with only five employees and some revolving lathes which nowadays have been developed into more than forty lathes accompanied with industrial robots and just as much as an increase in human resource. The production plant at Company A is nowadays highly automated with a continuous improvement mindset for achieving the highest Overall Equipment Effectiveness (OEE) and focusing on delivering high quality products in order to reach customer satisfaction. The company is focused mainly on the automotive industry for which parts are produced in high volume with a high degree of complexity, with the help of CNC turning machines, CNC milling machines, transfer machines and
11
machining centers. Innovation is the key driver at Company A, which can be observed from its high level of digitalization, 100% traceability capability, reduced wastes and high-performance machining process that adhere to Industry 4.0 practices.
1.3 Problem Statement
There are several maintenance practices that can be adopted in many ways, and it can either improve the efficiency in the production system or negatively impact its productivity. The problem with the current maintenance practices in the case study can be depicted by the inclusion of unnecessary human input towards the data collection of this maintenance policy. This is a concern as the workload and implication of time consumption is directly proportional with the number of tools that require manual scanning for the data gathering purpose. There is a potential for improvement by the means of technologically automating this task, which is stated as the ambition to strive for such a future state. Also, there is an increased need in academic research towards developing a guideline for SMEs to systematically develop and analyze conceptual solutions for implementation of Industry 4.0 practices and increase the automation degree in the production plant.
Achieving this conceptual goal and applying it to production further on it is considered to be a double perk, namely, decreasing human workload, as well as the requirement for assistance and increased reliability in data collection. This project is proposed for adding value to the efficiency of production by creating the conceptual interface contributing towards both the automation of maintenance data gathering and generalizing the solution for multi-purpose applicability.
1.4 Purpose and Aim
The purpose is to achieve the most efficient conceptual solution towards the automation of maintenance data gathering in a manufacturing industry context by scanning each tool including the activity to store the data in the local production data cloud. The aim is to focus on developing a conceptual interface which enables the possibility of automation in the presented case, adapting the existing technological resources while at the same time granting perspective angles from alternative concepts. Also, the project is aimed towards the development of an integrated concept solution guideline to be used within SMEs for enabling the generation, selection and analysis for reaching the most suitable solution for automation projects.
1.5 Research Questions
The first research question of this project draws the representation of the case company’s current state, as it represents a standing point for in-depth understanding of its capabilities and limitations. In this manner, the conceptual solutions that are to follow will provide a more understandable way of targeting the implementation. The reader is able to observe a current and future state, possible improvements and trade-offs being more discernable.
[1] How does the current state of the case company look like now concerning the maintenance data gathering from the machinery in focus?
12
The second research question of this project aids into achieving the aim stated earlier and refers to what kind of interface or carrier can be designed in order to read the sensors and manage different tool changers from CNC machines. Two possible case scenarios are going to be studied where the conceptual solutions are analyzed considering the fact that the CNC machines are (a) served by industrial robots or (b) without their assistance. [2] What type of carrier/interface that reads the sensors on request can be designed for managing many different tool changers from CNC machines in the following conditions:
a) Each CNC machine is served by a robot.
b) CNC machines are running without robot assistance.
Lastly, there is a need for identifying and analyzing from which point of view an automated solution has influence on the efficiency of the production process, therefore the last research question refers to the identification of which parameter can indicate improvements brought by the proposed automated solutions:
[3] Which parameter indicates that the automated solutions contribute to a more efficient production process?
1.6 Delimitations
This report focuses on maintenance policies that can be integrated in the Industry 4.0 movement. More in-depth, the phenomenon of the automation and digitalization of these policies will be studied and analyzed, in order to raise to the surface their implications, advantages, disadvantages and capabilities, and to which degree can be adopted by small-to-medium companies. However, the study will focus only on the automatization and digitalization of the maintenance of tools and tool holders that are found in CNC turning centers, where stainless steel parts are manufactured in high volumes.
1.7 Outline
This report consists of nine remaining chapters, where the second (Ch.2) will describe the methodology for implementation and what methods have been used for gathering and analyzing empirical data. The next chapter (Ch.3) focuses on setting the foundation of theoretical aspects that have been researched and used throughout the study, while the fourth chapter (Ch.4) aims at providing a detailed description of the surroundings in which this study as has been carried out and all the data that has been collected during the preliminary phases. The following three chapters (Ch.5, 6 & 7) are the main core of this study, where the purpose is aimed to be achieved and the purpose questions answered. They consist of the work and the processes that have been carried out for generating, selecting, and testing the maintenance policy solutions developed in respect to the needs found in SMEs. Lastly the final two chapters include the discussion of the analyzed findings (Ch.8) and conclusions on the main aspects about the carried work (Ch.9). At the end of the study, there will be made some recommendation for future work in this field.
13
2. Methodology
This chapter is dedicated to present the research methodology chosen to conduct this study as well as presenting the methods used for collecting and analyzing data. Furthermore, the methods are explained from the perspective of how they are used throughout the study.
2.1 Research Methodology: A Case Study
The methodology behind the realization of the thesis project revolves around an inductive research case study in a Swedish Small-to-Medium Enterprise (SME) that will enhance and add value to the future of academic research. The reasoning behind this choice is that within a case study, the ability to test out the latest theories related to maintenance policies, as well as investigate their applicability in the industry is provided (Saunders et al., 2016). The case study strategy has been observed to successfully interrelate a certain phenomenon with the context that it happens in, this fact enabling further development for study in the respective area, the real-life case aiding into collecting empirical data to base the theory on (Yin, 2014). This type of method has been shown to deliver successful results in the past, especially when researching a phenomenon that there is little known about. The logic flow of an inductive research as stated by Saunders et al. (2016) is the most suitable research approach as it is the most appropriate in terms of generalizability. Since the inductive inference is pursuing generalization from more specific to more broad applicability, its theory approach is also following to be generated and built rather than be accepted or denied. Also, the use of data attribute of the inductive approach is also checking the suitability of the project since it is pursuing to explore a phenomenon, identify themes, patterns and create a conceptual framework (Saunders et al., 2016).
2.2 Data Collection
Even though a case study is a commonly used method among researchers, its reliability may come in question, as one case study may not show enough evidence of a general applicability within the researched field. For that, in the following subchapters the methods for collecting qualitative and quantitative data are described, which are also scrutinized in Table 1. For the qualitative data collection, three methods are selected, which consist of semi-structured interviews at the case company, focus groups with the stakeholders and observations. On the other hand, quantitative data will be collected through a benchmarking of available products or solutions that are already on the market and a literature search, where numerous scientific papers will be analyzed, with the purpose of identifying the existing understand and knowledge in the studied field, and also for identifying the existing gaps in research.
In order to avoid uncertainty, biasedness, and lack of cooperation from the stakeholders, that represents the main source for collecting data, several methods have been used for assuring that this study is following an ethical conduct. As Saunders et al., (2016) mention, a clear overview of the required data has been presented to the stakeholders early in the project as well as the purpose behind collecting the required data. In the same manner, sensitive data has not been disclosed or shared and confidentiality has been assured throughout the entire study. Data that is sensitive or confidential has been presented with a level of generalization that makes identification impossible (Saunders et al., 2016).
14
Table 1 - Methods for Collecting Data Based on (Saunders et al., 2016; Ulrich & Eppinger, 2012)
Data Type Method Collected Data
Q ua lita tiv e Da ta
Observations • knowledge gain by observing the ongoing phenomenon • increased understanding of the implications of the
phenomenon
Semi-structured Interviews
• specific key information about the maintenance process • human resource’s thoughts and ideas about the process • clear understanding for the need for improvement • technical data about the process
• implications and limitations of the current tool maintenance process
Focus Groups • stakeholders’ feedback about the generated concepts • generation of new ideas and perspective
Q ua ntit a tiv e Da ta Literature
Search • overview of the current information available
• perspective from which the topic has been researched before Benchmarking • Available solution already on the market
• Assessing significant trade-off potential
2.2.1 Qualitative Data
In order to get a detailed overview of the phenomena that will be analyzed throughout this project there is a need for qualitative data methods such as observations, semi-structured interviews and focus groups. Semi-semi-structured interviews are chosen because the interviewer has the possibility to start from specific key questions with a clear purpose behind but can still shift to different areas to get more details if found necessary (Saunders et al., 2016). Semi-structured interviews will be held with the technical manager from the case company and other stakeholders, as they represent a valuable source of information required to get a detailed view over the studied phenomenon (See Appendix 1 & Appendix 2). An important stakeholder interviewed consists of a group of representatives from the companies involved in this project such as the current maintenance policy supplier. Both interviews are constructed in such a way that enables the interviewers the ability to adapt certain questions based on the flow of the interview and eliminate or discuss further others in the same manner. Another method for gathering qualitative data that will be used in this thesis project refers to observations. Observation research is a qualitative method through which the researcher gains knowledge by observing the ongoing phenomenon. The decision upon using this kind of method comes from the need for a close investigation that will be required for fully understanding the surroundings, limitations and implications of the studied phenomenon, in other words, to understand and to answer the “how?” in the research questions (Saunders et al., 2016). This type of method will also help in ensuring validity when observing the process in natural conditions at the case company.
An additional type of interview that will be used for the collection of data, consists of creating a focus group. This method is particularly helpful when there is a need for feedback during the evaluation process (Phillips & Stawarski, 2013). Focus groups mainly consist of a small group, usually stakeholders, that are involved in an open discussion where each individual makes its own input and gives feedback upon the issue that is presented. This kind of method helps in generating relevant data where multiple
15
aspects of the discussed topic are weighted and assessed, which will bring value and improve the concept generation process.
2.2.2 Quantitative Data
The reasoning behind using literature search as one of the methods for this thesis is that, through a literature search on a certain topic, the authors are able to get an overview of the current information available related to the main field of study. This method represents an important step for us to identify the current understanding and knowledge about maintenance policies that are in use in the context of Industry 4.0 (Pan, 2016). Table 2 depicts the literature search results summary including the combinations of keywords used, filters and the number of hits ensued.
Table 2 - Literature Search Process Search
No.
Database Search Combination Filters Used No. of
hits 1 Science Direct Predictive Maintenance AND Factory
Automation AND Industry 4.0 Year, Subject Area 246 2 Science Direct Predictive Maintenance AND Wireless
Data Transfer AND CNC Year 30
3 Science Direct
Internet of things AND Industrial wireless sensor networks OR Wireless protocols AND Predictive maintenance
Year, Article Type,
Subject Area 955
4 Science Direct
Predictive Maintenance AND SMEs AND Industrial robots AND machine monitoring
Year, Article Type,
Subject Area 43
5 Emerald Insight
Industrial Internet of Things AND Industrial Wireless Sensor Network AND Predictive Maintenance
Year, Journal, Type 53
6 Emerald Insight Industry 4.0 AND Machine Monitoring
AND Wireless Protocols Year, Article Type 93
7 SpringerLink SME AND Factory Automation AND Machine Monitoring
Year, Content Type, Discipline and Subdiscipline, Language
41
8 SpringerLink Machine Monitoring AND Wireless sensor networks
Year, Content Type, Discipline and Subdiscipline, Language
225
9 SpringerLink
Predictive Maintenance AND Tool Condition Monitoring AND Wireless sensor networks
Year, Content Type, Discipline and Subdiscipline, Language
11
10 SpringerLink Closed Loop Control Application AND Factory Automation
Year, Content Type, Discipline and Subdiscipline, Language
36
11 ProQuest Smart Tool Condition Monitoring AND
CNC Machine Year, Article Type 9
12 ProQuest Preventive Maintenance AND Tool
Condition Monitoring Year, Article Type 28
13 ProQuest Lean Manufacturing AND Total
Productive Maintenance Year, Article Type 54
14 ProQuest Lean Manufacturing AND Industry 4.0 Year, Article Type 78
16
Through this method, relevant publications will be found, analyzed, and discussed, in order to identify to which extent this topic has been researched before, but also to identify the gaps in knowledge and academic research (Pan, 2016). For this purpose, four databases are used for searching relevant literature in the field of study. The databases used are the following: “Science Direct”, “Emerald Insight”, “SpringerLink” and “ProQuest”.
The keywords used revolve around the topic of automation engineering, more in depth “factory automation” with an emphasis on maintenance policies, such as “preventive maintenance” and “predictive maintenance”. Other keywords found necessary for reaching relevant literature are “Tool Condition Monitoring”, “Wireless Sensor Networks”, “Closed Loop Control Application”, “Industrial Wireless Sensor Network” and “Wireless Protocols”. At the same time, it is significantly important for the study conducted to identify the available knowledge and research that has been carried out regarding “SMEs” in the context of “Industry 4.0” and “Internet of Things”, as well as their connection to “Lean Manufacturing” and “Total Productive Maintenance”. Also, to be mentioned that the keywords are used in relation to the database search operators “AND” and “OR” for limiting the search results to more specific searching needs. Since the number of hits was considered to be too high, the available literature was narrowed down. The principles behind narrowing down literature refer to selecting the literature which has a relevant similarity in keywords associated with this study’s purpose. Then, the next step involved scanning the abstract of each remaining literature paper and based on that, a decision was made if the paper might or might not add value to the present case study.
Correspondingly, the benchmarking method is used in this context to relate the relative performance of the developed conceptual solutions in this project to the existing solutions available on the market (Ulrich & Eppinger, 2012). This method delivers both a more in-depth view over what technologies are already offered by solution providers in this matter and the developed conceptual solutions potential success. In the current circumstances, since the developed solution is not a physical product itself but a structured interface of multiple hardware and software involvement, the core of the solution (wireless module) will be benchmarked. This decision is based on the availability of metrics developed for clearly discerning between the available wireless technologies offered by solution providers at the current time.
The solution core represents the starting point of the benchmarking sequence, the rest of the solution being dependent on it and not having any benchmarking impact since it may differ from concept to concept. The rest of the solution besides the wireless module is considered to be nonstandard, hence eliminated from the benchmarking sequence for reliability reasons. A competitive benchmarking chart will be developed for comparative marking, where the importance rating is an essential factor in differentiating between competitive solutions and assessing the significant trade-off potentials (Ulrich & Eppinger, 2012).
2.3 Data Analysis
In order to fulfil the purpose of this thesis project and answer the research questions mentioned earlier, the data collected both qualitatively and quantitatively, will contribute to further generation of concepts that revolve around automating the process of gathering data for maintenance of tools used in CNC machines at the case company.
17 2.3.1 Feasibility Analysis
The feasibility analysis process is provided within the project as a decision tool in order to aid the company which is pursuing the implementation of a conceptual solution to assess its viability. Ssegawa & Muzinda (2021) are providing a feasibility analysis process which pursues this objective of viability within different areas of feasibility. As different alternatives of solution implementation are emerging, each one is analyzed mainly throughout technical, economical, and socio-environmental areas, although, the operational and schedule areas of feasibility are also considered when performing the analyses. The slightly adapted model suggested for feasibility analysis in order to fit the current case scenario is based on the iterative process proposed by Ssegawa & Muzinda (2021) and it is pursuing the following methodological areas:
1. Technical impact – the technical area is represented by the perspective of solution building and implementation viability within the existing production layout. This impact shall be further analyzed through the Discrete Event Simulation (DES) outcomes.
2. Economic impact – the economic area is seeking the degree of viability from an investment angle where the main method used for assessment is the Cost-Benefit Analysis (CBA).
3. Socio-environmental impact – this area is providing an overview mainly upon the human and the environmental factors that have a tangency with the respective solution in focus through the filter of a risk assessment.
4. Operational impact – as the relationships between the companies involved in the supply chain of maintenance policy in focus are considered, the operational impact is provided as determining the position of each stakeholder in this operational chain. The operational impact is assessed through focus groups. 5. Schedule impact – this area is considered from a time-perspective as the
technology is in a continuous accelerated development and the feasibility of a project can change if its implementation requires an overextended period. 2.3.2 Risk Assessment Analysis
In parallel with the process of concept generation, a risk assessment analysis matrix will be developed for closely monitoring and detecting the potential risk factors that may affect the expected outcome. This kind of tool helps with identifying the preferred outcomes that are prioritized and assess which kind of risks there should be taken into consideration (Olson & Wu, 2010). A matrix design similar to the one proposed in Figure 1 provides an easy overview of the risks that the concepts may face, and, in this way, it will be easier to keep track and minimize them.
The matrix design consists of the “likelihood” of the risk occurring, and the tolerability of the “consequence” that occurs due to the risk. There are three green cells which refer to the risks that are improbable to occur, and if there is a possibility for occurrence, they are either acceptable or tolerable. The yellow middle area indicates the risk with a more severe effect than those in the green area, and they require more attention and monitoring, in such way that the risk will be minimized as much as possible. Lastly, the red area refers to those risks that are most certainly probable, but not acceptable or tolerable.
18
Figure 1 - Risk Assessment Matrix Design
2.3.3 Cost Analysis
Developing maintenance solutions or improving current maintenance techniques require extensive cost analyses as maintenance in a manufacturing plant holds great responsibility for delivering quality products and maintaining overall productivity high (Tevfik, 1996). The maintenance of equipment in a production plant could either save the company time and money, or it could potentially damage equipment, increase throughput time, increase costs, and lower productivity (Tevfik, 1996). A cost-benefit analysis (CBA) influences decision making from the early stages of system design, as it represents a key evaluator of alternatives processing and trade-offs. Cost-benefit analysis is a type of analysis developed for decision making and evaluation during project development phase (Tevfik, 1996). The methodology behind CBA refers to the identification of all the potential gains and losses and the conversion of those into monetary values. Based on this principle of assigning monetary value for each gain (benefit) and loss (cost) decision can be made if the project is feasible or not for that particular case (Boardman et al., 2001). As the authors mention, the analysis of the relationship between costs and benefits is more than a regular analysis of costs, as it brings in addition to other types of analyzation of costs, a social utility index that is addressed when comparing gains with losses, represented as follows:
NSB= B-C Where:
NSB= Net Social Value B= Social Benefits C=Social Costs
The general purpose of such analysis is to aid decision-making for the most efficient allocation of resources within an organization to promote a sustainable socio-economic development. For that, Boardman et al. (2001) create a systematic approach for CBA and explain all the major steps that need to be taken for achieving the desired results.
19 1. Specification of the set of alternatives.
This first step represents the foundation of the cost-benefit analysis, and it is required to present all the alternatives of the project. The main challenge that rises during this preliminary step relates to the multitude of alternatives that a project may have, which can create difficulties when deciding which alternatives should be included or not. Boardman et al. (2001) mentioned that when such an issue arises, it is best to focus on those alternatives that are feasible within the area of the project and based on the capabilities and limitations available in that particular environment. It does not add value to analyze alternatives that require a completely different set of resources which are not aligned to the environmental and socio-economic aspects in which the project will take place.
2. Decision on which benefits and costs count.
The second step reflects upon from which perspective the benefits and costs should be counted. That owes to the fact that a project has different impacts on a group of individuals, organizations, or environment, depending on which perspective the analysis is done. Therefore, it is of great importance to determine from which point of view the costs and benefits will be accounted, facilitating in this way a clear path towards which future decisions will be made.
3. Classification of inputs and outputs with measurement indicators.
Next, it is required to catalogue the impacts, other way put as inputs or required resources, of the alternatives in terms of benefits and costs. In addition to this step, it is mentioned by the authors, that the assignment of measurement indicators is also needed, for a clear understanding of what is it weighted and counted in both terms of benefits and costs (Boardman et al., 2001).
4. Prediction of the impacts over the lifetime of the project.
As mentioned by the authors (Boardman et al., 2001), a project has impacts that extend throughout a timeframe and therefore it is required to quantify these impacts for each alternative dependent on the lifetime of the project. One challenge that rises during this stage in particular, relates to the difficulty, although necessary, of prediction the impacts of the alternatives. However, there are methods for data collection that ease up the process of prediction based on other previous studies within the researched topic related to the case phenomena.
5. Monetization of costs and benefits.
The monetization process refers as a method for addressing a monetary value on each benefit and cost that the alternatives imply. In this process, parameters as time saved, material saved, labor hours saved or any other indicator that it is measured in other units, are given a monetary value. The purpose of this step is to translate all possible gains and losses into capital worth, making it possible for these values to be analyzed under the structure of CBA.
20
6. Discounting the costs of benefits in relation to time.
This step is described by the authors for a project that has costs and benefits that fluctuate over an extended period (Boardman et al., 2001). For that matter, future benefits and costs are discounted relatively to their present values (PV). The present value in terms of benefits at costs is calculated as follows:
PV(B)=∑ 𝐵𝑡 (1+𝑠)𝑡 𝑛 𝑡=0 PV(C)= ∑ 𝐶𝑡 (1+𝑠)𝑡 𝑛 𝑡=0 Where:
PV(B)=present value of the benefits PV(C)=present value of the costs t=year in which benefit, or cost appear n= project lifetime (years)
Bt= benefits in year t
Ct= costs in year t
7. Computation of the Net Present Value for each alternative.
The net present value (NPV) is calculated as the difference between the present value of the benefits and the present value of the costs. This step is important for ruling out alternatives in favor of others, and the most important rule to follow as stated by the authors is to “adopt the project if its NPV is positive” (Boardman et al., 2001). If the NPV is positive, it means that the value of benefits is higher than the value of cost, therefore, the gain is greater that the losses. The net present value (NPV) is calculated as follows:
NPV= PV(B)-PV(C) If NPV>0, it means that PV(B)>PV(C)
8. Make recommendation based on NPV.
As a general approach, Boardman et al. (2001) recommend going for the alternative with the highest positive value of NPV, however there might be cases where all the alternatives have negative values. This means that the alternatives analyzed do not necessarily bring more gains to the scope of the project, and for that, the recommendation made from the authors is to either remain with the status quo or develop other alternatives for the project, if possible.
2.3.4 Discrete Event Simulation (DES)
The Discrete Event Simulation (DES) in this case is represented by a digital animated model of the production layout with real case scenario machine inputs from a time-based process perspective, while also considering the possibility of disturbances
21
occurrence. This method is used as a reinforcement in the feasibility analysis for achieving introspection perception of the entire manufacturing system and observe the possible efficiency gains from comparison of different models proposed (Allen, 2011). The gains are specifically traced on the maintenance data gathering operation and the disturbances caused by this manual operational process versus what the simulation results are showing in a case where the data gathering process is performed automatically. Allen (2011) is representing the framework towards the development of a DES in a five step scenario with the following structure:
Phase 1: Defining – this phase is necessary, being the core planning of the simulation in order to pursue a clear scope and obtain a “big picture” of how the simulation will be designed.
Phase 2: Input analysis – this phase step represents the collection of relevant data for the simulation analysis purpose, furthermore, associating the data into the software’s proper fields.
Phase 3: Simulation or calculation – this phase is the actual experimental phase where the calculations needed for the respective simulation model are done by the software and results are shown to be analyzed. Further iterations can be performed in this step, in order to increase the reliability and validity of the model based on the real flow of the manufacturing system and the user’s expectations.
Phase 4: Output analysis – this step is used in the case company scenario to analyze different conceptual models of maintenance data gathering in order to offer perspective over the efficiency gains. Multiple models of the same manufacturing layout are resulted from the simulations and the outputs are further analyzed.
Phase 5: Decision support – this step is the actual graphical representation of the result obtained from the simulation analysis performed from a feasibility perspective. In this step, the results are depicted for a clearer image over the difference gap in results between the different simulation models and what are the take-aways from this perspective.
As the Phase 4 of DES involves the analysis of output fields, the relevant output fields for this case are the following: throughput, Work In Process (WIP), lead time and produced parts. These indicators are closely linked together since they are common points of interest in simulation, line balancing and flow optimization topics. Nahavandi (2009) is describing throughput and lead time as typical performance measures and it is arguing how their interdependence is influencing one another. In the simulation context, the throughput is considered as the “number of units produced in a time period”, likewise, the lead time is represented as the “average time a unit spends in the system”. Based on these criteria analyzed as outcomes, differentiation in performance measurement can be observed. Also, Curry & Feldman (2011) are defining WIP as “the number of jobs within a system that are either undergoing processing or waiting in a queue for processing”. Ultimately, the “Produced parts” outcome refers to the jobs that are successfully finished considering the time span of the simulation.
22
2.4 Concept Development
The concept development process consists of mainly seven stages, following the framework developed by Ulrich & Eppinger (2012) where several activities are performed in a chronological order, in such way that the outcome of the process result a well-documented and analyzed concept that will satisfy the customer needs. The concept development process starts by identifying the customer needs, mainly from lead users and stakeholders, followed by establishing the target specifications. Once the target specifications are set, the process of concept generation and selection are done. The concepts that manage to pass these two stages will be subjected to testing and when even this stage is passed, the final specification will be set for completing the fully process of concept development, as shown in Figure 2.
Figure 2 - Concept Development Flow Inspired by (Ulrich & Eppinger, 2012)
2.4.1 Concept Generation
Concept generation can be defined as a process that begins from a customer need, in this case the company’s need, from which by generating and analyzing of a set of alternatives, one or more conceptual solutions are chosen for further development and implementation (Taura & Nagai, 2013). The advantages of pursuing this kind of procedure in the thesis project are related to the ability for generating new perspectives and take into consideration new possibilities that may have not been used before, in such way that fills the gap in knowledge and creates a new bridge between academic research and the manufacturing industry.
In addition to the process of generating concepts, the methods of “Screening” and “Scoring” will be used for analyzing and identifying the most suitable concepts for achieving the purpose and aim, and lastly the concept with the highest score will go through the last phase of concept generation, which refers to testing (Ulrich & Eppinger, 2012).
2.4.2 Concept Screening
Concept screening is the first of the two-stage process of concept selection, after the concepts have been developed in the previous stage. This process of screening the generated concepts represents an evaluation of them in relation to one reference concept and using a matrix called, screening matrix (Ulrich & Eppinger, 2012). During this phase, the concepts are ranked and analyzed using the same criteria, in order to avoid unbiased selection. The criteria are based on the customer needs that have been identified in the previous steps, in this case, the case company, as well as the needs of its two supplier companies that participate in the concept development process.
23
After the selection criteria is set, the next step is to choose a reference concept, that usually represents a standard that it is used already in the industry or a direct concept that the team is already familiar with. Once a reference concept is chosen, the next steps consist of rating with “+” meaning “better than”, “-” meaning “worse than” and “0” which means ”same as” the reference concept (Ulrich & Eppinger, 2012). The rating and raking of the concepts will enable the team to analyze and choose which concept to pursue further into development, which could be combined and improved. Once this phase is complete, the result after concept screening consists of fewer improved concepts that can be taken into the next stage.
2.4.3 Concept Scoring
Concept Scoring is the seconds stage of concept selection, and it represents a more in-depth analysis of concepts that have been selected from the previous stage. The process is similar in steps with concept screening, however, more in-depth criteria it is used, such as added a rating index, as well as a weighted importance in the decision making for whether to pursue a concept or not (Ulrich & Eppinger, 2012). Even though, the steps are similar in both screening and scoring, in the last stage, the concepts are selected based on a total score that is calculated as follows:
𝑆𝑗 = ∑ 𝑟𝑖𝑗𝑤𝑖
𝑛
𝑖=1
Where:
rij= raw rating of concept j for the ith criterion
wi= weighting for ith criterion
n= number of criteria Sj= total score for concept j
During this phase of concept scoring, concepts are subjected to combinations and modifications in such way that more improvements are added than in the previous phase. Ulrich & Eppinger (2012) emphasize that throughout this phase, the development team is much more prone to realization and identification of better solutions, that could not have been foreseen during the earlier stages of concept generation, where certain features of the concepts are analyzed more in-depth.
2.4.4 Concept Testing
Concept testing is also a part of the concept selection process, where from the possible alternatives generated and ranked previously, only a narrow selection are taken for further investigations. This phase is important in particular, because during this process information from potential customers in the target market is gathered and also the sales potential is examined. The reasoning behind concept testing taking place after the screening and scoring phases owes to the fact that usually the development team cannot practically test out more than a few concepts with the potential customers. Therefore, the development team must limit the set of concepts and pursue with a fewer range from this stage on (Ulrich & Eppinger, 2012). A very important aspect is taken into consideration during this stage, and that refers to any costs that the concepts require for development and implementation. The costs represent an extremely valuable factor for customer when deciding on whether to choose the developed concept or not and can even represent the definitive reason for choosing one concept over the other. That is why the relevant types of costs are carefully analyzed during this phase.
24
2.5 Research Quality
According to Saunders et al. (2016), the research quality is checked through the notions of reliability and validity. As the current project is also liable to own an assessment of quality, the notions mentioned are reasoned through the following justifications. The reliability refers to the research’s replication and consistency. This component is marked throughout the study by its main goal of providing a guideline towards reaching the best possible solutions when designing concepts in a certain field of study. In the current circumstances, the guideline is focusing on the case of developing conceptual solutions for maintenance data gathering. Furthermore, from the perspective of validity, the notion is split into external and internal validity (Saunders et al., 2016). The external validity which is concerning if the performed study is able to be generalized is also applicable in this case with the main justification point being common with the one from the reliability perspective. The study generalization is kept in focus throughout the project development since one of its aims is to provide applicability to a wider area of interest within the manufacturing industry which tends to factory automation. For this fact, the processes performed within the project are thoroughly explained, also offering the correspondent framework that can be followed for similar projects. Moreover, the internal validity refers to the ability to validate the progressive outcomes throughout the project (Saunders et al., 2016).
In the current study, since there is available a constant collaboration between the authors, the case company and their suppliers, constant validation feedback is required from the involved companies before progressing in a certain direction with the study. This fact enables a three-way validity system that is considered since the companies possess different responsibilities regarding the involvement in the solutions provided within this study. Also, contributing to the same context, the study is using continuous literature research and the literature findings used are based on reviewed valid academical theories.
25
3. Theoretical Framework
To achieve the goal of this project, which was presented above, there must be mentioned the main area of expertise to be followed both for the case study resolution and research driven input, which are scrutinized in the theoretical framework tree found in Figure 3. While the main area of study is considered to be within the Industry 4.0, the focal subjects revolve around maintenance practices that adhere to this phenomenon. As an attempt towards deepening the knowledge about what type of maintenance policy can be integrated in Industry 4.0, other branches have been also studied, such as Cyber Physical Systems (CPS), Industrial Internet of Things (IIoT), Industrial Robots (IR) and end-effectors and wireless technologies. The CPS is considered as an integrated system containing IIoT elements which aids into achieving an integrated automated system where more efficient maintenance policies are enabled through the use of wireless technologies. The study goes more in depth towards linking the Lean Manufacturing (LM) and Total Productive Maintenance (TPM) in the context of Industry 4.0, facilitating the implementation of preventive and predictive maintenance in robotic cell manufacturing layouts where production is aided by industrial robot support.
Figure 3 - Theoretical Framework Tree
3.1 Maintenance
The concern of efficiency in production systems is known to be highlighted directly by the need of avoiding breakdowns, hence decreasing machine downtime (Mobley, 2002). The notion of Mean Time Between Failure (MTBF) is considered as relevant within the maintenance practices involved into the study as this is representing a basis of computation towards the proper scheduling of the predictive maintenance system. The MTBF is defined according to Lienig & Bruemmer (2017) as a calculated reliability parameter representing the mean lifetime of the component in focus within a repairable system.
26
This concept has been discussed along the years by scholars through the frame of proving the link between the expected result and maintenance practices that can be used by a company in an industrial environment. In order to understand the need for proper maintenance techniques in this case study considering the context of Industry 4.0 and high automatization ambition, the main maintenance types are reviewed from a theoretical perspective.
3.1.1 Corrective Maintenance
The corrective maintenance practice is represented by reaching the maintenance scenario in a case where the tool that require maintenance has failure symptoms or has already broken down (Wang et al., 2014). This maintenance policy is tending to be avoided as cost-wise, there is recommended to prevent total failure of a tool rather than utilize it until it is unusable. The corrective maintenance is also prone to higher downtime while at the same time is associated with loss in quality, since the tools necessary to perform manufacturing operations in the current case are worn-out in the critical range of tool’s life state. There are two types of corrective maintenance to be mentioned, immediate and deferred, according to the maintenance terminology standard EN13306:2017 (2017). This maintenance type, known to present lots of downsides, is unavoidable if no other maintenance approaches are introduced. For instance, the first layer of protection against total tool failure is the preventive maintenance phenomenon where scheduling of maintenance is introduced.
3.1.2 Preventive Maintenance
Before a fault in the operation tools occur, the symptoms of failure can be detected through check-ups on a scheduled basis. One of the most common definitions of preventive maintenance refers to this type of maintenance practice as “actions performed on a time or machine-run-based schedule that detect, preclude or mitigate degradation of a component or system with the aim of sustaining or extending its useful life through controlling degradation to an acceptable level” (Sullivan et al., 2010). The preventive maintenance can be performed on a pre-determined schedule when a certain asset’s measured levels are determined to be unsafe or on limit (Sullivan et al., 2010). The preventive maintenance is a good approach when pursuing the reduction of machine downtime and improving OEE, as its main benefit is observed in extending the equipment’s life and minimizing failures, which also helps in saving capital and reducing costs (Sullivan et al., 2010). This measure is also taken into consideration when there is a known cause of failure and it is considered that the check-ups and maintenance can be done periodically (Duffuaa & Raouf, 2015). The preventive maintenance also has a characteristic of being able to be time based when the cause of failure is known to be the wear of a tool for instance.
3.1.3 Predictive Maintenance
With the expansion of knowledge and research towards the Industry 4.0 (Boyes et al., 2018) phenomena, the most suitable type of maintenance for an enterprise which is striving for high automation in production is the predictive maintenance. The care for maintenance practices is a key enabler of long-term vision over reducing operations costs, maintenance proven to represent between 15% and 60% of these (Zonta et al., 2020). Taking that in consideration, predictive maintenance can be defined in general as “the regular monitoring of the actual mechanical condition, operating efficiency, and other indicators of the operating condition of machine-trains and process system will
27
provide the data required to ensure the maximum interval between repairs and minimize the number and cost of unscheduled outages created by machine-train failures” (Mobley, 2002). The difference between preventive and predictive maintenance is that, predictive maintenance does not rely only on average statistics about the life cycles’ machines, but also other relevant data is collected and that refer to the mechanical condition of the equipment, system capability and other indicators that reflect better the actual MTBF of the equipment (Mobley, 2002).
Now, in the context of Industry 4.0, the potential of performance in terms of machine availability is even higher since there are dedicated hardware and software applicable for this task. Sensors and maintenance monitoring software solutions are available for a multitude of cases where there is the aspiration towards quality, cost, and production efficiency. Predictive maintenance solutions are in constant development and as new technologies or improvements to older versions arise, new solutions develop and prove to be more and more suitable to the manufacturer’s needs.
3.2 Production Layout
Nowadays, a commonly found practice in the manufacturing industry refers to the implementation of a robotic cell layout that can enable companies to gain competitive advantage on the market (Zhang & Fang, 2017). Robotic cell layouts in a production plant are part of a newly studied design layout type that is called Cellular Manufacturing Systems (CMS), that imply using several articulated-arm industrial robots for a multitude of operations performed during manufacturing. Since there is a constant need for increasing the automation level in the production plants, the use of articulated arm industrial robots for performing a wide range of applications that would normally be done by human operations, is present in more and more companies, from small to large (Ji & Wang, 2019). Designing the production layout as a robotic cellular manufacturing system consists of many decision-making processes, as it can either have huge benefits or major drawbacks that result in higher costs, breakdowns, and production failures. Since cellular manufacturing systems are designed in groups that share the same applications, meaning that machines and industrial robots are placed together to collaboratively work, the setup times is reduced, as well as work-in-process, throughput time and material handling costs (Suemitsu et al., 2016). Also, a robotic cellular manufacturing layout is designed as cluster of shared applications and processes that can act as a smooth-running sole system, if implemented and optimized accordingly. When non-adding values are reduced or even eliminated in a CMS, it creates the optimal environment for lean manufacturing implementation, enabling companies to minimize costs and enhance productivity and quality of products (Pattanaik & Sharma, 2009).
3.3 Industry 4.0 Principles
The fourth industrial revolution is emerging as a phenomenon towards digitalization of traditional manufacturing and SMEs are driven towards adopting the new practices in order to remain competitive on a volatile market. The concept of Industry 4.0 was founded on the integration of the so-called Cyber Physical Systems (CPS) that enable fusion between information and communication technologies, for increasing traceability of goods, machines, products, and people (Matt et al., 2020). The competitiveness factor is distributed between the SME’s ability to decrease the lead time, increase flexibility and show a certain degree of customization towards the customers’ requests as there is stated by Matt et al. (2020).
28
The authors also mention the process of transitioning to Industry 4.0 as was observed to be more and more necessary for high wage countries in order to keep the competitive advantage. However, one of the highlights in specialized literature regarding the Industry 4.0 is related to the struggles of companies, especially SMEs, to properly adhere to smart manufacturing (SM) methods (Masood & Sonntag, 2020). The SMEs are also more prone to adopting Industry 4.0 practices in a slower pace than larger companies, investing more time into processes required to upgrade the technological factors towards SM (Matt et al., 2020). Masood & Sonntag (2020) also mention that in SMEs there are more financial constraints and resources limitations than in the case for multinational companies, therefore more challenges arise in areas of production, manufacturing, logistics and on managerial perspective. Orzes et al. (2019) highlight that some of the challenges most commonly found in SMEs when it comes to Industry 4.0 refer to limitations in knowledge and technology awareness and also due to the unclear return of investment of adopting Industry 4.0 practices. The lack of knowledge and understanding of the market potential create barriers for SMEs to adhere to Industry 4.0, as it perceived as a costly and prolonged process (Orzes et al., 2019). Moreover, other challenges that are empathized by Orzes et al. (2019), Masood & Sonntag (2020) and Matt et al. (2020) refer to the human labor resistance towards such a major change in the way of working and the lack of top management vision towards Industry 4.0. The capital required for new equipment acquisition and training employees to work in a more technological advanced environment are not always reasoned enough for the top management, as there is a lack of a proper toolset and guideline towards integration of Industry 4.0 practices in SMEs (Orzes et al., 2019).
As mentioned earlier, Industry 4.0 is based on the concepts of CPS, Industrial Internet of Things, and wireless technologies that enable a better and more efficient communication and knowledge sharing across the entire manufacturing plant. One of the Industry 4.0 fundamental characteristics relates to the interoperability between assets involved in the manufacturing system, fact on which the current case study is focusing upon. This type of data exchange combined with visualization capabilities towards automatization of a lean process, provides an SME the advantage of being one step closer to fully Industry 4.0 adaptation.
3.4 Industrial Internet of Things (IIoT)
Industrial Internet of Things (IIoT) is defined as “the use of Internet of Things technologies in manufacturing” according to Boyes et al. (2018) as it enables machine-to-machine collaboration, machine learning and big data usage in the industry for more efficient operations. The industrial applications included in IIoT refer to interconnection between information and operational technology that aid companies towards automation for better and more visible monitoring and controlling processes of operations throughout the supply chain (Boyes et al., 2018). More in depth, IIoT consists of a system which integrates Internet of Things principles, such as CPS, Big Data, information technology, Industrial Control Systems (ICS), and Supervisory Control and Data Acquisition (SCADA) in industrial systems (Arnold et al., 2016).The main promoter for the current performance potential of this type of condition-based maintenance is the uprising technology comprised by the IIoT. The industrial applications of sensors combined with cloud-based services and a holistic view over a proper maintenance philosophy is providing companies nowadays novel technological possibilities towards solving production challenges or optimizing existing layouts. One of the main challenges with IIoT at this moment is that it derives from Internet of Things (IoT) consumer-oriented appliances and some of the adaptations to the manufacturing scenario are in developmental stages (Butun, 2020).
