The Significance of Machine Learning in the Manufacturing Sector: An ISM Approach
Round 1
Reviewer 1 Report (New Reviewer)
When discussing the methodology, please prepare a flow chart for a better understanding of the problem at hand.
It is worth reading the literature published in MDPI.
Author Response
Response Sheet 1
(Changes have been highlighted in the manuscript)
Comment 1: When discussing the methodology, please prepare a flow chart for a better understanding of the problem at hand.
Response 1: As per the valuable suggestion by the reviewer we have added the flow chart for better understanding of the problem.
Author Response File: 
Author Response.docx
Reviewer 2 Report (New Reviewer)
My Comments and Suggestions for the Authors are given below:
1. Using Interpretive Structural Modelling (ISM), the researchers analyse the relationships between factors that are thought to be important in the development of machine learning in the manufacturing industry. The ISM used in the article is a well-known method and a correct application was made by the authors in the article.
2. However, the novelty and its original contributions are completely unclear. More emphasis should be placed on the innovative and original contributions of the study to the field and emphasis should be placed on why this study is important.
3. One of the most important points of the article is the criteria determined in the study. Because all the remaining modelling and analysis efforts of the study are based on these criteria. The authors state that a comprehensive literature review was conducted to identify the factors, and then, through a brainstorming session, a list of factors was drawn up, concluding the study with eleven critical factors. However, there is not enough detailed explanation about this process and how this process was conducted. In other words, it is not explained how the detailed evaluation process is related to the determination of these eleven criteria, and why the criteria that are not taken into account, if any, are ignored. This should be explained in more convincing detail based on analytical considerations.
4. In the second paragraph of the conclusion part, it is seen that the results are presented with reference to the studies done by other authors. As much as possible, the original results of the study should be presented without reference to the results. This paragraph should be presented in previous sections.
Author Response
Response Sheet 2
(Changes have been highlighted in the manuscript)
Comment 1: Novelty and original contributions are unclear. More emphasis should be placed on the innovative and original contribution in the field and emphasis should be placed why this study is important.
Response 1: Taking the valuable suggestion from the reviewer we have mentioned why our work is unique in itself. And highlighted the contribution in the field.
Less work has been done on the digital side of the manufacturing sector. Though some authors have mentioned machine learning in manufacturing sector, there were no prioritization of factors involved. Since no recent comprehensive work based on the general understanding of significance of machine learning in manufacture sector using ISM was published, this paper is a step ahead in that direction. This paper is an attempt to rank the different factors involved.
Comment 2: It is not explained how the detailed evaluation process is related to the determination of these eleven criteria, and why the criteria that are not taken into account, if any, are ignored. This should be explained in more convincing detail based on analytical considerations.
Response 2: Referring to the reviewer’s query, we have explained the process on how we concluded on the eleven factors.
A conference was conducted in our KIIT University consisting of academicians and industrial experts from the manufacturing sector. Where an open discussion was conducted to point out which were the significant factors affecting ML in MS. Through a brain-storming session the list of fifteen factors was established. These were further ranked by the experts, and concluded with the eleven critical factors. The eleven factors concluded got score from 8 to 4 in the ranking procedure. Other factors from the short listed factors like Data storing capacity of the industry, Selection of algorithms, Flexibility of industrial capability and Ignorance to the technology scored too low ( Score =1)had not been included in the final list.
Comment 3: In the second paragraph of the conclusion part, it is seen that the results are presented with reference to the studies done by other authors. As much as possible, the original results of the study should be presented without reference to the results. This paragraph should be presented in previous sections.
Response 3: As per the reviewer’s request the second part of the conclusion has been omitted. And it’s discussion is mentioned in the previous section.
Author Response File: Author Response.docx
Reviewer 3 Report (New Reviewer)
The article is interesting. It raises important contemporary issues. However, the Authors have not done everything right and therefore corrections are needed.
First of all, the colour of the font needs to be adjusted because of the Publisher's requirements. I understand that the paper was written by several people, but the colour should be changed before submission.
The abstract lacks a clearly stated purpose of the study. Please add it so that it is clear exactly what the manuscript is written for. The objectives formulated in the introduction are appropriate and can be copied into the abstract.
Extensive literature review.
Section title should be changed: 4. Results and Discussions. These are the results of the analyses, but there is no discussion. The discussion is comparing the results obtained with studies published by other scholars. This could also be divided into two parts by adding a discussion.
Author Response
Response Sheet 3
(Changes have been highlighted in the manuscript)
Comment 1: the colour of the font needs to be adjusted because of the Publisher's requirements. I understand that the paper was written by several people, but the colour should be changed before submission.
Response 1: Addressing the reviewer’s concern we have modified document as per the format. And only recently made changes are highlighted.
Comment 2: The abstract lacks a clearly stated purpose of the study. Please add it so that it is clear exactly what the manuscript is written for.
Response 2: As per the reviewer’s valuable solution the purpose of the study has been mentioned in the abstract. Giving a clear indication on what the manuscript is written for.
The aim of this research was to identify these significant factors and develop a conceptual relationship between them.
Comment 3: Section title should be changed: 4. Results and Discussions. These are the results of the analyses, but there is no discussion. The discussion is comparing the results obtained with studies published by other scholars. This could also be divided into two parts by adding a discussion.
Response 3: We have identified the issue raised by the reviewer. To which we have added a paragraph as discussion to Section 4.
Researchers in the past have analyzed the significance of machine learning in the manufacturing industries. And since then critical factors were highlighted which showed it’s effect on the growth of the industry. Paturi and Cheruku [58] in their work recorded the development of machine learning in the manufacturing sector. Which showcased that it has been most useful in the optimization of the work processes because of it’s ability to predict situations with accuracy. Since data has been the main basis of operation for machine learning, Huo and Chaudhry [59] noted that use of algorithm is critical. Their work focused on recognizing profit making factors for a Chinese manufacturing company. In contrast, this research identifies factors which would help improve the significance of machine learning in the manufacturing sector. A contextual interrelationship was developed between the factors through the ISM technique. The result reveled that Demand of the Market, Revenue of the Industry, Skilled Labour, Lack of Man Machine and Manpower are the most determining factors. Which implies that customer’s demand and industrial capacity needs in depth understanding before the application of technology. Result further reported that Order Fulfillment, Significance of Machine Learning, Process Automation, Diversified Product Generation, Just-in-time Product Generation are dependent factors. It suggests that these factors are objectives which the industry should aim to fulfill. The work provides a structural development will help industries make step by step evaluation of their work. Helping policy formulation to direct the need to fill the gap where required. Thus this research contributes a distinctive analysis through the relationship developed between the factors that would help industries focus on their limitations to improve the situation of machine learning. Mohapatra et al. [60] used ISM methodology to study the prioritizing factors in the industry by integrating machine learning. And concluded revenue generation as the top priority. Solke et al. [61] mentioned flexibility of working machines as top priority using the ISM methodology.
Author Response File: Author Response.docx
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
This paper tries to investigate different scenarios where Machine Learning models/techniques/algorithms/solutions could be used in the Manufacturing sector. Unfortunately, this quite a straightforward yet vast problem is tackled with a very chaotic and lacking proper organisation approach.
On of the main issues that I have managed to identify in the paper, is the lack of proper definition of the used terms and ideas. For instance, from the very beginning of the paper it is unclear what do the authors mean by the term "Machine Learning" - sometimes it is used as a discipline, sometimes as a "model" (in this context I do not understand in what sense IoT can be a model), sometimes as a set of detailed technical solutions and sometimes as a "technology".
Second important issue is the presentation quality. Unfortunately, the paper contains multiple language errors. Moreover, half of the paper is a huge table that is supposed to present literature review, however to the lack of proper introduction and organisation it is unclear what information, and more importantly, what is the purpose of such a presentation.
Third issues revolves around lack of proper introduction to the ISM methodology and process, as the main (even in the title) part of the paper. Unfortunately, it is left to the reader to overview the Interpretive Structural Modelling (ISM). It is also unclear, why the authors have focused on this particular approach - the paper tries to also use two different approaches SWOT and MICMAC.
Fourth issue is the lack of proper organisation and flow of the paper. Even though, the paper has 19 pages, it seems that the content is very rushed and chaotic. Major revision is required to better steer the reader.
Fifth issue is the lack of significant results. In a way the paper ends with a conclusions that it is, in general, beneficial to use ML in Manufacturing Sector, yet people have some security and damage concerns.
The topic of the paper is really interesting and seems relevant. However, the paper requires a complete redesign, with a more concrete and well-defined goal.
Author Response
Response Sheet 1
(all the revision in the manuscript has been highlighted in yellow)
Comment 1- Unclear definition of the term “Machine Learning”
Response to comment 1- As per the valuable suggestion of the reviewer we have added the brief view on ISM in the abstract and explained the idea of what Machine Learning is in the introduction of the paper to clarify the concept and how it is related to the Manufacturing Sector in the revised manuscript. We are mentioning the added sentence for the sake of clarity.
Machine learning, when simply explained a branch of Artificial Intelligence(AI) and Computer Science that focus on the use of data and algorithms to imitate the way humans learn. Through this computer systems are developed which are thus able to learn and perform without the need of explicit instruction.One of the most advanced technologies today that are being implemented in industries where they are used through different work processes especially in the manufacturing sector. This implementation has helped better monitoring and execution of various processes by reading and analyzing through previous collected data and perform through simple instructions. In several cases Machine Learning can be introduced through different models of the working environment in any industry which will not only help as in above mentioned spots but also will work through the strategic distribution and manpower utility within the industry.
Comment 2- Appearance of language error and presence of huge table for literature review.
Response to comment 2- As per the reviewer’s concern for the language error, the paper has been thoroughly revised and reviewed and any such occurrence has been immediately addressed which are highlighted in yellow in the revised manuscript. Regarding the literature review, the tabular format has been suggested to convey a detailed overview of the industries/sectors through the research work done which have implemented Machine Learning and what benefits and challenges were faced by them.
Comment 3- lack of proper introduction to ISM
Response to comment 3- As per the viewer suggestion we have added a brief introduction to the ISM (Interpretive Structural Modelling) and specified why it has been used for the research work.We are mentioning the added sentence for the sake of clarity.
Interpretive Structural Model (ISM) is a methodology which simply puts forward the relation between the factors by comparing the effectiveness among themselves. The result form the ISM Model helped position the factors according to their effectiveness to purpose as short term focus and long term focus.The ISM methodology has been thus implemented as it helps give a rational result and help develop a graphical representation for a complex situation. Since the aspect “Machine Learning in Manufacturing Sector” is vast and only a defined set of factors has been considered the ISM methodology fits the perfect analysis for the situation.
Comment 4- lack of significant results
Response to the comment 4- As per the suggestion the result has been reviewed and made specific as per the research work done, highlighted in yellow in the revised manuscript . The goal of this paper is to give an underline to the future prospects who wish to implement Machine Learning in their work. The factors considered and the results declared will give an idea of what to look for when one ventures into the corporation of Machine Language in their system. We are mentioning the added sentence for the sake of clarity.
Though the ISM Analysis has been on the focus we have also tried to add two more analysis, the SWOT Analysis and the MICMAC Analysis. In the ISM Analysis, through the study five levels have resulted in the ISM Model .
The MICMAC Analysis represents four quadrants, Driver, Linkage, Autonomous and Dependence. The representation of each quadrants is as; Driver: the factors present in this quadrants impact all the other factors present, Linkage: factors present in this quadrant are the most vulnerable as slight changes to their activity will impact the whole of the factors in the system and may change their working functionality, Autonomous: with weaker dependence power and driving power the factors here function on their own and do not impact the other factors, Dependence: the factors in this quadrant are highly dependent on the performance efficiency of the other factors.
we realize that the basic requirement for each industry is same. The factors that has been suggested in the study gives the idea on how an industry could initiate the implementation of Machine Learning through their processes. Through the ISM Model and MICMAC Analysis industries could clearly help themselves on which factors needs the action referring to their scenario. Not only just focusing on the long term focus and the short term focus, managers can take help at specific stages according to their requirement, and train their systems to work accordingly. Beginning with the initial stages of implementation of Machine Learning cannot be an easy task as it requires a large amount of background data to process it’s systems and make the overall task efficient.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The current manuscript, Significance of Machine Learning in Manufacturing Sector: An ISM Approach, is written and presented with few details in the research steps and results.
Some points are required to improve or clarify:
1. In the Methodology section, a brief presentation of the methods used in the study is required. Here are the empirical results obtained.
2. It would also be useful to provide more details on the results in Tables 3, 4 and 5.
3 .What are the weaknesses of this research?
Author Response
Response Sheet 2
(all the revision in the manuscript has been highlighted in yellow)
Comment 1- In the Methodology section, a brief presentation of the methods used in the study is required. Here are the empirical results obtained.
Response to comment 1- As per the valuable suggestions put forward by our reviewer we have added a brief overview of the methods in the methodology section. Thus showing our results obtained.We are mentioning the added sentence for the sake of clarity.
The SWOT Analysis is a strategic technique to facilitate realistic, fact based and data driven outlook on the strength, weakness, opportunities and treats in an industry.
The ISM methodology has been thus implemented as it helps give a rational result and help develop a graphical representation for a complex situation. Since the aspect “Machine Learning in Manufacturing Sector” is vast and only a defined set of factors has been considered the ISM methodology fits the perfect analysis for the situation.
The MICMAC analysis is made from the reachability matrix (Table 4), is the graphical development of the factors based on their driving power and dependence power.
Comment 2- It would also be useful to provide more details on the results in Tables 3, 4 and 5
Response to comment 2- The table 3, 4 and 5 show the steps following the Interpretive Structural Modelling(ISM). Table 3 shows the comparison of factors on the basis of A, V, X and O considered in ISM method. Following which Table 4 is the binary conversion of the Table 3 as per steps followed. And Table 5 compares the dependency and driving power of each factor, comparing which we get the different levels for the ISM model.
Comment 3- What are the weaknesses of this research?
Response to comment 3- As per the reviewer’s proposal to state weakness of the research, we have tried to give a brief idea as required in accordance to the research. We are mentioning the added sentence for the sake of clarity.
Though this research has tried to showcase certain factors which are influencing the Machine Learning in Manufacturing Sector, there are different aspects even to the Manufacturing Sector. Manufacturing Industries of different products may have their own specific requirements when incorporating the Machine Learning. This paper only expresses the general requirement of any Manufacturing Sector but much more background details will be required if it is made industry specific.
Author Response File: Author Response.docx
Reviewer 3 Report
The authors show a very good coverage of several applications of ML in Manufacturing sector. Very useful and interesting to read about!
Author Response
Response Sheet 3
(all the revision in the manuscript has been highlighted in yellow)
Comment 1- The authors show a very good coverage of several applications of ML in Manufacturing sector. Very useful and interesting to read about!
Response to comment 1- We are extremely gratified by the reviewer’s response to our research work. We are glad, we were able to put forward this idea of “Machine Learning in Manufacturing Industry” which instigated the interest.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
- this work reviews findings from 24 papers, yet the references contain 25 positions - it means that only 1 additional (besides the studied ones) reference is used. - it is unclear how the particular 24 papers were selected - formatting is off in several places - paper still presents a huge table with very condensed and general description of some selected aspects tackled in each paper. Unfortunately, it is unclear what are individual parts of the table - what do the authors mean by "Factors Affected" (and later "supporting" and "affecting"). Moreover, how the application area was derived, for instance what kind of application area is "Enterprises" or "Industry 4.0"? - presented descriptions (in the table) are not adequate for an "in depth review" and lack proper organisation. For instance, findings in the last paper state "the study identifies subjects for better strategic planning in finance. Indicated improved forecast and returns." How this is related to ML? How ML was used there? I think the authors should better organize this fragment and provide some more goal oriented and focused descriptions. - I think it would be beneficial to convert the table into in-line text. Why? The tables is not using a certain condensed representation of selected aspects and provides just a short textual description of each aspect. There is no particular structure to the description, as such there is no benefit of tabular representation, where some elements can be quickly identified and compared. Here all descriptions are just short paragraphs. Inline organisation could save a lot of empty space introduced by tabular representation. - still major language mistakes, like missing verbs, etc. - several sentences are off, for instance "Machine learning, when simply explained a branch of Artificial Intelligence(AI) and Computer Science that focus on the use of data and algorithms to imitate the way humans learn." - several sentences are still unclear, for instance "All the tasks performed can be monitored, controlled and can be implemented to train professionals by adopting various models like Machine Learning, Internet Of Things, Artificial Intelligence and others." In what sense, "collection of raw material till the delivery of the goods" "can be implemented to train professionals by adopting" an advanced model? - there is a huge methodological gap between presentation of 24 papers' findings and analysis step of the study. It is unclear, how individual conclusions and characteristics (used in analysis) were derived from the performed literature review. - presented conclusions, from the SWOT analysis, seem to be so general that they fit nearly every application area for ML - as they state some obvious aspects of specificity of ML approaches. Moreover, it is unclear how they related to the studied papers. - "background research" managed to identify 11 factors "influencing the manufacturing sector through the aspect of Machine Learning" - but these factors are not introduced (listing is not an introduction) and presented in the paper. It is unclear, how they were derived and what is their "meaning". - resultant SSIM table is "completed on the basis of proper research and review of papers published by different authors in various fields" - I do not think that it is a good justification, you should present your methodology and way how these individual keynotes were assigned. Several questions arise here - for instance, is one paper enough to name such a type of relation, are two papers enough, what in the case in which two papers have different opinions, etc. - presented Result and discussions section lacks discussion part - it just briefly introduced the results - presented conclusions are not justified, for instance I would say that "the study gives the idea on how an industry could initiate the implementation of Machine Learning through their process". In particular, this is the first time the idea of "how an industry could initiate implementation" is mentioned.