Application of industry 4.0 enablers in supply chain management: Scientometric analysis and critical review

Abstract

The supply chain management (SCM) environment is rapidly evolving as a result of the critical role industry 4.0 enablers are playing. Consequently, and to leverage the power of industry 4.0 enablers (I4Es) including; artificial intelligence (AI), machine learning (ML), internet of things (IoT) and big data (BD), researchers and industry practitioners have employed these I4Es to resolve several pain points in supply chain management at all levels, improve operational efficiency, manage demand volatility, tackle cost fluctuations, and make data-driven decisions. Thus, I4Es are working as an evolutionary catalyst for supply chain management in myriads of ways. As such, the application of I4Es in supply chain management (I4Es-in-SCM) research has witnessed tremendous growth over the past years. This study conducted a scientometric analysis and critical review of the I4Es-in-SCM research to monitor trends, visualize the structure of knowledge, identify gaps, and highlight future research avenues. The paper recruited and analysed bibliographic data of 786 papers from Scopus on the application of I4Es-in-SCM research. Analysis showed that the last two decades witnessed a phenomenal growth in research on the application of I4Es-in-SCM, with at least 42 % of all countries making contributions. The analysis showed wider collaboration between countries and noticed a rather significant collaboration among researchers within a given continent. The study also identified the most influential researchers, journals, and countries as well as trending themes and topics in the application of I4Es-in-SCM research. After delineating boundaries of scientific knowledge, the study proffered areas that require further research. The novelty of this study lies in providing a more holistic statistical and visualized analysis of the structure of knowledge, productivity, and scientific collaborations of researchers, journals and countries in the application of I4Es-in-SCM management research. Accordingly, the study outcomes may serve as a useful reference to supply chain academics, early-stage researchers, practitioners, policymakers, and organizations in understanding the structure of knowledge on the application of I4Es-in-SCM research and may constitute a basis for future research.

1. Introduction

In recent times, industry 4.0 enablers (I4Es) has become omnipresent, gaining the attention of corporate leaders, government officials, academic scholars [1] and decision-makers [2]. Although few researchers argue that I4Es can be divided into organizational and technological enablers [[3], [4], [5]], this research considers four critical technological enablers, comprising big data, artificial intelligence, machine learning and IoT.

Data is ubiquitous in the modern digital age, and its volume is expanding at an exponential rate. Companies now have access to massive volumes of data from a wide variety of sources, facilitated by IoT. The process of making informed decisions involves more than just gathering information. The true value is revealed through processing, analyzing, and interpreting the data. This establishes the fertile ground for complimentary of artificial intelligence (AI), machine learning (ML), and large datasets in the industry 4.0 disruption. According to Ref. [6] organizations in a variety of industries stand to gain trillions of dollars from adopting these technologies.

Big data is the term used to describe the massive amounts of data, both organized and unstructured, that are created on a daily basis. It encompasses social media information, client conversations, and IoT sensor data. On the other side, technologies like artificial intelligence and machine learning allow computers to learn and develop autonomously over time. Insights gained from real-time analysis of the massive amounts of data produced by IoT devices can help businesses make better decisions.

According to Ref. [7] 75 % of enterprise data will be generated and processed at the Edge, including IoT devices by the year 2025. This is up from 10 % in 2018. In tandem with the traction in the industrial application I4Es is a growing scholarly interest in the I4Es in the supply chain management literature [8]. The growing body of literature on the application of industry 4.0 enablers in supply chain management (I4Es-in-SCM) demands scientific monitoring and tracking to cope with the rapidly evolving nature of the research.

The scientific community already researchers recognized this necessity and synthesized current literature on the applications of I4Es-in-SCM research, but mostly covering short periods or certain enablers. For example, Ref. [9] investigated big data analytics within logistics and supply chains covering the period from 2004 to 2014. Similarly, Ref. [8] presented a bibliometric and thematic analysis of big data research papers from 2008 to 2016. Equally, Ref. [10] presented a comprehensive literature review paper focusing on different applications of machine learning algorithms and Natural Language Processing (NLP) techniques. Finally, Ref. [11] investigated big data analytics research and its application in supply chain management between 2010 and 2016 and provided insights to industries.

While relevant, these prevailing reviews failed to capture the application of a comprehensive set of the I4Es-in-SCM. Capturing a relevant set of the I4Es is necessary because most of the enablers are hardly independent in the industry 4.0 revolution. To extend the extant reviews and build on what has been achieved, this paper aims to a scientometric analysis and critical review of I4Es-in-SCM research. This holistic approach aims to: (i) identify and evaluate the most active and influential researchers, journals, and countries in the application of I4Es-in-SCM research; (ii) identify high-impact journals and landmark articles on the application of I4Es-in-SCM research; (iii) explore the scientific collaboration patterns and networks of researchers and countries in the application of I4Es-in-SCM research; (iv) evaluate and shed light on trending research topics and themes on the application of I4Es-in-SCM research; and (v) suggests areas that require further research. These aspects of scientometric analysis are sufficient to provide clear insights and understanding of the structure of scientific knowledge [12,13] on the application of I4Es-in-SCM research.

It is expected that this substantive assessment will establish the broader themes of research studies on the application of I4Es-in-SCM and highlight areas that require additional research and investigation as well as opportunities for collaboration between researchers and research institutions. The all-inclusive assessment provides concrete evidence to both the academic community and industrial practitioners on this state-of-the-art research on the application of I4Es-in-SCM. The novelty of this study lies in providing a more holistic statistical and visualized analysis of the structure of knowledge, productivity, and scientific collaboration of researchers, journals, and countries in the application of I4Es-in-SCM research. The outcomes of the study serve as a useful reference to supply chain academics, early-stage researchers and practitioners on the structure of scientific knowledge in the application of I4Es-in-SCM and may provide a basis for future research. The overall paper is structured as follows: 1 – Introduction, 2 – Conceptual background, 3 – Methods and materials, 4 – Results and discussions, 5 – Managerial implications and 6 – Conclusions.

2. Conceptual background

2.1. Big data

Businesses today are dealing with big datasets characterized by 4Vs: large volume, velocity, variety, and veracity, accordingly, big data analytics techniques have emerged as an approach to gain competitive advantage for organizations in recent years [14,15]. The amount of structured and unstructured data emerging from their interactions increases exponentially [16] fuelled by the introduction of social media with millions of photos, videos and tweets posted daily and exchanged between more than 3 billion users [17].

Many researchers argue that big data is not only the “next big thing” in management but might be the “fourth paradigm of science” [1] or even the “new paradigm of knowledge assets” [18] with others going further to claim that it is “the next frontier for innovation, competition, and productivity” [19]. Furthermore, Ref. [17] stressed that big data will act as the new oil and, consequently, Artificial Intelligence (AI) companies will probably be the utility providers of the future, transforming data into information similar to how current utility providers transform the oil into energy. Several fields, activities and industries, including the supply chain environment industry generate copious quantities of datasets. In the wake of the big data revolution, supply chain management researchers have sought to understand and take advantage of the widely available data in the supply chain industry to inform data-powered decisions and management.

2.2. Machine learning (ML)

Machine learning is playing a key role in reshaping many industries. Employing IT systems is enabling the recognition of patterns within existing databases and algorithms, which leads to providing adequate solutions and informing the decision-making process [20]. Likewise, ML can help in the demand forecasting approach which in turn improves the efficiency performance of supply chains [21,22].

2.3. Artificial intelligence (AI)

Researchers argue that artificial intelligence improves internal processes and helps organizations manage their financial resources [23]. Moreover, Ref. [24] claims that Al is serving as a key enabler for enterprises to gain the required flexibility and agility and become more competitive in cost. Many organizations employed AI to gain competitive advantage. For example, Kiwibots (a US-based restaurant) claims that their AI-powered robotic vehicles that seamlessly mesh into the fabric of the urban landscape to deliver meals are very fast, cheap and make people's lives more convenient [25].

2.4. IoT

IoT from another hand, can help automate the identification process of products, trace and track products globally, achieve transparency and reduce time and cost in supply chains [26]. Furthermore, the application of IoT can yield new opportunities to improve supply chain integrity and operational performance [27]. Finally, IoT can enable real-time data collection and increase data efficiency as long as enables real-time communication within the supply chain [28].

The applications of I4Es in supply chains are too many and can add value upstream, downstream or within any of the main stages or processes of supply chains. For example, Ref. [29] asseverated that I4Es can assist in strategic sourcing activities and help businesses select appropriate suppliers. I4Es are an important procurement enabler as they allow for conducting spend analysis and consequently segmenting suppliers and big-ticket items [9,30]. I4Es can also play a vital role in supply chain network design [9], help to improve demand planning and forecast, improve delivery times by synchronizing shipments, identify better ways of reducing communication gap between manufacturers and suppliers, improve supply chain and logistics process transparency, and tackle cost fluctuations in the supply chain [31]. Furthermore, I4Es can support product design and development ([32,33], contribute to the production and operational efficiency [34] and provide unlimited capabilities in inventory management [9,35]. Additionally, some scholars stress that I4Es are valuable in logistics and distribution including vehicle routing and transport modal split [36] while others consider I4Es as a strategic asset that can help in managing supply chain risks [37].

3. Methods and materials

3.1. Scientometric analysis

The scientific literature on the application of I4Es-in-SCM can be reviewed through manual approaches or software-aided approaches. However, each approach to the literature review has its unique limitations and strengths. Manual reviews are feasible when the corpus of documents is few, whereas science mapping approaches can analyse a large corpus of documents [13,38]. The large quantity of papers on the application of I4Es-in-SCM renders it inefficient when using manual reviews. Manual qualitative reviews have also been widely censured for the higher degree of subjectivity and lack of reproducibility [39]. Additionally, qualitative manual reviews cannot examine the collaboration between authors, institutions and countries [40] to facilitate knowledge exchange, innovation diffusion and cross-cultural research.

However, science mapping can address the limitations of using a manual qualitative approach to analyse the large corpus of documents [41]. Scientometric review constitutes a powerful science mapping tool that employs computational and statistical techniques to analyse, model, detect, and visualize the intellectual structure of a scientific research domain [12]. This study employed a scientometric approach to conduct a robust quantitative statistical analysis and visualization of research on the application of I4Es-in-SCM. Scientometric analysis has been widely used to describe the knowledge structure and distributions patterns in the contribution of authors, documents, journals, countries, and keywords in the scientific research domain of green building [40], artificial intelligence [38], building information modelling [42] and safety culture research [43].

3.2. Database selection

Several literature databases are used to conduct scientometric analysis in different fields, including Elsevier's Scopus, Clarivate Analytics' Web of Science, PubMed, Engineering Village, EBSCO, Taylor and Francis, and Emerald Insight. The suitability of the database depends on the context, scientific research field, and scientometric tools adopted [13]. After a preliminary comparison of the adumbrated databases, this study adopted Scopus to recruit the bibliographic datasets of research documents on the application of I4Es-in-SCM. Scopus has been used in previous scientometric reviews because it is comprehensive and provides a wider coverage of research documents and a structured interface for queries [13,40].

3.3. Literature search strategy

After selecting a literature database, the study developed two sets of search terms. One set represented the different terms used by researchers to describe I4Es in the literature and the other set represented terms used by researchers to describe supply chain management in the literature. These terms were used to recruit the bibliographic records of research articles on the application of I4Es-in-SCM. The following search string was implemented in Scopus to extract the bibliographic records.

(TITLE (“big data” OR “internet of things” OR iot OR “data analytics” OR “unstructured data” OR “data-driven” OR media OR “machine learning” OR “deep learning” OR “artificial intelligence” OR “predictive analytics” OR “data science” OR “data-based”) AND TITLE (“supply chain” OR “supply chain management” OR logistics)) AND (LIMIT-TO (DOCTYPE, “ar”) OR LIMIT-TO (DOCTYPE, “cp”) OR LIMIT-TO (DOCTYPE, “re”)) AND (LIMIT-TO (LANGUAGE, “English”)) AND (LIMIT-TO (SRCTYPE, “j")).

The search was conducted on 20 November 2021 and generated bibliographic data from 786 research documents. As shown in the search string, the results were filtered by document type, source type, and language of publication. For wider coverage, the study filtered articles, review articles, and conference papers. The source type was restricted to Journals only and the language of publication was restricted to English. The dataset comprised bibliographic records of 786 articles, review articles and conference papers on the application of I4Es-in-SCM, published between 1994 and 2020.

3.4. Data attributes

The dataset was downloaded as a CSV file and each bibliographic record contained the metadata of a published article, review article and conference paper. The metadata included a list of authors, paper title, abstract, a set of keywords, and a set of references cited by the article, review article and conference paper. Each reference contained the name of the first author, year of publication, source type, volume number and digital object identifier (DOI) reference. The validity of the data was acceptable due to two reasons. First, the study relied articles indexed in Scopus, which is widely recognized as a reliable database. Second, the study did not include grey literature, which may not usually be peer-reviewed. The study also minimized the risk of bias because it included studies across all time span indexed in Scopus, and hence both recent and earlier studies have been analysed. However, the study was restricted to English language publication, which could be considered a bias, but English is the most widely reported scientific language.

3.5. Scientometric tool

There are several commercial and open-access scientometric tools used to analyse scientific literature, but each has its unique strength and limitations. The most widely used scientometric tools include; Cite-Space, VOSviewer, Gephi, Bibliometrics, and HistCite. Due to the limitations of each tool, existing scientometric analyses combined two or more tools to improve the depth of analysis. However, this research adopted VOSviewer as the sole scientometric tool to analyse the literature. VOSviewer is a powerful software tool with text-mining functionality that can construct and visualize bibliometric networks [12]. The tool accepts CSC files of bibliographic records from Scopus and is suitable for visualizing networks of authors, journals, and countries, based on citation, bibliographic coupling, co-citation or co-authorship relations. The text mining functionality enables science mapping researchers to construct and visualize co-occurrence networks of trending topics and themes to spot main research areas [12]. VOSviewer has been solely used in previous scientometric studies to analyse research on green building [40], sustainability of offsite construction [13], and safety culture [43]. Overall, VOSviewer was used because it offers the basic functionality required for producing, visualizing, and exploring bibliometric networks, accepts bibliographic records from Scopus, and is known to the authors.

3.6. Analytical procedure

The VOSviewer software tool uses authors, documents, journals, institutions, and keywords as the units of analysis [ [12,44]]. Based on the metadata of these units of analysis, VOSviewer can generate both maps and networks. VOSviewer supports co-authorship analysis, document co-citation analysis, co-occurrence network of keywords, bibliographic coupling, temporal analysis, geospatial analysis, source citation analysis, bootstrap resampling, spectral clustering and thematic areas visualization [ [12,13]]. The scientometric analysis involved the construction and visualization of networks of authors, institutions, journals, and countries. From these networks, the study conducted four types of analyses.

First, the study analysed the scientific research productivity of authors, journals, and countries to identify the most contributing researchers and economies in the application of I4Es-in-SCM research. Second, the study analysed the scientific collaborations of authors and countries, as well as the co-citation analysis of journals. Third, the study analysed the impact of articles, authors, journals and countries in the application of I4Es-in-SCM research. In particular, the study computed the Field-Weighted Citation Impact (FWCI) of the most cited articles. FWCI is a novel metric that considers the disparity in research behavior across disciplines and indicates how the number of citations received by an article compares with the average number of citations received by all other similar articles indexed in the Scopus database. It defines similar articles are those articles in the Scopus database that have the same publication year, type and discipline. FWCI captures citations received in the year of publication plus the following 3 years and measures the unbiased prestige of an article's citation performance. A FWCI of 1.00 indicates that the article has been cited at world average for similar publications. A FWCI >1.00 indicates that the article has been cited more than would be expected based on the world average for similar publications. For instance, an FWCI of 1.65 indicates that the article has been cited 65 % more times than expected. A FWCI <1.00 indicates that the article has been cited less that would be expected based on the world average for similar publications. For instance, an FWCI of 0.75 indicates 25 % less cited than world average.

Finally, the study conducted a co-occurrence network of keywords to map the intellectual structure of knowledge and trending research areas on the application I4Es-in-SCM. Overall, the study analysed the annual growth pattern of the publications, co-authorship networks, document citations, source citations, co-occurrence network of keywords, and geospatial network of countries to understand the publication trends, trending topics and themes, landmark articles, and most productive researchers, research outlets, research outlets and countries. As such, these analytical techniques were considered adequate in understanding the knowledge structure from the scientific body of literature on the application of I4Es-in-SCM [12].

4. Results and discussion

4.1. Annual growth of publications on the application of I4Es-in-SCM research

It is useful to track the growing trend in the interests of researchers in understanding the application of I4Es-in-SCM research. Fig. 1 shows the annual publications trend on the application of I4Es-in-SCM research over a period of 2.5 decades. As shown in Fig. 1, researchers have shown interest in the application of I4Es-in-SCM since 1994. This period also marked the renewed renaissance of the I4Es revolution and shows that supply chains are reservoirs of a large, unstructured and rapidly accumulative datasets. The second-degree polynomial trend line in Fig. 1 depicts the accumulation of scientific knowledge and progress on the application of I4Es-in-SCM research. As shown in Fig. 1, the period between 1994 and 2007 witnessed a minimal annual growth in publications.

Fig. 1.

Annual publication growth pattern between 1994 and 2020.

However, the significant and sustained interest in the application of I4Es-in-SCM research started to be recorded from the year 2008. The growth sustained and reached the highest ever in the year 2020 when at least 230 articles were published on the application of I4Es-in-SCM research. The last decade (2010–2020) witnessed the most significant growth in the application of I4Es-in-SCM research and could be considered the first decade of the I4Es-in-SCM research revolution. This exponential trend witnessed during the last decade indicates that research on the application of I4Es-in-SCM has received considerable attention from the supply chain scientific community in recent times. The all-time peak of 230 published studies in 2020 is not surprising since supply chains are now generating significant volumes of datasets, and domain researchers have sought to leverage data science, artificial intelligence and machine learning to inform data-driven supply chain decision-making, operations, optimization, and management. Overall, Fig. 1 shows that supply chain management may be fully driven by big data analytics in the coming decades. Fig. 2 provides a distribution of the 786 papers for the I4Es-in-SCM. It shows that BD, IoT, and ML are the most researched I4Es in SCM. It also highlights the relatively limited studies on the application of AI in SCM.

Fig. 2.

Distribution of papers across the I4Es-in-SCM.

4.2. Scientific productivity

4.2.1. Most productive authors

The bibliographic dataset of the 786 articles was contributed by 2040 authors. However, some researchers are more active and productive than others in the application of I4Es-in-SCM research. Identifying the most productive and high-impact researchers could facilitate future collaboration in research and grant applications [40]. Table 1 shows the researchers who published at least 5 articles on the application of I4Es-in-SCM research. The minimum number of 5 documents was a default specification in the VOSviewer software [12] and has been used in previous scientometric reviews.

Table 1.

Most productive researchers in the application of I4Es-in-SCM research.

Author	Articles	Citations	Total Link Strength
Gunasekaran A.	18	1407	11
Wang Y.	11	52	3
Zhang Y.	11	116	3
Dubey R.	10	778	10
Liu P.	9	85	0
Huang G.Q.	8	779	2
Lim M.K.	8	181	1
Wang J.	8	64	4
Childe S.J.	7	719	7
Li J.	6	8	0
Liu X.	6	74	1
Liu Y.	6	212	4
Papadopoulos T.	6	990	6
Wang H.	6	24	2
Wang M.	6	40	1
Wang X.	6	37	2
Zhang X.	6	10	4
Zhang Z.	6	21	1
Hazen B.T.	5	587	1
Jermsittiparsert K.	5	21	0
Lee S.	5	90	0
Li L.	5	26	0
Singh S·P.	5	128	0
Wang L.	5	41	3
Yan B.	5	77	1
Yu W.	5	84	0
Zhang N.	5	120	2

As shown in Table 1, 27 of the 2040 researchers published at least 5 articles on the application of I4Es-in-SCM research. Based on the number of publications, the top 5 most productive academics in the application of I4Es-in-SCM research include Gunasekaran A. who published 18 articles, followed by Wang Y., Zhang Y., Dubey R., and Liu P. who published 11, 11, 10, and 9 articles on the application of I4Es-in-SCM, respectively. Table 1 also shows the impact of the 27 authors. Based on citations, the top 5 most cited authors in the application of I4Es-in-SCM research include Gunasekaran A. with 1407 citation counts, followed by Papadopoulos T., Huang G.Q., Dubey R., and Hazen B.T. with 990, 779, 778, and 587 citations counts. The values of the total link strength in Table 1 represent the number of publications a researcher co-authored with others [12]. Based on the total link strength indices, the 4 most collaborative authors include Gunasekaran A. who co-authored 11 articles, followed by Dubey R., Childe S.J., and Papadopoulos T. who co-authored 10, 7, and 6 papers with others.

4.2.2. Most productive countries

At least 82 countries have contributed 786 papers on the application of I4Es-in-SCM research. It is useful to identify the most productive countries in the application of I4Es-in-SCM research to facilitate international collaboration and innovation [40]. Of the 82 countries, 25 published at least 8 articles. Based on the number of documents, the 5 most productive countries in the application of I4Es-in-SCM research include China, the United States, India, the United Kingdom, and Australia which contributed 251, 119, 87, 81, and 33 articles. This outcome is realistic and expected because these countries are tech giants and have made large investments in complementary fields such as big data, data science, artificial intelligence, machine learning and supply chain studies. The outcome further suggests that these countries are promoting data-driven supply chain decision-making and management.

Table 2 also shows that these countries have made different impacts in the application of I4Es-in-SCM research. Based on the citation counts, the 5 economies with the highest academic impact include the United States (4131), the United Kingdom (3006), China (2635), India (1347), and Hong Kong (1320). This indicates that articles published by researchers from these economies are well-received and referenced by other researchers in the application of I4Es-in-SCM research. The values of the total link strength in Table 2 represent the number of publications a country co-authored with other countries [12]. Based on the total link strength, the 5 most collaborative countries in the application of I4Es-in-SCM research include China (74), the United States (60), the United Kingdom (52), Hong Kong (24), and France (24). Thus, researchers in these countries are excellent academics to collaborate in I4Es-in-SCM funding applications and research. Overall, Table 2 shows that researchers from developing, transition, and developed economies all contribute to the research on the application of I4Es-in-SCM. Although unequal, there are contributions of countries in both the global south and north in the application of I4Es-in-SCM research, indicating a global interest in data-driven supply chain management.

Table 2.

Most productive countries in the application of I4Es-in-SCM research.

Country	Articles	Citations	Total Link Strength
China	251	2635	74
United States	119	4131	60
India	87	1347	20
United Kingdom	81	3006	52
Australia	33	1247	18
Hong Kong	29	1320	24
France	28	937	24
Indonesia	27	184	6
South Korea	27	893	8
Taiwan	26	711	12
Iran	24	287	7
Malaysia	23	116	12
Canada	22	939	10
Germany	21	444	9
Spain	17	167	8
Thailand	17	32	7
South Africa	14	121	3
Italy	12	157	7
Denmark	10	157	5
Singapore	10	234	8
Netherlands	9	448	3
Turkey	9	596	5
Brazil	8	55	1
Sweden	8	196	5
United Arab Emirates	8	187	4
Vietnam	8	62	5

4.3. Scientific collaborations

4.3.1. Co-authorship network analysis

The total link strength in Table 1 shows the collaborative strength of the 27 authors. However, the table did not show the collaborative landscape and network between and among the authors. Fig. 3 show the collaborative network of the authors. The different colour codes in Fig. 3 shows the clusters of active collaborations between and among the researchers. The size of a node represents the collaborative strength of a researcher within a cluster [12]. As VOSviewer produces distance-based networks and maps [40], the distance between nodes represents the frequency at which researchers collaborate within the cluster.

Fig. 3.

Co-authorship network in the application of I4Es-in-SCM research.

Fig. 3 shows five clusters of collaborative networks. The results show that authors such as Zhang Y., Liu Y., Wang H., and Zhang Z. frequently collaborate and co-author articles on the application of I4Es-in-SCM research. Researchers such as Zhang X., Zhang N., and Wang L. frequently collaborate in research publications on the application of I4Es-in-SCM research. Similarly, Wang Y. and Wang M. tend to collaborate frequently. Similarly, a collaboration network occurs between Liu X. and Huang G.Q. as well as Wang J. and Wang X. A compelling observation is that these authors all have Asian names. This does not however suggest that the collaboration is internal; it could be that these Asian authors work in different countries but collaborate frequently in publishing articles on the application of I4Es-in-SCM research. Thus, it is useful to explore the collaboration network of countries.

4.3.2. Network of countries

International collaborations facilitate knowledge exchange, innovation transfer, and cross-cultural research development [ [13,40]]. Thus, it is useful to explore the collaboration of countries in the application of I4Es-in-SCM research to provide information to other researchers looking to collaborate with countries keen on understanding the application of I4Es-in-SCM. Fig. 4 show the collaboration network of most productive and influential countries in the application of I4Es-in-SCM research. The different colour codes in Fig. 4 shows the clusters of active collaborations between countries. As VOSviewer generates size and distance-based maps and networks, the size of a node represents the collaborative strength of a country within a cluster and the distance between nodes represents the frequency at which countries collaborate within the cluster [12].

Fig. 4.

Collaboration network of countries in the application of I4Es-in-SCM research.

The colour schemes in Fig. 4 show seven clusters of collaboration networks between and among countries. Cluster #1 includes Belgium, Denmark, France, Germany, Ireland, Italy, Netherlands, Portugal, Romania, and Sweden. The collaboration network in Cluster #1 comprises only European economies and suggests a strong internal collaboration among researchers in the continent. Cluster #2 includes Canada, Indonesia, Japan, Malaysia, Singapore, South Korea, Thailand, United Arab Emirates, and Vietnam. From Canada, the rest are Asian countries. This also suggests a strong internal collaboration between and among researchers in Asia. Cluster #3 includes Australia, Finland, Iran, South Africa, and Turkey. The collaboration network of Cluster #3 includes countries from Australia, Europe, Africa and Asia. This is an intercontinental collaboration and highlights the wider interest in the application of I4Es-in-SCM research. Cluster #4 includes China, Hong Kong, Switzerland, and the United States. Cluster #4 also suggests that there is some degree of collaboration between researchers from Asia, Europe and North America.

Cluster #5 includes India, Pakistan, Russian Federation, and Saudi Arabia. Even though this network includes Russia, it suggests a collaboration between researchers within Asia and the Pacific because the Russian Federation is a transcontinental country located in Eastern Europe and Northern Asia. Cluster #6 includes Brazil, Egypt, Greece, and Spain. This network also suggests the existence of collaboration between researchers in South America, Europe, and Africa. Cluster #7 includes Taiwan, and the United Kingdom, indicating collaboration between researchers in Europe and Asia. The collaboration networks in Fig. 4 show that researchers in all continents collaborate with other researchers in other continents in the application of I4Es-in-SCM research. However, there are limited collaboration networks of researchers from Africa and South America. This may reflect the lower interest and commitment of the countries to big data and supply chain management research. The significant dominance of researchers from Asia, North America and Europe in the collaboration networks shows the increasing attempts to export innovations from these continents to others. However, the significant internal collaboration of researchers within the continents does not promote enough global collaboration, knowledge sharing and innovation diffusion, and should be improved.

4.4. Research impact analysis

4.4.1. Prominent research outlets

Researchers are usually keen to have full knowledge of the research outlets that publishes articles on a trending research area. Analysis of prominent research outlets for a given trending subject provides a relevant submission guide to future researchers [13]. As such, Table 3 shows the prominent journals that publish articles on the application of I4Es-in-SCM research. The information in Table 3 could assist institutions and libraries with limited resources to subscribe to key Journals that publish high-impact papers on the application of I4Es-in-SCM research.

Table 3.

Prominent Journals with articles on the application of I4Es-in-SCM research.

Research outlet	Articles	Citations	Total Link Strength
International Journal of Supply Chain Management	37	91	18
International Journal of Production Research	22	582	139
International Journal of Production Economics	19	1743	171
Computers and Industrial Engineering	18	714	94
Industrial Management and Data Systems	16	270	22
Journal of Cleaner Production	16	625	81
Sustainability (Switzerland)	16	175	49
International Journal of Logistics Management	14	348	102
Annals of Operations Research	13	234	67
IEEE Access	12	48	9
Journal of Advanced Oxidation Technologies	10	0	0
Transportation Research Part E: Logistics and Transportation Review	10	248	37
Computers and Operations Research	8	202	38
International Journal of Physical Distribution and Logistics Management	8	259	30
Chemical Engineering Transactions	7	15	0
International Journal of Logistics Research and Applications	7	111	36
Journal of Business Logistics	7	867	103
Neural Computing and Applications	7	50	4
Production Planning and Control	7	148	64
Agro Food Industry Hi-Tech	6	9	0
International Journal of Logistics Systems and Management	6	12	0
Journal of Enterprise Information Management	6	65	24

Of the 398 research outlets, the 21 Journals in Table 3 published at least 5 papers on the application of I4Es-in-SCM research [12]. Based on the number of articles on the application of I4Es-in-SCM research published between 1994 and 2020, the seven most productive research outlets include the International Journal of Supply Chain Management (37), International Journal of Production Research (22), International Journal of Production Economics (19), Computers and Industrial Engineering (18), Industrial Management and Data Systems (16), Journal of Cleaner Production (16), and Sustainability (Switzerland) (16). This indicates that these research outlets frequently publish articles on the application of I4Es-in-SCM. This information could guide authors when submitting their articles for publication. However, these research outlets are not necessarily publishing the most cited and referenced papers. Based on the total citations count in Table 3, the 5 top most high-impact research outlets include the International Journal of Production Economics (1743), the Journal of Business Logistics (867), the Computers and Industrial Engineering (714), the Journal of Cleaner Production (625), and the International Journal of Production Research (582). This suggests that articles published in these Journals are well-received by researchers and frequently referenced in the I4Es-in-SCM scientific community. Fig. 5 shows the co-citation network of the research outlets.

Fig. 5.

Co-citation network of e prominent research outlets.

The values of the total link strength in Table 3 represent the frequency at which the journal is cited with other prominent journals and reflect the sizes of nodes in Fig. 5. Based on the total link strength indices and the sizes of the nodes in Fig. 5, International Journal of Production Economics (171), International Journal of Production Research (139), Journal of Business Logistics (103), International Journal of Logistics Management (102), and Computers and Industrial Engineering (94) are the most co-cited research outlets in the application of I4Es-in-SCM. This suggests that the Journals publish articles whose findings complement each other. It also implies researchers who can publish articles in these journals are more likely to be frequently cited by papers submitted to the same sets of journals.

4.4.2. High-impact articles in the application of I4Es-in-SCM research

Citations are the most widely used measure of the impact of research output. It is useful to identify and analyse the most cited articles in each domain as they have the greatest influence and impact in the field [ [13,40]]. Based on total citation counts, Table 4 shows the top 10 landmark articles on the application of I4Es-in-SCM research. Table 4 shows three common metrics used to measure the impact of a paper – total citations, normalized citations [12], and field-weighted citation impact [40]. While total citation measures the citations accumulated by the paper since its publication, the normalized citations eliminate the impact of the year of publication on the total citation count. As shown in Table 4, even though the first publication emerged in 1994, the landmark articles were published between 2008 and 2017, suggesting that articles published during the last two decades have laid a strong foundation for improved understanding in the field.

Table 4.

Most cited articles in the application of I4Es-in-SCM research.

Authors	Title of document	Journal	Total Citations	FWCI	Norm. Citations
Waller and Fawcett (2013)	Data science, predictive analytics, and big data: A revolution that will transform supply chain design and management	Journal of Business Logistics	561	18.27*	14.45
Wang et al. (2016)	Big data analytics in logistics and supply chain management: Certain investigations for research and applications	International Journal of Production Economics	423	26.97*	9.92
Hazen et al. (2014)	Data quality for data science, predictive analytics, and big data in supply chain management: An introduction to the problem and suggestions for research and applications	International Journal of Production Economics	373	24.06*	10.80
Gunasekaran et al. (2017)	Big data and predictive analytics for supply chain and organizational performance	Journal of Business Research	259	20.75*	11.19
Zhong et al. (2015)	A big data approach for logistics trajectory discovery from RFID-enabled production data	International Journal of Production Economics	235	26.52*	5.82
Zhong et al. (2016)	Big Data for supply chain management in the service and manufacturing sectors: Challenges, opportunities, and future perspectives	Computers and Industrial Engineering	216	29.16*	5.06
Carbonneau et al. (2008)	Application of machine learning techniques for supply chain demand forecasting	European Journal of Operational Research	195	5.70*	2.18
Schoenherr and Speier-Pero (2015)	Data science, predictive analytics, and big data in supply chain management: Current state and future potential	Journal of Business Logistics	190	16.18*	4.71
Tan et al. (2015)	Harvesting big data to enhance supply chain innovation capabilities: An analytic infrastructure based on deduction graph	International Journal of Production Economics	182	21.06*	4.51
Papadopoulos et al. (2017)	The role of Big Data in explaining disaster resilience in supply chains for sustainability	Journal of Cleaner Production	179	15.11*	7.74

Note: *FWCI - Field-Weighted Citation Impact shows how well cited this document is when compared to similar documents. A value greater than 1.00 means the document is more cited than expected according to the average.

Although the articles are ordered in descending order of total citations, it is useful to highlight the impact based on the normalized citations and FWCI. Based on the normalized citations, the works of Ref. [45] have made a higher impact than the works of Ref. [9] and the work of Ref. [46] which were ranked higher based on total citations count. The FWCI in Table 4 represents the ratio of the document's citations to the average number of citations received by all similar documents over three years period. The FWCI considers the year of publication, document type, and disciplines associated with its sources. This metric is useful because these articles may have been frequently referenced by papers addressing broader I4Es or supply chain management issues and not necessarily the applications of I4Es-in-SCM. It offers the ability to compare ‘apples and apples’ based on internal field impact. Based on the FWCI values as shown in Table 4, the works of Ref. [9,47], and [48] have made the highest impact within the application of I4Es-in-SCM research.

Moreover, the majority of the landmark articles in Table 4 have addressed the predictive analytics of the application of I4Es-in-SCM research. This suggests that researchers have mainly focused on predictive analytics, with minimal attention to descriptive, prescriptive, and cognitive analytics. Table 4 shows that the landmark articles were published in landmark research outlets, including the International Journal of Production Economics (4), Journal of Business Logistics (2), Journal of Business Research (1), the Journal of Cleaner Production (1), the European Journal of Operational Research (1), and the Computers and Industrial Engineering (1). The citation counts of the topmost cited journals significantly contribute to the higher impacts of the International Journal of Production Economics and the Journal of Business Logistics.

4.5. Main research areas in the application of I4Es-in-SCM

According to Ref. [40], keywords reflect the main contents of the papers and denote the trending topics and themes in the given field of research. In the context of science mapping, analysis of the co-occurrence network of keywords can provide a precise mental map of the trending topics and themes in the research field [12]. A network of keywords provides supply chain researchers and academics with a snapshot and panoramic view of frequently studied themes and existing research methodologies employed and offers a sound basis to delineate boundaries of existing scientific knowledge. This approach offers a structured and objective methodology for drawing valid conclusions. Effective analysis of all keywords within the 786 research items could provide a basis to identify the main research areas and hot topics in the application of I4Es-in-SCM research. Of the total of 5003 indexed keywords, Table 5 shows the 30 most frequently cited keywords in the application of I4Es-in-SCM research.

Table 5.

Trending topics and themes in the application of I4Es-in-SCM research.

Keyword	Occurrences	Total Link Strength
supply chain management	198	183
supply chains	161	158
big data	160	141
The internet of things	158	140
supply chain	80	71
logistics	75	69
machine learning	70	63
decision making	61	61
artificial intelligence	42	40
information management	39	39
data analytics	33	33
risk assessment	33	33
big data analytics	31	28
logistic regression	29	24
sustainability	29	26
forecasting	27	27
sustainable development	27	26
learning systems	26	26
logistic models	26	26
manufacture	26	26
radio frequency identification (RFID)	26	26
regression analysis	26	25
SMEs	26	21
statistical model	25	25
electronic commerce	22	22
sales	22	22
industry 4.0	21	20
information technology	21	20
optimization	21	21
logistic regression analysis	20	20

As shown in Table 5, big data is used in supply chain management to inform decision-making, optimization, risk assessment, sales, sustainability management, forecasting, and performance evaluation, among others. The keywords shown in Table 5 reflect the most frequently researched topics and themes in the application of I4Es-in-SCM research. As such, a holistic analysis of the keywords could assist to identify the main research areas. A network analysis of the most cited keywords is useful in identifying the clusters of research themes and trending topics [12]. The different colour schemes shown in Fig. 6 denote the different clusters of the trending topics and themes in the application of I4Es-in-SCM research. Each node in Fig. 6 represents a keyword, and its connections reflects how many times the keywords co-occurred with other keywords in the included studies. The distance between two nodes denotes the frequency at which the keywords have been cited together [12].

Fig. 6.

Co-occurrence network of keywords in the application of I4Es-in-SCM research.

As shown in Fig. 6, there are six clusters of keywords. Cluster #1 includes terms such as analytics, big data, collaboration, data analytics, data science, integration, logistics, performance, predictive analytics, radio frequency identification, supply chain management, SME, and supply chain. Cluster #2 includes terms such as analytic hierarchy process, artificial intelligence, bullwhip effect, classification, demand forecasting, genetic algorithm, logistic regression, machine learning, neural networks, performance evaluation, risk management, and sentiment analysis. From Fig. 6, Cluster #3 includes terms such as adoption, challenges, data-driven, information technology, inventory management, production logistics, simulation, small to medium-sized enterprises, supply chain finance, supply chain network, supply chains, and technology.

Cluster #4 includes terms such as e-commerce, green supply chain management, information sharing, innovation, logistics management, reverse logistics, small and medium-enterprise, SMEs, supply chain performance, sustainability, and sustainable supply chain. Cluster #5 includes terms such as big data analytics, organizational performance, resilience, resource-based view, small and medium-sized enterprises, social media, supply chain disruption, supply chain integration, supply chain management, and supply chain risk management. Cluster #6 includes keywords such as blockchain, cloud computing, deep learning, food supply chain, industry 4.0, internet of things, internet of things (IoT), smart logistics, supply chain management, and traceability.

• (a)The keywords show that big data is frequently leveraged for performance evaluation, predictive analytics, sentiment analysis, demand forecasting, risk management, supply chain resilience and disruption assessment, logistics management, organizational performance analysis, inventory management, and supply chain traceability. This highlights the wider application of I4Es-in-SCM research and promises more informed and data-driven supply chain management in the future.
• (b)Existing research has frequently used methodologies such as analytic hierarchy process, regression analysis, classification, machine learning, genetic algorithm, artificial neural networks, simulation, and deep learning. As expected, different methods of artificial intelligence have been deployed to accurately analysed the widely available data in the supply chain environment
• (c)The keywords also highlight different supply chain industries such as e-commerce, food, and small and medium-sized enterprises, among others leveraging big data to inform supply chain management.
• (d)Existing research has also leveraged blockchain technology, platforms (e.g. social media) and digital technologies such as radio frequency identification, the internet of things, cloud computing, and smart logistics objects to derive big data for the supply chain management.
• (e)The extant literature has also considered several aspects of supply chain management such as logistics management, sustainable supply chain management, supply chain finance, and production logistics.

4.6. Gaps and areas for further research

This research lays the foundations for future research since it discussed the importance of industry 4.0 enablers and deeply elaborated on the areas of potential collaboration among researchers, institutions and countries. For example, the study provides clear guidance for early-stage researchers on the significant works, trending topics and researchers in this area so they can collaborate with for future studies.

Having considered applications I4Es-in-SCM, it is also reasonable to look at their implementations and employment within the different stages of supply chains namely, plan, source, make, deliver and return as proposed by the Supply Chain Operations Reference (SCOR) model.

Another fruitful research avenue could be supply chain uncertainty and how enablers such as BD, AI, IoT and ML can help reduce perceived risks and support in supply chain indecision situations. Information has always been considered as key supply chain driver allowing the supply chain to become more efficient and effective, however, the extent to which these information-driven enablers are currently employed to attenuate supply chain uncertainty requires further research and investigation.

Robotic systems and the value they are currently adding to supply chain operations can be a fertile area of research, especially with the mounting dependence of many manufacturers on the capability of both humanoid and nonhumanoid robots. Organizations such as Amazon depend on thousands of “drive” robots to help fulfil customers’ requests within their fulfilment centres while some restaurants in Japan are completely managed and operated by humanoid robots. Therefore, it would be intriguing to measure the effectiveness of implementing and using these robots within supply chains.

4.7. Implications

The paper's findings have several important theoretical and practical implications for the field of supply chain management (SCM) and the application of Industry 4.0 technologies (I4Es). Firstly, the analysis of trends and the significant growth in I4Es-in-SCM research over the last two decades indicate a growing recognition of the importance of integrating advanced technologies into supply chain operations. This suggests a paradigm shift in how supply chains are managed and optimized, emphasizing the potential benefits of incorporating Industry 4.0 concepts like big data, predictive analytics, and IoT into SCM practices.

Secondly, the identification of the most productive and influential researchers and research outlets provides valuable insights into the academic landscape of I4Es-in-SCM. Understanding which researchers have made significant contributions and which journals attract the most citations underscores the importance of collaboration and dissemination strategies in driving the advancement of knowledge in this area.

Moreover, the geospatial distribution and network analysis revealing global participation and collaborations among researchers from diverse economies highlight the global significance of I4Es-in-SCM research. This finding suggests that Industry 4.0 technologies are not confined to developed economies but are also gaining traction in developing and transition countries. This has implications for the adoption and implementation of I4Es in different global contexts, potentially leading to a more inclusive and comprehensive understanding of SCM practices worldwide.

The paper's analysis of trending themes and hot topics offers valuable insights into the direction of future research in I4Es-in-SCM. By identifying areas with the highest research interest, scholars and practitioners can focus their efforts on emerging challenges and opportunities in applying Industry 4.0 technologies to supply chain management. This can lead to the development of innovative solutions and frameworks that address current and anticipated challenges faced by supply chain professionals.

Overall, the theoretical and practical implications of this paper highlight the evolving landscape of supply chain management in the era of Industry 4.0. The growing interest, collaborations, and global involvement indicate the importance of embracing technological advancements to enhance supply chain efficiency, sustainability, and resilience. The identified gaps and trending themes provide a roadmap for future research, guiding scholars towards the areas that can have the most significant impact on improving supply chain performance and competitiveness in the digital age.

5. Conclusions

This paper conducted a scientometric analysis and critical review of the literature on the application of I4Es-in-SCM between 1994 and 2020 to monitor trends, visualize the structure of knowledge, identify gaps, and highlight future research areas. The study recruited bibliographic records of 786 studies from Scopus on the application of I4Es-in-SCM and analysed the datasets using several science mapping techniques. Analysis showed that the last two decades witnessed a phenomenal growth in research on the application of I4Es-in-SCM. Significant growth is witnessed since 2008 and reached the highest ever in the year 2020 when at least 230 articles were published on the application of I4Es-in-SCM research.

Of 2040 researchers, the analysis showed that the top five most productive researchers include Gunasekaran A. (18), Wang Y. (11), Zhang Y. (11), Dubey R. (10), Liu P. (9) whereas the top five most influential researchers in terms of total citations in the application of I4Es-in-SCM research include Gunasekaran A. (1407), Papadopoulos T. (990), Huang G.Q. (779), Dubey R. (778), and Hazen B.T. (587). These researchers may constitute priority academics when considering future research and grant applications on I4Es-in-SCM. Further analysis showed that the 5 topmost cited and influential research outlets in the application of I4Es-in-SCM research include the International Journal of Production Economics (1743), the Journal of Business Logistics (867), the Computers and Industrial Engineering (714), the Journal of Cleaner Production (625), and International Journal of Production Research (582). I4Es-in-SCM research articles published in these journals have been well-received and frequently referenced in new studies. As such, publishing I4Es-in-SCM research works in these outlets could influence the impact of the works.

A geospatial distribution and network analysis showed that at least 82 (42 %) of the global 193 countries are involved in the application of I4Es-in-SCM research. It is found that the interest and contributions are global since there is participation of researchers from developing, transition, and developed economies in both the global north and global south. In terms of scientific productivity, the top five most productive countries in the application of I4Es-in-SCM research include China, the United States, India, the United Kingdom, and Australia. The study found seven clusters of international collaborations but noticed significant collaboration among researchers within each continent and suggesting the need for intercontinental collaborations in future research. Additionally, there are limited contributions and collaborations among researchers from Africa and South America, which could be due to minimal interest in the application of I4Es-in-SCM research in both regions.

The study further revealed trending themes and hot topics in the application of I4Es-in-SCM research, delineated boundaries of the existing scientific knowledge and proffered areas that require further research. The novelty of this study lies in providing a more holistic statistical and visualized analysis of the structure of knowledge, productivity, and scientific collaborations of researchers, journals and countries in the application of I4Es-in-SCM management research. The study outcomes may serve as a useful reference to supply chain academics, early-stage researchers, practitioners, policymakers, and organizations in understanding the structure of knowledge on the application of I4Es-in-SCM research and may constitute a basis for future research. Despite the realization of the study's aim, some limitations are noteworthy. First, the scientometric analysis provides an overview of the structure of the knowledge and does not provide a detailed evaluation of each included study. Effectively, the authors did not manually review each of the 786 articles and hence, some critical considerations may have been missed in the conclusions. Second, the study did not provide in-depth discussions of the trending themes and topics due to space constraints. Nevertheless, these limitations do not override the validity and relevance of scientometric reviews as they are more objective and quantitative than manual reviews.

Data availability statement

No data was used for the research described in the article.

CRediT authorship contribution statement

Hassan Younis: Conceptualization, Funding acquisition, Project administration, Supervision, Writing – review & editing. Ibrahim Yahaya Wuni: Software, Visualization, Writing – original draft.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.