https://doi.org/10.1086/258100
+1 323-425-8868 | |
+86 18163351462(WhatsApp) | |
Paper Publishing WeChat |
Copyright © 2024 by authors and Scientific Research Publishing Inc.
This work and the related PDF file are licensed under a Creative Commons Attribution 4.0 International License .
You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.
All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .
Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.
Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.
Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.
Original Submission Date Received: .
Find support for a specific problem in the support section of our website.
Please let us know what you think of our products and services.
Visit our dedicated information section to learn more about MDPI.
Analysing near-miss incidents in construction: a systematic literature review.
3. research methodology, 4.1. a statistical analysis of publications, 4.2. methods used to obtain information about near misses, 4.2.1. traditional methods.
4.3.1. quantitative and qualitative statistical methods, 4.3.2. analysis using artificial intelligence (ai), 4.3.3. building information modelling, 4.4. key aspects of near-miss investigations in the construction industry, 4.4.1. occupational risk assessment, 4.4.2. causes of hazards in construction, 4.4.3. time series of near misses, 4.4.4. material factors of construction processes, 4.5. a comprehensive overview of the research questions and references on near misses in the construction industry, 5. discussion, 5.1. interest of researchers in near misses in construction (question 1), 5.2. methods used to obtain near-miss information (question 2), 5.3. methods used to analyse the information and data sets (question 3), 5.4. key aspects of near-miss investigations in the construction industry (question 4), 6. conclusions.
Institutional review board statement, informed consent statement, data availability statement, conflicts of interest.
Year | Source Title | DOI/ISBN/ISSN | Reference |
---|---|---|---|
1999 | Construction Management and Economics | 10.1080/014461999371691 | [ ] |
2002 | Structural Engineer | 14665123 | [ ] |
2009 | Building a Sustainable Future—Proceedings of the 2009 Construction Research Congress | 10.1061/41020(339)4 | [ ] |
2010 | Safety Science | 10.1016/j.ssci.2010.04.009 | [ ] |
2010 | Automation in Construction | 10.1016/j.autcon.2009.11.017 | [ ] |
2010 | Safety Science | 10.1016/j.ssci.2009.06.006 | [ ] |
2012 | Journal of Construction Engineering and Management | 10.1061/(ASCE)CO.1943-7862.0000518 | [ ] |
2013 | ISARC 2013—30th International Symposium on Automation and Robotics in Construction and Mining, Held in Conjunction with the 23rd World Mining Congress | 10.22260/isarc2013/0113 | [ ] |
2014 | Proceedings of the Institution of Civil Engineers: Civil Engineering | 10.1680/cien.14.00010 | [ ] |
2014 | Safety Science | 10.1016/j.ssci.2013.12.012 | [ ] |
2014 | Journal of Construction Engineering and Management | 10.1061/(ASCE)CO.1943-7862.0000795 | [ ] |
2014 | 31st International Symposium on Automation and Robotics in Construction and Mining, ISARC 2014—Proceedings | 10.22260/isarc2014/0115 | [ ] |
2014 | Construction Research Congress 2014: Construction in a Global Network—Proceedings of the 2014 Construction Research Congress | 10.1061/9780784413517.0181 | [ ] |
2014 | Construction Research Congress 2014: Construction in a Global Network—Proceedings of the 2014 Construction Research Congress | 10.1061/9780784413517.0235 | [ ] |
2014 | Construction Research Congress 2014: Construction in a Global Network—Proceedings of the 2014 Construction Research Congress | 10.1061/9780784413517.0096 | [ ] |
2015 | Automation in Construction | 10.1016/j.autcon.2015.09.003 | [ ] |
2015 | 32nd International Symposium on Automation and Robotics in Construction and Mining: Connected to the Future, Proceedings | 10.22260/isarc2015/0062 | [ ] |
2015 | ASSE Professional Development Conference and Exposition 2015 | - | [ ] |
2015 | Congress on Computing in Civil Engineering, Proceedings | 10.1061/9780784479247.019 | [ ] |
2016 | Automation in Construction | 10.1016/j.autcon.2016.03.008 | [ ] |
2016 | Automation in Construction | 10.1016/j.autcon.2016.04.007 | [ ] |
2016 | IEEE IAS Electrical Safety Workshop | 10.1109/ESW.2016.7499701 | [ ] |
2016 | Journal of Construction Engineering and Management | 10.1061/(ASCE)CO.1943-7862.0001100 | [ ] |
2016 | Safety Science | 10.1016/j.ssci.2015.11.025 | [ ] |
2016 | Journal of Construction Engineering and Management | 10.1061/(ASCE)CO.1943-7862.0001049 | [ ] |
2016 | IEEE Transactions on Industry Applications | 10.1109/TIA.2015.2461180 | [ ] |
2017 | Safety Science | 10.1016/j.ssci.2017.06.012 | [ ] |
2017 | ENR (Engineering News-Record) | 8919526 | [ ] |
2017 | 6th CSCE-CRC International Construction Specialty Conference 2017—Held as Part of the Canadian Society for Civil Engineering Annual Conference and General Meeting 2017 | 978-151087841-9 | [ ] |
2017 | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | 10.1007/978-3-319-72323-5_12 | [ ] |
2017 | Journal of Construction Engineering and Management | 10.1061/(ASCE)CO.1943-7862.0001209 | [ ] |
2017 | Safety Science | 10.1016/j.ssci.2016.08.027 | [ ] |
2017 | Safety Science | 10.1016/j.ssci.2016.08.022 | [ ] |
2018 | Safety Science | 10.1016/j.ssci.2018.04.004 | [ ] |
2018 | International Journal of Construction Management | 10.1080/15623599.2017.1382067 | [ ] |
2018 | Journal of Construction Engineering and Management | 10.1061/(ASCE)CO.1943-7862.0001420 | [ ] |
2018 | Proceedings of SPIE—The International Society for Optical Engineering | 10.1117/12.2296548 | [ ] |
2019 | Automation in Construction | 10.1016/j.autcon.2019.102854 | [ ] |
2019 | Physica A: Statistical Mechanics and its Applications | 10.1016/j.physa.2019.121495 | [ ] |
2019 | Sustainability (Switzerland) | 10.3390/su11051264 | [ ] |
2019 | Computing in Civil Engineering 2019: Data, Sensing, and Analytics—Selected Papers from the ASCE International Conference on Computing in Civil Engineering 2019 | 978-078448243-8 | [ ] |
2019 | Journal of Health, Safety and Environment | 18379362 | [ ] |
2019 | Computing in Civil Engineering 2019: Data, Sensing, and Analytics—Selected Papers from the ASCE International Conference on Computing in Civil Engineering 2019 | 978-078448243-8 | [ ] |
2019 | Computing in Civil Engineering 2019: Smart Cities, Sustainability, and Resilience—Selected Papers from the ASCE International Conference on Computing in Civil Engineering 2019 | 10.1061/9780784482445.026 | [ ] |
2019 | Journal of Construction Engineering and Management | 10.1061/(ASCE)CO.1943-7862.0001582 | [ ] |
2019 | Advances in Intelligent Systems and Computing | 10.1007/978-3-030-02053-8_107 | [ ] |
2020 | Accident Analysis and Prevention | 10.1016/j.aap.2020.105496 | [ ] |
2020 | Advanced Engineering Informatics | 10.1016/j.aei.2020.101062 | [ ] |
2020 | Advanced Engineering Informatics | 10.1016/j.aei.2020.101060 | [ ] |
2020 | ARCOM 2020—Association of Researchers in Construction Management, 36th Annual Conference 2020—Proceedings | 978-099554633-2 | [ ] |
2020 | International Journal of Building Pathology and Adaptation | 10.1108/IJBPA-03-2020-0018 | [ ] |
2020 | Communications in Computer and Information Science | 10.1007/978-3-030-42852-5_8 | [ ] |
2021 | Journal of Architectural Engineering | 10.1061/(ASCE)AE.1943-5568.0000501 | [ ] |
2021 | Safety Science | 10.1016/j.ssci.2021.105368 | [ ] |
2021 | ACM International Conference Proceeding Series | 10.1145/3482632.3487473 | [ ] |
2021 | Reliability Engineering and System Safety | 10.1016/j.ress.2021.107687 | [ ] |
2021 | Proceedings of the 37th Annual ARCOM Conference, ARCOM 2021 | - | [ ] |
2022 | Buildings | 10.3390/buildings12111855 | [ ] |
2022 | Safety Science | 10.1016/j.ssci.2022.105704 | [ ] |
2022 | Sensors | 10.3390/s22093482 | [ ] |
2022 | Proceedings of International Structural Engineering and Construction | 10.14455/ISEC.2022.9(2).CSA-03 | [ ] |
2022 | Journal of Information Technology in Construction | 10.36680/j.itcon.2022.045 | [ ] |
2022 | Forensic Engineering 2022: Elevating Forensic Engineering—Selected Papers from the 9th Congress on Forensic Engineering | 10.1061/9780784484555.005 | [ ] |
2022 | Computational Intelligence and Neuroscience | 10.1155/2022/4851615 | [ ] |
2022 | International Journal of Construction Management | 10.1080/15623599.2020.1839704 | [ ] |
2023 | Journal of Construction Engineering and Management | 10.1061/JCEMD4.COENG-13979 | [ ] |
2023 | Heliyon | 10.1016/j.heliyon.2023.e21607 | [ ] |
2023 | Accident Analysis and Prevention | 10.1016/j.aap.2023.107224 | [ ] |
2023 | Safety | 10.3390/safety9030047 | [ ] |
2023 | Engineering, Construction and Architectural Management | 10.1108/ECAM-09-2021-0797 | [ ] |
2023 | Advanced Engineering Informatics | 10.1016/j.aei.2023.101929 | [ ] |
2023 | Engineering, Construction and Architectural Management | 10.1108/ECAM-05-2023-0458 | [ ] |
2023 | Intelligent Automation and Soft Computing | 10.32604/iasc.2023.031359 | [ ] |
2023 | International Journal of Construction Management | 10.1080/15623599.2020.1847405 | [ ] |
2024 | Heliyon | 10.1016/j.heliyon.2024.e26410 | [ ] |
Click here to enlarge figure
No. | Name of Institution/Organization | Definition |
---|---|---|
1 | Occupational Safety and Health Administration (OSHA) [ ] | “A near-miss is a potential hazard or incident in which no property was damaged and no personal injury was sustained, but where, given a slight shift in time or position, damage or injury easily could have occurred. Near misses also may be referred to as close calls, near accidents, or injury-free events.” |
2 | International Labour Organization (ILO) [ ] | “An event, not necessarily defined under national laws and regulations, that could have caused harm to persons at work or to the public, e.g., a brick that falls off scaffolding but does not hit anyone” |
3 | American National Safety Council (NSC) [ ] | “A Near Miss is an unplanned event that did not result in injury, illness, or damage—but had the potential to do so” |
4 | PN-ISO 45001:2018-06 [ ] | A near-miss incident is described as an event that does not result in injury or health issues. |
5 | PN-N-18001:2004 [ ] | A near-miss incident is an accident event without injury. |
6 | World Health Organization (WHO) [ ] | Near misses have been defined as a serious error that has the potential to cause harm but are not due to chance or interception. |
7 | International Atomic Energy Agency (IAEA) [ ] | Near misses have been defined as potentially significant events that could have consequences but did not due to the conditions at the time. |
No. | Journal | Number of Publications |
---|---|---|
1 | Safety Science | 10 |
2 | Journal of Construction Engineering and Management | 8 |
3 | Automation in Construction | 5 |
4 | Advanced Engineering Informatics | 3 |
5 | Construction Research Congress 2014 Construction in a Global Network Proceedings of the 2014 Construction Research Congress | 3 |
6 | International Journal of Construction Management | 3 |
7 | Accident Analysis and Prevention | 2 |
8 | Computing in Civil Engineering 2019 Data Sensing and Analytics Selected Papers From The ASCE International Conference | 2 |
9 | Engineering Construction and Architectural Management | 2 |
10 | Heliyon | 2 |
Cluster Number | Colour | Basic Keywords |
---|---|---|
1 | blue | construction, construction sites, decision making, machine learning, near misses, neural networks, project management, safety, workers |
2 | green | building industry, construction industry, construction projects, construction work, human, near miss, near misses, occupational accident, occupational safety, safety, management, safety performance |
3 | red | accident prevention, construction equipment, construction, safety, construction workers, hazards, human resource management, leading indicators, machinery, occupational risks, risk management, safety engineering |
4 | yellow | accidents, risk assessment, civil engineering, near miss, surveys |
Number of Question | Question | References |
---|---|---|
Q | Are near misses in the construction industry studied scientifically? | [ , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ] |
Q | What methods have been used to obtain information on near misses and systems for recording incidents in construction companies? | [ , , , , , , , , , , , , , , , , , , , , ] |
Q | What methods have been used to analyse the information and figures that have been obtained? | [ , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ] |
Q | What are the key aspects of near misses in the construction industry that have been of interest to the researchers? | [ , , , , , , , , , , , , ] |
The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
Woźniak, Z.; Hoła, B. Analysing Near-Miss Incidents in Construction: A Systematic Literature Review. Appl. Sci. 2024 , 14 , 7260. https://doi.org/10.3390/app14167260
Woźniak Z, Hoła B. Analysing Near-Miss Incidents in Construction: A Systematic Literature Review. Applied Sciences . 2024; 14(16):7260. https://doi.org/10.3390/app14167260
Woźniak, Zuzanna, and Bożena Hoła. 2024. "Analysing Near-Miss Incidents in Construction: A Systematic Literature Review" Applied Sciences 14, no. 16: 7260. https://doi.org/10.3390/app14167260
Article access statistics, further information, mdpi initiatives, follow mdpi.
Subscribe to receive issue release notifications and newsletters from MDPI journals
You have full access to this open access article
Our paper analyzes the current state of research on artificial intelligence (AI) adoption from a production perspective. We represent a holistic view on the topic which is necessary to get a first understanding of AI in a production-context and to build a comprehensive view on the different dimensions as well as factors influencing its adoption. We review the scientific literature published between 2010 and May 2024 to analyze the current state of research on AI in production. Following a systematic approach to select relevant studies, our literature review is based on a sample of articles that contribute to production-specific AI adoption. Our results reveal that the topic has been emerging within the last years and that AI adoption research in production is to date still in an early stage. We are able to systematize and explain 35 factors with a significant role for AI adoption in production and classify the results in a framework. Based on the factor analysis, we establish a future research agenda that serves as a basis for future research and addresses open questions. Our paper provides an overview of the current state of the research on the adoption of AI in a production-specific context, which forms a basis for further studies as well as a starting point for a better understanding of the implementation of AI in practice.
Avoid common mistakes on your manuscript.
The technological change resulting from deep digitisation and the increasing use of digital technologies has reached and transformed many sectors [ 1 ]. In manufacturing, the development of a new industrial age, characterized by extensive automation and digitisation of processes [ 2 ], is changing the sector’s ‘technological reality’ [ 3 ] by integrating a wide range of information and communication technologies (such as Industry 4.0-related technologies) into production processes [ 4 ].
Although the evolution of AI traces back to the year 1956 (as part of the Dartmouth Conference) [ 5 ], its development has progressed rapidly, especially since the 2010s [ 6 ]. Driven by improvements, such as the fast and low-cost development of smart hardware, the enhancement of algorithms as well as the capability to manage big data [ 7 ], there is an increasing number of AI applications available for implementation today [ 8 ]. The integration of AI into production processes promises to boost the productivity, efficiency as well as automation of processes [ 9 ], but is currently still in its infancy [ 10 ] and manufacturing firms seem to still be hesitant to adopt AI in a production-context. This appears to be driven by the high complexity of AI combined with the lack of practical knowledge about its implementation in production and several other influencing factors [ 11 , 12 ].
In the literature, many contributions analyze AI from a technological perspective, mainly addressing underlying models, algorithms, and developments of AI tools. Various authors characterise both machine learning and deep learning as key technologies of AI [ 8 , 13 ], which are often applied in combination with other AI technologies, such as natural language recognition. While promising areas for AI application already exist in various domains such as marketing [ 14 ], procurement [ 15 ], supply chain management [ 16 ] or innovation management [ 17 ], the integration of AI into production processes also provides significant performance potentials, particularly in the areas of maintenance [ 18 ], quality control [ 19 ] and production planning and management [ 20 ]. However, AI adoption requires important technological foundations, such as the provision of data and the necessary infrastructure, which must be ensured [ 11 , 12 , 21 ]. Although the state of the art literature provides important insights into possible fields of application of AI in production, the question remains: To what extent are these versatile applications already in use and what is required for their successful adoption?
Besides the technology perspective of AI, a more human-oriented field of discussion is debated in scientific literature [ 22 ]. While new technologies play an essential role in driving business growth in the digital transformation of the production industry, the increasing interaction between humans and intelligent machines (also referred to as ‘augmentation’) creates stress challenges [ 23 ] and impacts work [ 24 ], which thus creates managerial challenges in organizations [ 25 , 26 ]. One of the widely discussed topics in this context is the fear of AI threatening jobs (including production jobs), which was triggered by e.g. a study of Frey, Osborne [ 27 ]. Another issue associated to the fear of machines replacing humans is the lack of acceptance resulting from the mistrust of technologies [ 28 , 29 ]. This can also be linked to the various ethical challenges involved in working with AI [ 22 ]. This perspective, which focuses on the interplay between AI and humans [ 30 ], reveals the tension triggered by AI. Although this is discussed from different angles, the question remains how these aspects influence the adoption of AI in production.
Another thematic stream of current literature can be observed in a series of contributions on the organizational aspects of the technology. In comparison to the two research areas discussed above, the number of publications in this area seems to be smaller. This perspective focuses on issues to implement AI, such as the importance of a profound management structure [ 31 , 32 ], leadership [ 33 ], implications on the organizational culture [ 34 ] as well as the need for digital capabilities and special organizational skills [ 33 ]. Although some studies on the general adoption of AI without a sectoral focus have already been conducted (such as by Chen, Tajdini [ 35 ] or Kinkel, Baumgartner, Cherubini [ 36 ]) and hence, some initial factors influencing the adoption of AI can be derived, the contributions from this perspective are still scarce, are usually not specifically analyzed in the context of production or lack a comprehensive view on the organization in AI adoption.
While non-industry specific AI issues have been researched in recent years, the current literature misses a production-specific analysis of AI adoption, providing an understanding of the possibilities and issues related to integrating AI into the production context. Moreover, the existing literature tells us little about relevant mechanisms and factors underlying the adoption of AI in production processes, which include both technical, human-centered as well as organizational issues. As organizational understanding of AI in a business context is currently still in its early stages, it is difficult to find an aggregate view on the factors that can support companies in implementing AI initiatives in production [ 37 , 38 ]. Addressing this gap, we aim to systematise the current scientific knowledge on AI adoption, with a focus on production. By drawing on a systematic literature review (SLR), we examine existing studies on AI adoption in production and explore the main issues regarding adoption that are covered in the analyzed articles. Building on these findings, we conduct a comprehensive analysis of the existing studies with the aim of systematically investigating the key factors influencing the adoption of AI in production. This systematic approach paves the way for the formulation of a future research agenda.
Our SLR addresses three research questions (RQs). RQ1: What are the statistical characteristics of existing research on AI adoption in production? To answer this RQ, we conduct descriptive statistics of the analyzed studies and provide information on time trends, methods used in the research, and country specifications. RQ2: What factors influence the adoption of AI in production? RQ2 specifies the adoption factors and forms the core component of our analysis. By adoption factors, we mean the factors that influence the use of AI in production (both positively and negatively) and that must therefore be analyzed and taken into account. RQ3: What research topics are of importance to advance the research field of AI adoption in production? We address this RQ by using the analyzed literature as well as the key factors of AI adoption as a starting point to derive RQs that are not addressed and thus provide an outlook on the topic.
In order to create a sound information base for both policy makers and practitioners on the topic of AI adoption in production, this paper follows the systematic approach of a SLR. For many fields, including management research, a SLR is an important tool to capture the diversity of existing knowledge on a specific topic for a scientific investigation [ 39 ]. The investigator often pursues multiple goals, such as capturing and assessing the existing environment and advancing the existing body of knowledge with a proprietary RQ [ 39 ] or identifying key research topics [ 40 ].
Our SLR aims to select, analyze, and synthesize findings from the existing literature on AI adoption in production over the past 24 years. In order to identify relevant data for our literature synthesis, we follow the systematic approach of the Preferred Reporting Items for Systematic reviews (PRISMA) [ 41 ]. In evaluating the findings, we draw on a mixed-methods approach, combining some quantitative analyses, especially on the descriptive aspects of the selected publications, as well as qualitative analyses aimed at evaluating and comparing the contents of the papers. Figure 1 graphically summarizes the methodological approach that guides the content of the following sub-chapters.
Methodical procedure of our SLR following PRISMA [ 41 ]
Following the development of the specific RQs, we searched for suitable publications. To locate relevant studies, we chose to conduct a publication analysis in the databases Scopus, Web of Science and ScienceDirect as these databases primarily contain international scientific articles and provide a broad overview of the interdisciplinary research field and its findings. To align the search with the RQs [ 42 ], we applied predefined key words to search the titles, abstracts, and keywords of Scopus, Web of Science and ScienceDirect articles. Our research team conducted several pre-tests to determine the final search commands for which the test results were on target and increased the efficiency of the search [ 42 ]. Using the combination of Boolean operators, we covered the three topics of AI, production, and adoption by searching combinations of ‘Artificial Intelligence’ AND ‘production or manufacturing’ AND ‘adopt*’ in the three scientific databases. Although ‘manufacturing’ tends to stand for the whole sector and ‘production’ refers to the process, the two terms are often used to describe the same context. We also follow the view of Burbidge, Falster, Riis, Svendsen [ 43 ] and use the terms synonymously in this paper and therefore also include both terms as keywords in the study location as well as in the analysis.
AI research has been credited with a resurgence since 2010 [ 6 ], which is the reason for our choice of time horizon. Due to the increase in publications within the last years, we selected articles published online from 2010 to May 8, 2024 for our analysis. As document types, we included conference papers, articles, reviews, book chapters, conference reviews as well as books, focusing exclusively on contributions in English in the final publication stage. The result of the study location is a list of 3,833 documents whose titles, abstracts, and keywords meet the search criteria and are therefore included in the next step of the analysis.
For these 3,833 documents, we then conducted an abstract analysis, ‘us[ing] a set of explicit selection criteria to assess the relevance of each study found to see if it actually does address the research question’ [ 42 ]. For this step, we again conducted double-blind screenings (including a minimum of two reviewers) as pilot searches so that all reviewers have the same understanding of the decision rules and make equal decisions regarding their inclusion for further analysis.
To ensure the paper’s focus on all three topics regarded in our research (AI, production, and adoption), we followed clearly defined rules of inclusion and exclusion that all reviewers had to follow in the review process. As a first requirement for inclusion, AI must be the technology in focus that is analysed in the publication. If AI was only mentioned and not further specified, we excluded the publication. With a second requirement, we checked the papers for the context of analysis, which in our case must be production. If the core focus is beyond production, the publication was also excluded from further analysis. The third prerequisite for further consideration of the publication is the analysis of the adoption of a technology in the paper. If technology adoption is not addressed or adoption factors are not considered, we excluded the paper. An article was only selected for full-text analysis if, after analyzing the titles, abstracts, and keywords, a clear focus on all three research areas was visible and the inclusion criteria were met for all three contexts.
By using this tripartite inclusion analysis, we were able to analyse the publications in a structured way and to reduce the 3,833 selected documents in our double-blind approach to 300 articles that were chosen for the full-text analysis. In the process of finding full versions of these publications, we had to exclude three papers as we could not access them. For the rest of the 297 articles we obtained full access and thus included them for further analysis. After a thorough examination of the full texts, we again had to exclude 249 publications because they did not meet our content-related inclusion criteria mentioned above, although the abstract analysis gave indications that they did. As a result, we finally obtained 47 selected papers on which we base the literature analysis and synthesis (see Fig. 1 ).
Figure 2 summarises the results of the descriptive analysis on the selected literature regarding AI adoption in production that we analyse in our SLR. From Fig. 2 a), which illustrates annual publication trends (2010–2024), the increase in publications on AI adoption in production over the past 5 years is evident, yet slightly declining after a peak in 2022. After a steady increase until 2022, in which 11 articles are included in the final analysis, 2023 features ten articles, followed by three articles for 2024 until the cut-off date in May 2024. Of the 47 papers identified through our search, the majority (n = 33) are peer-reviewed journal articles and the remaining thirteen contributions conference proceedings and one book chapter (see Fig. 2 b)).
Descriptive analyses of the selected articles addressing AI adoption in production
The identified contributions reveal some additional characteristics in terms of the authors country base (Fig. 2 c)) and research methods used (Fig. 2 d)). Almost four out of ten of the publications were written in collaboration with authors from several countries (n = 19). Six of the papers were published by authors from the United States, five from Germany and four from India. In terms of the applied research methods used by the researchers, a wide range of methods is used (see Fig. 2 c), with qualitative methods (n = 22) being the most frequently used.
In order to derive a comprehensive list of factors that influence the use of AI in production at different levels, we follow a qualitative content analysis. It is based on inductive category development, avoiding prefabricated categories in order to allow new categories to emerge based on the content at hand [ 44 , 45 ]. To do this, we first read the entire text to gain an understanding of the content and then derive codes [ 46 ] that seem to capture key ideas [ 45 ]. The codes are subsequently sorted into distinct categories, each of which is clearly defined and establishes meaningful connections between different codes. Based on an iterative process with feedback loops, the assigned categories are continuously reviewed and updated as revisions are made [ 44 ].
Various factors at different levels are of significance to AI and influence technology adoption [ 47 , 48 ]. To identify the specific factors that are of importance for AI adoption in production, we analyze the selected contributions in terms of the factors considered, compare them with each other and consequently obtain a list of factors through a bottom-up approach. While some of the factors are based on empirical findings, others are expected factors that result from the research findings of the respective studies. Through our analysis, a list of 35 factors emerges that influence AI adoption in production which occur with varying frequency in the studies analyzed by our SLR. Table 1 visualizes each factor in the respective contributions sorted by the frequency of occurrence.
The presence of skills is considered a particularly important factor in AI adoption in the studies analyzed (n = 35). The availability of data (n = 25) as well as the need for ethical guidelines (n = 24) are also seen as key drivers of AI adoption, as data is seen as the basis for the implementation of AI and ethical issues must be addressed in handling such an advanced technology. As such, these three factors make up the accelerants of AI adoption in production that are most frequently cited in the studies analyzed.
Also of importance are issues of managerial support (n = 22), as well as performance measures and IT infrastructure (n = 20). Some factors were also mentioned, but only addressed by one study at a time: government support, industrial sector, product complexity, batch size, and R&D Intensity. These factors are often used as quantitatively measurable adoption factors, especially in empirical surveys, such the study by Kinkel, Baumgartner, Cherubini [ 36 ].
The 35 factors presented characteristically in Sect. 2.4 serve as the basis for our in-depth analysis and for developing a framework of influences on AI adoption in production which are grouped into supercategories. A supercategory describes a cluster of topics to which various factors of AI adoption in production can be assigned. We were able to define seven categories that influence AI adoption in production: the internal influences of ‘business and structure’, ‘organizational effectiveness’, ‘technology and system’, ‘data management’ as well as the external influences of the ‘regulatory environment’, ‘business environment’ and ‘economic environment’ (see Fig. 3 ). The factors that were mentioned most frequently (occurrence in at least half of the papers analyzed) are marked accordingly (*) in Fig. 3 .
Framework of factors influencing AI adoption in production
The internal influences on AI adoption in production refer to factors that an organization carries internally and that thus also influence adoption from within. Such factors can usually be influenced and clearly controlled by the organization itself.
The supercategory ‘business and structure’ includes the various factors and characteristics that impact a company’s performance, operations, and strategic decision-making. By considering and analyzing these business variables when implementing AI in production processes, companies can develop effective strategies to optimize their performance, increase their competitiveness, and adapt to changes in the business environment.
To understand and grasp the benefits in the use of AI, quantitative performance measures for the current and potential use of AI in industrial production systems help to clarify the value and potential benefits of AI use [ 49 , 54 , 74 , 79 , 91 ]. Assessing possible risks [ 77 ] as well as the monetary expected benefits for AI (e.g. Return on Investment (ROI)) in production plays an important role for adoption decisions in market-oriented companies [ 57 , 58 , 63 , 65 , 78 ]. Due to financial constraints, managers behave cautiously in their investments [ 78 ], so they need to evaluate AI adoption as financially viable to want to make the investment [ 61 , 63 , 93 ] and also drive acceptance [ 60 ]. AI systems can significantly improve cost–benefit structures in manufacturing, thereby increasing the profitability of production systems [ 73 ] and making companies more resilient [ 75 ]. However, in most cases, the adoption of AI requires high investments and the allocation of resources (s.a. personnel or financial) for this purpose [ 50 , 51 , 57 , 80 , 94 ]. Consequently, a lack of budgets and high expected transition costs often hinder the implementation of smart concepts [ 56 , 62 , 67 , 82 , 84 , 92 ]. It is up to management to provide necessary funding for AI adoption [ 53 , 59 , 79 ], which is required, for example, for skill development of employees [ 59 , 61 , 63 ], IT adaptation [ 62 , 66 ], AI development [ 74 ] or hardware deployment [ 68 ]. In their empirical study, Kinkel, Baumgartner, Cherubini [ 36 ] confirm a positive correlation between company size and the intensity in the use of AI technologies. Large companies generally stand out with a higher propensity to adopt [ 53 ] as they have less difficulties in comparison to small firms regarding the availability of resources [ 69 ], such as know-how, budget [ 68 , 84 ] and general data organization [ 68 ]. Others argue that small companies tend to be more open to change and are characterized by faster decision-making processes [ 68 , 93 ]. Product complexity also influences a company’s propensity for AI. Companies that produce rather simple products are more likely to digitize, which in turn offers good starting points for AI adoption. On the other hand, complex product manufacturers (often characterized by small batch sizes) are often less able to standardize and automate [ 36 ]. The company’s produced batch size has a similar influence on AI adoption. Small and medium batch sizes in particular hinder the integration of intelligent technologies, as less automation often prevails here as well. Nevertheless, even small and medium lot sizes can benefit economically from AI [ 36 ]. Since a high R&D intensity indicates a high innovation capability of a company, it is assumed to have a positive influence on AI adoption, as companies with a high R&D intensity already invest heavily in and use new innovations. This in turn speaks for existing competencies, know how and structures [ 36 ].
This supercategory focuses on the broader aspects that contribute to the effectiveness, development, and success of an organization when implementing AI in a production context. As the factors are interconnected and influence each other, decision makers should consider them carefully.
Users´ trust in AI is an essential factor to enable successful AI adoption and use in production [ 52 , 68 , 78 , 79 , 88 , 90 ]. From the users´ perspective, AI often exhibits the characteristics of a black box because its inherent processes are not fully understood [ 50 , 90 ] which can lead individuals to develop a fear towards the unknown [ 71 ]. Because of this lack of understanding, successful interaction between humans and AI is not guaranteed [ 90 ], as trust is a foundation for decisions that machines are intended to make autonomously [ 52 , 91 ]. To strengthen faith in AI systems [ 76 , 80 ], AI users can be involved in AI design processes in order to understand appropriate tools [ 54 , 90 ]. In this context, trust is also discussed in close connection with transparency and regulation [ 79 ]. User resistance is considered a barrier to implementing new information technologies, as adoption requires change [ 53 , 62 , 92 ]. Ignorance, as a kind of resistance to change, is a main obstacle to successful digital transformation [ 51 , 56 , 65 ]. Some employees may resist the change brought about by AI because they fear losing their jobs [ 52 ] or have other concerns [ 78 ]. Overcoming resistance to technology adoption requires organizational change and is critical for the success of adoption [ 50 , 51 , 62 , 67 , 71 , 80 ]. Therefore, change management is important to create awareness of the importance of AI adoption and increase acceptance of the workforce [ 66 , 68 , 74 , 83 ]. Management commitment is seen as a significant driver of technology adoption [ 53 , 59 , 81 , 82 , 86 ] and a lack of commitment can negatively impact user adoption and workforce trust and lead to skepticism towards technology [ 86 ]. The top management’s understanding and support for the benefits of the adopted technology [ 53 , 56 , 67 , 78 , 93 , 94 ] enhances AI adoption, can prioritize its implementation and also affects the performance of the AI-enabled application [ 55 , 60 , 83 ]. Preparing, enabling, and thus empowering the workforce, are considered the management’s responsibility in the adoption of digital technologies [ 59 , 75 ]. This requires intelligent leadership [ 52 ] as decision makers need to integrate their workforce into decision-making processes [ 75 ]. Guidelines can support managers by providing access to best practices that help in the adoption of AI [ 50 ]. Critical measures to manage organizational change include the empowerment of visionaries or appointed AI champions leading the change and the collaborative development of digital roadmaps [ 54 , 62 ]. To demonstrate management commitment, managers can create such a dedicated role, consisting of an individual or a small group that is actively and enthusiastically committed to AI adoption in production. This body is considered the adoption manager, point of contact and internal driver of adoption [ 62 , 74 , 80 ]. AI initiatives in production do not necessarily have to be initiated by management. Although management support is essential for successful AI adoption, employees can also actively drive integration initially and thus realize pilot projects or initial trials [ 66 , 80 ]. The development of strategies as well as roadmaps is considered another enabling and necessary factor for the adoption of AI in production [ 50 , 53 , 54 , 62 , 71 , 93 ]. While many major AI strategies already exist at country level to further promote research and development of AI [ 87 ], strategy development is also important at the firm level [ 76 , 77 , 81 ]. In this context, strategies should not be delegated top-down, but be developed in a collaborative manner, i.e. by engaging the workforce [ 75 ] and be in alignment with clear visions [ 91 , 94 ]. Roadmaps are used to improve planning, support implementation, facilitate the adoption of smart technologies in manufacturing [ 93 ] and should be integrated into both business and IT strategy [ 62 , 66 ]. In practice, clear adoption roadmaps that provide approaches on how to effectively integrate AI into existing strategies and businesses are often lacking [ 56 , 87 ]. The need for AI-related skills in organizations is a widely discussed topic in AI adoption analyses [ 79 ]. In this context, the literature points both at the need for specific skills in the development and design of AI applications [ 57 , 71 , 72 , 73 , 76 , 93 ] as well as the skills in using the technology [ 53 , 65 , 73 , 74 , 75 , 84 , 93 ] which availability in the firm is not always given [ 49 ]. AI requires new digital skills [ 36 , 50 , 52 , 55 , 56 , 59 , 61 , 63 , 66 , 78 , 80 ], where e.g. advanced analytics [ 64 , 75 , 81 ], programming skills [ 68 ] and cybersecurity skills [ 78 , 93 ] gain importance. The lack of skills required for AI is seen as a major challenge of digital transformation, as a skilled workforce is considered a key resource for companies [ 51 , 54 , 56 , 60 , 62 , 67 , 69 , 70 , 82 , 93 ]. This lack of a necessary skillset hinders the adoption of AI tools in production systems [ 58 , 77 ]. Closely related to skills is the need for new training concepts, which organizations need to consider when integrating digital technologies [ 49 , 50 , 51 , 56 , 59 , 63 , 71 , 74 , 75 ]. Firms must invest in qualification in order to create necessary competences [ 73 , 78 , 80 , 81 , 92 ]. Additionally, education must target and further develop the skills required for effectively integrating intelligent technologies into manufacturing processes [ 54 , 61 , 62 , 83 ]. Regarding this issue, academic institutions must develop fitting curricula for data driven manufacturing engineering [ 64 ]. Another driving factor of AI adoption is the innovation culture of an organization, which is influenced by various drivers. For example, companies that operate in an environment with high innovation rates, facing intense competitive pressures are considered more likely to see smart technologies as a tool for strategic change [ 83 , 91 , 93 ]. These firms often invest in more expensive and advanced smart technologies as the pressure and resulting competition forces them to innovate [ 93 ]. Another way of approach this is that innovation capability can also be supported and complemented by AI, for example by intelligent systems supporting humans in innovation or even innovating on their own [ 52 ].The entrepreneurial orientation of a firm is characterized in particular by innovativeness [ 66 ], productivity [ 63 ], risk-taking [ 86 ] as well as continuous improvement [ 50 ]. Such characteristics of an innovating culture are considered essential for companies to recognise dynamic changes in the market and make adoption decisions [ 51 , 71 , 81 , 84 , 86 , 94 ]. The prevalence of a digital mindset in companies is important for technology adoption, as digital transformation affects the entire organizational culture and behavior [ 59 , 80 , 92 ] and a lack of a digital culture [ 50 , 65 ] as well as a ‘passive mindset’ [ 78 ] can hinder the digital transformation of firms. Organizations need to develop a corresponding culture [ 66 , 67 , 71 ], also referred to as ‘AI-ready-culture’ [ 54 ], that promotes development and encourages people and data through the incorporation of technology [ 71 , 75 ]. With the increasing adoption of smart technologies, a ‘new digital normal’ is emerging, characterized by hybrid work models, more human–machine interactions and an increased use of digital technologies [ 75 , 83 ].
The ‘technology and system’ supercategory focuses on the broader issues related to the technology and infrastructure that support organizational operations and provide the technical foundation for AI deployment.
By IT infrastructure we refer to issues regarding the foundational systems and IT needed for AI adoption in production. Industrial firms and their IT systems must achieve a mature technological readiness in order to enable successful AI adoption [ 51 , 60 , 67 , 69 , 83 ]. A lack of appropriate IT infrastructure [ 68 , 71 , 78 , 91 ] or small maturity of Internet of Things (IoT) technologies [ 70 ]) hinders the efficient use of data in production firms [ 56 ] which is why firms must update their foundational information systems for successful AI adoption [ 53 , 54 , 62 , 66 , 72 , 75 ]. IT and data security are fundamental for AI adoption and must be provided [ 50 , 51 , 68 , 82 ]. This requires necessary developments that can ensure security during AI implementation while complying with legal requirements [ 52 , 72 , 78 ]. Generally, security concerns are common when implementing AI innovations [ 72 , 79 , 91 , 94 ]. This fear of a lack of security can also prevent the release of (e.g. customer) data in a production environment [ 56 ]. Additionally, as industrial production systems are vulnerable to failures as well as cyberattacks, companies need to address security and cybersecurity measures [ 49 , 76 , 88 , 89 ]. Developing user-friendly AI solutions can facilitate the adoption of smart solutions by increasing user understanding and making systems easy to use by employees as well as quick to integrate [ 50 , 72 , 84 ]. When developing user-friendly solutions which satisfy user needs [ 76 ], it is particularly important to understand and integrate the user perspective in the development process [ 90 ]. If employees find technical solutions easy to use, they are more confident in its use and perceived usefulness increases [ 53 , 67 , 68 ]. The compatibility of AI with a firm and its existing systems, i.e., the extent to which AI matches existing processes, structures, and infrastructures [ 53 , 54 , 56 , 60 , 78 , 80 , 82 , 83 , 93 , 94 ], is considered an important requirement for the adoption of AI in IT systems [ 91 ]. Along with compatibility also comes connectivity, which is intended to ensure the links within the overall network and avoid silo thinking [ 59 ]. Connectivity and interoperability of AI-based processes within the company’s IT manufacturing systems must be ensured at different system levels and are considered key factors in the development of AI applications for production [ 50 , 72 , 89 ]. The design of modular AI solutions can increase system compatibility [ 84 ]. Firms deciding for AI adoption must address safety issues [ 51 , 54 , 59 , 72 , 73 , 78 ]. This includes both safety in the use and operation of AI [ 60 , 69 ]. In order to address safety concerns of integrating AI solutions in industrial systems [ 49 ], systems must secure high reliability [ 71 ]. AI can also be integrated as a safety enabler, for example, by providing technologies to monitor health and safety in the workplace to prevent fatigue and injury [ 75 ].
Since AI adoption in the organization is strongly data-driven, the ‘data management’ supercategory is dedicated to the comprehensive aspects related to the effective and responsible management of data within the organization.
Data privacy must be guaranteed when creating AI applications based on industrial production data [ 49 , 58 , 59 , 60 , 72 , 76 , 78 , 79 , 82 , 88 , 89 , 91 , 94 ] as ‘[M]anufacturing industries generate large volumes of unstructured and sensitive data during their daily operations’ [ 89 ]. Closely related to this is the need for anonymization and confidentiality of data [ 61 , 69 , 70 , 78 ]. The availability of large, heterogeneous data sets is essential for the digital transformation of organizations [ 52 , 59 , 78 , 80 , 88 , 89 ] and is considered one of the key drivers of AI innovation [ 62 , 68 , 72 , 86 ]. In production systems, lack of data availability is often a barrier to AI adoption [ 58 , 70 , 77 ]. In order to enable AI to establish relationships between data, the availability of large input data that is critical [ 62 , 76 , 81 ]. New AI models are trained with this data and can adapt as well as improve as they receive new data [ 59 , 62 ]. Big data can thus significantly improve the quality of AI applications [ 59 , 71 ]. As more and more data is generated in manufacturing [ 85 ], AI opens up new opportunities for companies to make use of it [ 62 ]. However, operational data are often unstructured, as they come from different sources and exist in diverse formats [ 85 , 87 ]. This challenges data processing, as data quality and origin are key factors in the management of data [ 78 , 79 , 80 , 88 , 89 , 91 ]. To make production data valuable and usable for AI, consistency of data and thus data integrity is required across manufacturing systems [ 50 , 62 , 77 , 84 ]. Another key prerequisites for AI adoption is data governance [ 56 , 59 , 67 , 68 , 71 , 78 , 88 ] which is an important asset to make use of data in production [ 50 ] and ensure the complex management of heterogenous data sets [ 89 ]. The interoperability of data and thus the foundation for the compatibility of AI with existing systems, i.e., the extent to which AI matches existing processes, structures, and infrastructures [ 53 , 56 , 84 , 93 ], is considered another important requirement for the adoption of AI in IT systems. Data interoperability in production systems can be hindered by missing data standards as different machines use different formats [ 87 ]. Data processing refers to techniques used to preparing data for analysis which is essential to obtain consistent results from data analytics in production [ 58 , 72 , 80 , 81 , 84 ]. In this process, the numerous, heterogeneous data from different sensors are processed in such a way that they can be used for further analyses [ 87 ]. The capability of production firms to process data and information is thus important to enable AI adoption [ 77 , 86 , 93 ]. With the increasing data generation in the smart and connected factory, the strategic relevance of data analytics is gaining importance [ 55 , 69 , 78 ], as it is essential for AI systems in performing advanced data analyses [ 49 , 67 , 72 , 86 , 88 ]. Using analytics, valuable insights can be gained from the production data obtained using AI systems [ 58 , 77 , 87 ]. In order to enable the processing of big data, a profound data infrastructure is necessary [ 65 , 75 , 87 ]. Facilities must be equipped with sensors, that collect data and model information, which requires investments from firms [ 72 ]. In addition, production firms must build the necessary skills, culture and capabilities for data analytics [ 54 , 75 , 87 , 93 ]. Data storage, one of the foundations and prerequisites for smart manufacturing [ 54 , 68 , 71 , 74 ], must be ensured in order to manage the larg amounts of data and thus realize the adoption of intelligent technologies in production [ 50 , 59 , 72 , 78 , 84 , 87 , 88 , 89 ].
The external drivers of AI adoption in production influence the organization through conditions and events from outside the firm and are therefore difficult to control by the organization itself.
This supercategory captures the broader concept of establishing rules, standards, and frameworks that guide the behavior, actions, and operations of individuals, organizations, and societies when implementing AI.
AI adoption in production faces many ethical challenges [ 70 , 72 , 79 ]. AI applications must be compliant with the requirements of organizational ethical standards and laws [ 49 , 50 , 59 , 60 , 62 , 75 ] which is why certain issues must be examined in AI adoption and AI design [ 62 , 73 , 82 , 91 ] so that fairness and justice are guaranteed [ 78 , 79 , 92 ]. Social rights, cultural values and norms must not be violated in the process [ 49 , 52 , 53 , 81 ]. In this context, the explainability and transparency of AI decisions also plays an important role [ 50 , 54 , 58 , 70 , 78 , 89 ] and can address the characteristic of AI of a black box [ 90 ]. In addition, AI applications must be compliant with legal and regulatory requirements [ 51 , 52 , 59 , 77 , 81 , 82 , 91 ] and be developed accordingly [ 49 , 76 ] in order to make organization processes using AI clear and effective [ 65 ]. At present, policies and regulation of AI are still in its infancy [ 49 ] and missing federal regulatory guidelines, standards as well as incentives hinder the adoption of AI [ 67 ] which should be expanded simultaneously to the expansion of AI technology [ 60 ]. This also includes regulations on the handling of data (e.g. anonymization of data) [ 61 , 72 ].
The factors in the ‘business environment’ supercategory refer to the external conditions and influences that affect the operations, decision making, and performance of the company seeking to implement AI in a production context.
Cooperation and collaboration can influence the success of digital technology adoption [ 52 , 53 , 59 , 72 ], which is why partnerships are important for adoption [ 53 , 59 ] and can positively influence its future success [ 52 , 67 ]. Both intraorganizational and interorganizational knowledge sharing can positively influence AI adoption [ 49 ]. In collaborations, companies can use a shared knowledge base where data and process sharing [ 51 , 59 , 94 ] as well as social support systems strengthen feedback loops between departments [ 79 , 80 ]. With regard to AI adoption in firms, vendors as well as service providers need to collaborate closely to improve the compatibility and operational capability of smart technologies across different industries [ 82 , 93 ]. Without external IT support, companies can rarely integrate AI into their production processes [ 66 ], which is why thorough support from vendors can significantly facilitate the integration of AI into existing manufacturing processes [ 80 , 91 ]. Public–private collaborations can also add value and governments can target AI dissemination [ 60 , 74 ]. The support of the government also positively influences AI adoption. This includes investing in research projects and policies, building a regulatory setting as well as creating a collaborative environment [ 60 ]. Production companies are constantly exposed to changing conditions, which is why the dynamics of the environment is another factor influencing the adoption of AI [ 52 , 63 , 72 , 86 ]. Environmental dynamics influence the operational performance of firms and can favor an entrepreneurial orientation of firms [ 86 ]. In order to respond to dynamics, companies need to develop certain capabilities and resources (i.e. dynamic capabilities) [ 86 ]. This requires the development of transparency, agility, as well as resilience to unpredictable changes, which was important in the case of the COVID-19 pandemic, for example, where companies had to adapt quickly to changing environments [ 75 ]. A firm’s environment (e.g. governments, partners or customers) can also pressure companies to adopt digital technologies [ 53 , 67 , 82 , 91 ]. Companies facing intense competition are considered more likely to invest in smart technologies, as rivalry pushes them to innovate and they hope to gain competitive advantages from adoption [ 36 , 66 , 82 , 93 ].
By considering both the industrial sector and country within the subcategory ‘economic environment’, production firms can analyze the interplay between the two and understand how drivers can influence the AI adoption process in their industrial sector’s performance within a particular country.
The industrial sector of a firm influences AI adoption in production from a structural perspective, as it indicates variations in product characteristics, governmental support, the general digitalization status, the production environment as well as the use of AI technologies within the sector [ 36 ]. Another factor that influences AI adoption is the country in which a company is located. This influences not only cultural aspects, the availability of know-how and technology orientation, but also regulations, laws, standards and subsidies [ 36 ]. From another perspective, AI can also contribute to the wider socio-economic growth of economies by making new opportunities easily available and thus equipping e.g. more rural areas with advanced capabilities [ 78 ].
The analysis of AI adoption in production requires a comprehensive analysis of the various factors that influence the introduction of the innovation. As discussed by Kinkel, Baumgartner, Cherubini [ 36 ], our research also concludes that organizational factors have a particularly important role to play. After evaluating the individual drivers of AI adoption in production in detail in this qualitative synthesis, we draw a conclusion from the results and derive a research agenda from the analysis to serve as a basis for future research. The RQs emerged from the analyzed factors and are presented in Table 2 . We developed the questions based on the literature review and identified research gaps for every factor that was most frequently mentioned. From the factors analyzed and RQs developed, the internal environment has a strong influence on AI adoption in production, and organizational factors play a major role here.
Looking at the supercategory ‘business and environment’, performance indicators and investments are considered drivers of AI adoption in production. Indicators to measure the performance of AI innovations are necessary here so that managers can perform cost–benefit analyses and make the right decision for their company. There is a need for research here to support possible calculations and show managers a comprehensive view of the costs and benefits of technology in production. In terms of budget, it should be noted that AI adoption involves a considerable financial outlay that must be carefully weighed and some capital must be available to carry out the necessary implementation efforts (e.g., staffing costs, machine retrofits, change management, and external IT service costs). Since AI adoption is a complex process and turnkey solutions can seldom be implemented easily and quickly, but require many changes (not only technologically but also on an organizational level), it is currently difficult to estimate the necessary budgets and thus make them available. Especially the factors of the supercategory ‘organizational effectiveness’ drive AI adoption in production. Trust of the workforce is considered an important driver, which must be created in order to successfully implement AI. This requires measures that can support management in building trust. Closely related to this are the necessary change management processes that must be initiated to accompany the changes in a targeted manner. Management itself must also play a clear role in the introduction of AI and communicate its support, as this also influences the adoption. The development of clear processes and measures can help here. Developing roadmaps for AI adoption can facilitate the adoption process and promote strategic integration with existing IT and business strategy. Here, best practice roadmaps and necessary action steps can be helpful for companies. Skills are considered the most important driver for AI adoption in manufacturing. Here, there is a lack of clear approaches that support companies in identifying the range of necessary skills and, associated with this, also opportunities to further develop these skills in the existing workforce. Also, building a culture of innovation requires closer research that can help companies foster a conducive environment for AI adoption and the integration of other smart technologies. Steps for developing a positive mindset require further research that can provide approaches for necessary action steps and measures in creating a positive digital culture. With regard to ‘technology and system’, the factors of IT infrastructure and security in particular are driving AI adoption in production. Existing IT systems must reach a certain maturity to enable AI adoption on a technical level. This calls for clear requirements that visualize for companies which systems and standards are in place and where developments are needed. Security must be continuously ensured, for which certain standards and action catalogs must be developed. With regard to the supercategory ‘data management’, the availability of data is considered the basis for successful AI adoption, as no AI can be successfully deployed without data. In the production context in particular, this requires developments that support companies in the provision of data, which usually arises from very heterogeneous sources and forms. Data analytics must also be closely examined, and production companies usually need external support in doing so. The multitude of data also requires big data storage capabilities. Here, groundwork is needed to show companies options about the possibilities of different storage options (e.g., on premis vs. cloud-based).
In the ‘regulatory environment’, ethics in particular is considered a driver of AI adoption in production. Here, fundamental ethical factors and frameworks need to be developed that companies can use as a guideline to ensure ethical standards throughout the process. Cooperations and environmental dynamism drive the supercategory ‘business environment’. Collaborations are necessary to successfully implement AI adoption and action is needed to create the necessary contact facilitation bodies. In a competitive environment, companies have to make quick decisions under strong pressure, which also affects AI adoption. Here, guidelines and also best practice approaches can help to simplify decisions and quickly demonstrate the advantage of the solutions. There is a need for research in this context.
The use of AI technologies in production continues to gain momentum as managers hope to increase efficiency, productivity and reduce costs [ 9 , 13 , 20 ]. Although the benefits of AI adoption speak for themselves, implementing AI is a complex decision that requires a lot of knowledge, capital and change [ 95 ] and is influenced by various internal and external factors. Therefore, managers are still cautious about implementing the technology in a production context. Our SLR seeks to examine the emergent phenomenon of AI in production with the precise aim of understanding the factors influencing AI adoption and the key topics discussed in the literature when analyzing AI in a production context. For this purpose, we use the current state of research and examine the existing studies based on the methodology of a systematic literature analysis and respond to three RQs.
We answer RQ1 by closely analyzing the literature selected in our SLR to identify trends in current research on AI adoption in production. In this process, it becomes clear that the topic is gaining importance and that research has increased over the last few years. In the field of production, AI is being examined from various angles and current research addresses aspects from a business, human and technical perspective. In our response to RQ2 we synthesized the existing literature to derive 35 factors that influence AI adoption in production at different levels from inside or outside the organization. In doing so, we find that AI adoption in production poses particularly significant challenges to organizational effectiveness compared to other digital technologies and that the relevance of data management takes on a new dimension. Production companies often operate more traditionally and are sometimes rigid when it comes to change [ 96 , 97 ], which can pose organizational challenges when adopting AI. In addition, the existing machines and systems are typically rather heterogeneous and are subject to different digitalization standards, which in turn can hinder the availability of the necessary data for AI implementation [ 98 , 99 ]. We address RQ3 by deriving a research agenda, which lays a foundation for further scientific research and deepening the understanding of AI adoption in production. The results of our analysis can further help managers to better understand AI adoption and to pay attention to the different factors that influence the adoption of this complex technology.
Our paper takes the first step towards analysing the current state of the research on AI adoption from a production perspective. We represent a holistic view on the topic, which is necessary to get a better understanding of AI in a production-context and build a comprehensive view on the different dimensions as well as factors influencing its adoption. To the best of our knowledge, this is the first contribution that systematises research about the adoption of AI in production. As such, it makes an important contribution to current AI and production research, which is threefold:
First, we highlight the characteristics of studies conducted in recent years on the topic of AI adoption in production, from which several features and developments can be deduced. Our results confirm the topicality of the issue and the increasing relevance of research in the field.
Having laid the foundations for understanding AI in production, we focused our research on the identification and systematization of the most relevant factors influencing AI adoption in production at different levels. This brings us to the second contribution, our comprehensive factor analysis of AI adoption in production provides a framework for further research as well as a potential basis for managers to draw upon when adopting AI. By systematizing the relevant factors influencing AI adoption in production, we derived a set of 35 researched factors associated with AI adoption in production. These factors can be clustered in two areas of analysis and seven respective supercategories. The internal environment area includes four levels of analysis: ‘business and structure’ (focusing on financial aspects and firm characteristics), ‘organizational effectiveness’ (focusing on human-centred factors), ‘technology and system’ (based on the IT infrastructure and systems) as well as ‘data management’ (including all data related factors). Three categories are assigned to the external environment: the ‘regulatory environment’ (such as ethics and the regulatory forms), the ‘business environment’ (focused on cooperation activities and dynamics in the firm environment) and the ‘economic environment’ (related to sectoral and country specifics).
Third, the developed research plan as outlined in Table 2 serves as an additional outcome of the SLR, identifying key RQs in the analyzed areas that can serve as a foundation for researchers to expand the research area of AI adoption in production. These RQs are related to the mostly cited factors analyzed in our SLR and aim to broaden the understanding on the emerging topic.
The resulting insights can serve as the basis for strategic decisions by production companies looking to integrate AI into their processes. Our findings on the factors influencing AI adoption as well as the developed research agenda enhance the practical understanding of a production-specific adoption. Hence, they can serve as the basis for strategic decisions for companies on the path to an effective AI adoption. Managers can, for example, analyse the individual factors in light of their company as well as take necessary steps to develop further aspects in a targeted manner. Researchers, on the other hand, can use the future research agenda in order to assess open RQs and can expand the state of research on AI adoption in production.
Since a literature review must be restricted in its scope in order to make the analyses feasible, our study provides a starting point for further research. Hence, there is a need for further qualitative and quantitative empirical research on the heterogeneous nature of how firms configure their AI adoption process. Along these lines, the following aspects would be of particular interest for future research to improve and further validate the analytical power of the proposed framework.
First, the lack of research on AI adoption in production leads to a limited number of papers included in this SLR. As visualized in Fig. 2 , the number of publications related to the adoption of AI in production has been increasing since 2018 but is, to date, still at an early stage. For this reason, only 47 papers published until May 2024 addressing the production-specific adoption of AI were identified and therefore included in our analysis for in-depth investigation. This rather small number of papers included in the full-text analysis gives a limited view on AI adoption in production but allows a more detailed analysis. As the number of publications in this research field increases, there seems to be a lot of research happening in this field which is why new findings might be constantly added and developed as relevant in the future [ 39 ]. Moreover, in order to research AI adoption from a more practical perspective and thus to build up a broader, continuously updated view on AI adoption in production, future literature analyses could include other publication formats, e.g. study reports of research institutions and companies, as well discussion papers.
Second, the scope of the application areas of AI in production has been increasing rapidly. Even though our overview of the three main areas covered in the recent literature serves as a good basis for identifying the most dominant fields for AI adoption in production, a more detailed analysis could provide a better overview of possibilities for manufacturing companies. Hence, a further systematisation as well as evaluation of application areas for AI in production can provide managers with the information needed to decide where AI applications might be of interest for the specific company needs.
Third, the systematisation of the 35 factors influencing AI adoption in production serve as a good ground for identifying relevant areas influenced by and in turn influencing the adoption of AI. Further analyses should be conducted in order to extend this view and extend the framework. For example, our review could be combined with explorative research methods (such as case studies in production firms) in order to add the practical insights from firms adopting AI. This integration of practical experiences can also help exploit and monitor more AI-specific factors by observing AI adoption processes. In enriching the factors through in-depth analyses, the results of the identified AI adoption factors could also be examined in light of theoretical contributions like the technology-organization-environment (TOE) framework [ 47 ] and other adoption theories.
Fourth, in order to examine the special relevance of identified factors for AI adoption process and thus to distinguish it from the common factors influencing the adoption of more general digital technologies, there is a further need for more in-depth (ethnographic) research into their impacts on the adoption processes, particularly in the production context. Similarly, further research could use the framework introduced in this paper as a basis to develop new indicators and measurement concepts as well as to examine their impacts on production performance using quantitative methods.
Benner MJ, Waldfogel J (2020) Changing the channel: digitization and the rise of “middle tail” strategies. Strat Mgmt J 86:1–24. https://doi.org/10.1002/smj.3130
Article Google Scholar
Roblek V, Meško M, Krapež A (2016) A complex view of industry 4.0. SAGE Open. https://doi.org/10.1177/2158244016653987
Oliveira BG, Liboni LB, Cezarino LO et al (2020) Industry 4.0 in systems thinking: from a narrow to a broad spectrum. Syst Res Behav Sci 37:593–606. https://doi.org/10.1002/sres.2703
Li B, Hou B, Yu W et al (2017) Applications of artificial intelligence in intelligent manufacturing: a review. Frontiers Inf Technol Electronic Eng 18:86–96. https://doi.org/10.1631/FITEE.1601885
Dhamija P, Bag S (2020) Role of artificial intelligence in operations environment: a review and bibliometric analysis. TQM 32:869–896. https://doi.org/10.1108/TQM-10-2019-0243
Collins C, Dennehy D, Conboy K et al (2021) Artificial intelligence in information systems research: a systematic literature review and research agenda. Int J Inf Manage 60:102383. https://doi.org/10.1016/j.ijinfomgt.2021.102383
Chien C-F, Dauzère-Pérès S, Huh WT et al (2020) Artificial intelligence in manufacturing and logistics systems: algorithms, applications, and case studies. Int J Prod Res 58:2730–2731. https://doi.org/10.1080/00207543.2020.1752488
Chen H (2019) Success factors impacting artificial intelligence adoption: perspective from the telecom industry in China, Old Dominion University
Sanchez M, Exposito E, Aguilar J (2020) Autonomic computing in manufacturing process coordination in industry 4.0 context. J Industrial Inf Integr. https://doi.org/10.1016/j.jii.2020.100159
Lee J, Davari H, Singh J et al (2018) Industrial artificial intelligence for industry 4.0-based manufacturing systems. Manufacturing Letters 18:20–23. https://doi.org/10.1016/j.mfglet.2018.09.002
Heimberger H, Horvat D, Schultmann F (2023) Assessing AI-readiness in production—A conceptual approach. In: Huang C-Y, Dekkers R, Chiu SF et al. (eds) intelligent and transformative production in pandemic times. Springer, Cham, pp 249–257
Horvat D, Heimberger H (2023) AI Readiness: An Integrated Socio-technical Framework. In: Deschamps F, Pinheiro de Lima E, Da Gouvêa Costa SE et al. (eds) Proceedings of the 11 th international conference on production research—Americas: ICPR Americas 2022, 1 st ed. 2023. Springer Nature Switzerland; Imprint Springer, Cham, pp 548–557
Wang J, Ma Y, Zhang L et al (2018) Deep learning for smart manufacturing: methods and applications. J Manuf Syst 48:144–156. https://doi.org/10.1016/J.JMSY.2018.01.003
Davenport T, Guha A, Grewal D et al (2020) How artificial intelligence will change the future of marketing. J Acad Mark Sci 48:24–42. https://doi.org/10.1007/s11747-019-00696-0
Cui R, Li M, Zhang S (2022) AI and procurement. Manufacturing Serv Operations Manag 24(691):706. https://doi.org/10.1287/msom.2021.0989
Pournader M, Ghaderi H, Hassanzadegan A et al (2021) Artificial intelligence applications in supply chain management. Int J Prod Econ 241:108250. https://doi.org/10.1016/j.ijpe.2021.108250
Su H, Li L, Tian S et al (2024) Innovation mechanism of AI empowering manufacturing enterprises: case study of an industrial internet platform. Inf Technol Manag. https://doi.org/10.1007/s10799-024-00423-4
Venkatesh V, Raman R, Cruz-Jesus F (2024) AI and emerging technology adoption: a research agenda for operations management. Int J Prod Res 62:5367–5377. https://doi.org/10.1080/00207543.2023.2192309
Senoner J, Netland T, Feuerriegel S (2022) Using explainable artificial intelligence to improve process quality: evidence from semiconductor manufacturing. Manage Sci 68:5704–5723. https://doi.org/10.1287/mnsc.2021.4190
Fosso Wamba S, Queiroz MM, Ngai EWT et al (2024) The interplay between artificial intelligence, production systems, and operations management resilience. Int J Prod Res 62:5361–5366. https://doi.org/10.1080/00207543.2024.2321826
Uren V, Edwards JS (2023) Technology readiness and the organizational journey towards AI adoption: an empirical study. Int J Inf Manage 68:102588. https://doi.org/10.1016/j.ijinfomgt.2022.102588
Berente N, Gu B, Recker J (2021) Managing artificial intelligence special issue managing AI. MIS Quarterly 45:1433–1450
Google Scholar
Scafà M, Papetti A, Brunzini A et al (2019) How to improve worker’s well-being and company performance: a method to identify effective corrective actions. Procedia CIRP 81:162–167. https://doi.org/10.1016/j.procir.2019.03.029
Wang H, Qiu F (2023) AI adoption and labor cost stickiness: based on natural language and machine learning. Inf Technol Manag. https://doi.org/10.1007/s10799-023-00408-9
Lindebaum D, Vesa M, den Hond F (2020) Insights from “the machine stops ” to better understand rational assumptions in algorithmic decision making and its implications for organizations. Acad Manag Rev 45:247–263. https://doi.org/10.5465/amr.2018.0181
Baskerville RL, Myers MD, Yoo Y (2020) Digital first: the ontological reversal and new challenges for information systems research. MIS Quarterly 44:509–523
Frey CB, Osborne MA (2017) The future of employment: How susceptible are jobs to computerisation? Technol Forecast Soc Chang 114:254–280. https://doi.org/10.1016/J.TECHFORE.2016.08.019
Jarrahi MH (2018) Artificial intelligence and the future of work: human-AI symbiosis in organizational decision making. Bus Horiz 61:577–586. https://doi.org/10.1016/j.bushor.2018.03.007
Fügener A, Grahl J, Gupta A et al (2021) Will humans-in-the-loop become borgs? Merits and pitfalls of working with AI. MIS Quarterly 45:1527–1556
Klumpp M (2018) Automation and artificial intelligence in business logistics systems: human reactions and collaboration requirements. Int J Log Res Appl 21:224–242. https://doi.org/10.1080/13675567.2017.1384451
Schrettenbrunnner MB (2020) Artificial-Intelligence-driven management. IEEE Eng Manag Rev 48:15–19. https://doi.org/10.1109/EMR.2020.2990933
Li J, Li M, Wang X et al (2021) Strategic directions for AI: the role of CIOs and boards of directors. MIS Quarterly 45:1603–1644
Brock JK-U, von Wangenheim F (2019) Demystifying AI: What digital transformation leaders can teach you about realistic artificial intelligence. Calif Manage Rev 61:110–134. https://doi.org/10.1177/1536504219865226
Lee J, Suh T, Roy D et al (2019) Emerging technology and business model innovation: the case of artificial intelligence. JOItmC 5:44. https://doi.org/10.3390/joitmc5030044
Chen J, Tajdini S (2024) A moderated model of artificial intelligence adoption in firms and its effects on their performance. Inf Technol Manag. https://doi.org/10.1007/s10799-024-00422-5
Kinkel S, Baumgartner M, Cherubini E (2022) Prerequisites for the adoption of AI technologies in manufacturing—evidence from a worldwide sample of manufacturing companies. Technovation 110:102375. https://doi.org/10.1016/j.technovation.2021.102375
Mikalef P, Gupta M (2021) Artificial intelligence capability: Conceptualization, measurement calibration, and empirical study on its impact on organizational creativity and firm performance. Inf Manag 58:103434. https://doi.org/10.1016/j.im.2021.103434
McElheran K, Li JF, Brynjolfsson E et al (2024) AI adoption in America: Who, what, and where. Economics Manag Strategy 33:375–415. https://doi.org/10.1111/jems.12576
Tranfield D, Denyer D, Smart P (2003) Towards a methodology for developing evidence-informed management knowledge by means of systematic review. Br J Manag 14:207–222. https://doi.org/10.1111/1467-8551.00375
Cooper H, Hedges LV, Valentine JC (2009) Handbook of research synthesis and meta-analysis. Russell Sage Foundation, New York
Page MJ, McKenzie JE, Bossuyt PM et al (2021) The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 372:n71. https://doi.org/10.1136/bmj.n71
Denyer D, Tranfield D (2011) Producing a systematic review. In: Buchanan DA, Bryman A (eds) The Sage handbook of organizational research methods. Sage Publications Inc, Thousand Oaks, CA, pp 671–689
Burbidge JL, Falster P, Riis JO et al (1987) Integration in manufacturing. Comput Ind 9:297–305. https://doi.org/10.1016/0166-3615(87)90103-5
Mayring P (2000) Qualitative content analysis. Forum qualitative Sozialforschung/Forum: Qualitative social research, Vol 1, No 2 (2000): Qualitative methods in various disciplines I: Psychology. https://doi.org/10.17169/fqs-1.2.1089
Hsieh H-F, Shannon SE (2005) Three approaches to qualitative content analysis. Qual Health Res 15:1277–1288. https://doi.org/10.1177/1049732305276687
Miles MB, Huberman AM (2009) Qualitative data analysis: An expanded sourcebook, 2nd edn. Sage, Thousand Oaks, Calif
Tornatzky LG, Fleischer M (1990) The processes of technological innovation. Issues in organization and management series. Lexington Books, Lexington, Mass.
Alsheibani S, Cheung Y, Messom C (2018) Artificial Intelligence Adoption: AI-readiness at Firm-Level: Research-in-Progress. Twenty-Second Pacific Asia Conference on Information Systems
Akinsolu MO (2023) Applied artificial intelligence in manufacturing and industrial production systems: PEST considerations for engineering managers. IEEE Eng Manag Rev 51:52–62. https://doi.org/10.1109/EMR.2022.3209891
Bettoni A, Matteri D, Montini E et al (2021) An AI adoption model for SMEs: a conceptual framework. IFAC-PapersOnLine 54:702–708. https://doi.org/10.1016/j.ifacol.2021.08.082
Boavida N, Candeias M (2021) Recent automation trends in portugal: implications on industrial productivity and employment in automotive sector. Societies 11:101. https://doi.org/10.3390/soc11030101
Botha AP (2019) A mind model for intelligent machine innovation using future thinking principles. Jnl of Manu Tech Mnagmnt 30:1250–1264. https://doi.org/10.1108/JMTM-01-2018-0021
Chatterjee S, Rana NP, Dwivedi YK et al (2021) Understanding AI adoption in manufacturing and production firms using an integrated TAM-TOE model. Technol Forecast Soc Chang 170:120880. https://doi.org/10.1016/j.techfore.2021.120880
Chiang LH, Braun B, Wang Z et al (2022) Towards artificial intelligence at scale in the chemical industry. AIChE J. https://doi.org/10.1002/aic.17644
Chouchene A, Carvalho A, Lima TM et al. (2020) Artificial intelligence for product quality inspection toward smart industries: quality control of vehicle Non-conformities. In: Garengo P (ed) 2020 9th International Conference on Industrial Technology and Management: ICITM 2020 February 11–13, 2020, Oxford, United Kingdom. IEEE, pp 127–131
Corti D, Masiero S, Gladysz B (2021) Impact of Industry 4.0 on Quality Management: identification of main challenges towards a Quality 4.0 approach. In: 2021 IEEE International Conference on Engineering, Technology and Innovation (ICE/ITMC). IEEE, pp 1–8
Demlehner Q, Schoemer D, Laumer S (2021) How can artificial intelligence enhance car manufacturing? A Delphi study-based identification and assessment of general use cases. Int J Inf Manage 58:102317. https://doi.org/10.1016/j.ijinfomgt.2021.102317
Dohale V, Akarte M, Gunasekaran A et al (2022) (2022) Exploring the role of artificial intelligence in building production resilience: learnings from the COVID-19 pandemic. Int J Prod Res 10(1080/00207543):2127961
Drobot AT (2020) Industrial Transformation and the Digital Revolution: A Focus on artificial intelligence, data science and data engineering. In: 2020 ITU Kaleidoscope: Industry-Driven Digital Transformation (ITU K). IEEE, pp 1–11
Ghani EK, Ariffin N, Sukmadilaga C (2022) Factors influencing artificial intelligence adoption in publicly listed manufacturing companies: a technology, organisation, and environment approach. IJAEFA 14:108–117
Hammer A, Karmakar S (2021) Automation, AI and the future of work in India. ER 43:1327–1341. https://doi.org/10.1108/ER-12-2019-0452
Hartley JL, Sawaya WJ (2019) Tortoise, not the hare: digital transformation of supply chain business processes. Bus Horiz 62:707–715. https://doi.org/10.1016/j.bushor.2019.07.006
Kyvik Nordås H, Klügl F (2021) Drivers of automation and consequences for jobs in engineering services: an agent-based modelling approach. Front Robot AI 8:637125. https://doi.org/10.3389/frobt.2021.637125
Mubarok K, Arriaga EF (2020) Building a smart and intelligent factory of the future with industry 4.0 technologies. J Phys Conf Ser. https://doi.org/10.1088/1742-6596/1569/3/032031
Muriel-Pera YdJ, Diaz-Piraquive FN, Rodriguez-Bernal LP et al. (2018) Adoption of strategies the fourth industrial revolution by micro, small and medium enterprises in bogota D.C. In: Lozano Garzón CA (ed) 2018 Congreso Internacional de Innovación y Tendencias en Ingeniería (CONIITI). IEEE, pp 1–6
Olsowski S, Schlögl S, Richter E et al. (2022) Investigating the Potential of AutoML as an Instrument for Fostering AI Adoption in SMEs. In: Uden L, Ting I-H, Feldmann B (eds) Knowledge Management in Organisations: 16th International Conference, KMO 2022, Hagen, Germany, July 11–14, 2022, Proceedings, 1st ed. 2022, vol 1593. Springer, Cham, pp 360–371
Rodríguez-Espíndola O, Chowdhury S, Dey PK et al (2022) Analysis of the adoption of emergent technologies for risk management in the era of digital manufacturing. Technol Forecast Soc Chang 178:121562. https://doi.org/10.1016/j.techfore.2022.121562
Schkarin T, Dobhan A (2022) Prerequisites for Applying Artificial Intelligence for Scheduling in Small- and Medium-sized Enterprises. In: Proceedings of the 24 th International Conference on Enterprise Information Systems. SCITEPRESS—Science and Technology Publications, pp 529–536
Sharma P, Shah J, Patel R (2022) Artificial intelligence framework for MSME sectors with focus on design and manufacturing industries. Mater Today: Proc 62:6962–6966. https://doi.org/10.1016/j.matpr.2021.12.360
Siaterlis G, Nikolakis N, Alexopoulos K et al. (2022) Adoption of AI in EU Manufacturing. Gaps and Challenges. In: Katalinic B (ed) Proceedings of the 33 rd International DAAAM Symposium 2022, vol 1. DAAAM International Vienna, pp 547–550
Tariq MU, Poulin M, Abonamah AA (2021) Achieving operational excellence through artificial intelligence: driving forces and barriers. Front Psychol 12:686624. https://doi.org/10.3389/fpsyg.2021.686624
Trakadas P, Simoens P, Gkonis P et al (2020) An artificial intelligence-based collaboration approach in industrial IoT manufacturing: key concepts. Architectural Ext Potential Applications Sens. https://doi.org/10.3390/s20195480
Vernim S, Bauer H, Rauch E et al (2022) A value sensitive design approach for designing AI-based worker assistance systems in manufacturing. Procedia Computer Sci 200:505–516. https://doi.org/10.1016/j.procs.2022.01.248
Williams G, Meisel NA, Simpson TW et al (2022) Design for artificial intelligence: proposing a conceptual framework grounded in data wrangling. J Computing Inf Sci Eng 10(1115/1):4055854
Wuest T, Romero D, Cavuoto LA et al (2020) Empowering the workforce in Post–COVID-19 smart manufacturing systems. Smart Sustain Manuf Syst 4:20200043. https://doi.org/10.1520/SSMS20200043
Javaid M, Haleem A, Singh RP (2023) A study on ChatGPT for Industry 4.0: background, potentials, challenges, and eventualities. J Economy Technol 1:127–143. https://doi.org/10.1016/j.ject.2023.08.001
Rathore AS, Nikita S, Thakur G et al (2023) Artificial intelligence and machine learning applications in biopharmaceutical manufacturing. Trends Biotechnol 41:497–510. https://doi.org/10.1016/j.tibtech.2022.08.007
Jan Z, Ahamed F, Mayer W et al (2023) Artificial intelligence for industry 4.0: systematic review of applications, challenges, and opportunities. Expert Syst Applications 216:119456
Waschull S, Emmanouilidis C (2023) Assessing human-centricity in AI enabled manufacturing systems: a socio-technical evaluation methodology. IFAC-PapersOnLine 56:1791–1796. https://doi.org/10.1016/j.ifacol.2023.10.1891
Stohr A, Ollig P, Keller R et al (2024) Generative mechanisms of AI implementation: a critical realist perspective on predictive maintenance. Inf Organ 34:100503. https://doi.org/10.1016/j.infoandorg.2024.100503
Pazhayattil AB, Konyu-Fogel G (2023) ML and AI Implementation Insights for Bio/Pharma Manufacturing. BioPharm International 36:24–29
Ronaghi MH (2023) The influence of artificial intelligence adoption on circular economy practices in manufacturing industries. Environ Dev Sustain 25:14355–14380. https://doi.org/10.1007/s10668-022-02670-3
Rath SP, Tripathy R, Jain NK (2024) Assessing the factors influencing the adoption of generative artificial intelligence (GenAI) in the manufacturing sector. In: Sharma SK, Dwivedi YK, Metri B et al (eds) Transfer, diffusion and adoption of next-generation digital technologies, vol 697. Springer Nature Switzerland, Cham
Bonnard R, Da Arantes MS, Lorbieski R et al (2021) Big data/analytics platform for Industry 4.0 implementation in advanced manufacturing context. Int J Adv Manuf Technol 117:1959–1973. https://doi.org/10.1007/s00170-021-07834-5
Confalonieri M, Barni A, Valente A et al. (2015) An AI based decision support system for preventive maintenance and production optimization in energy intensive manufacturing plants. In: 2015 IEEE international conference on engineering, technology and innovation/ international technology management conference (ICE/ITMC). IEEE, pp 1–8
Dubey R, Gunasekaran A, Childe SJ et al (2020) Big data analytics and artificial intelligence pathway to operational performance under the effects of entrepreneurial orientation and environmental dynamism: a study of manufacturing organisations. Int J Prod Econ 226:107599. https://doi.org/10.1016/j.ijpe.2019.107599
Lee J, Singh J, Azamfar M et al (2020) Industrial AI: a systematic framework for AI in industrial applications. China Mechanical Eng 31:37–48
Turner CJ, Emmanouilidis C, Tomiyama T et al (2019) Intelligent decision support for maintenance: an overview and future trends. Int J Comput Integr Manuf 32:936–959. https://doi.org/10.1080/0951192X.2019.1667033
Agostinho C, Dikopoulou Z, Lavasa E et al (2023) Explainability as the key ingredient for AI adoption in Industry 5.0 settings. Front Artif Intell. https://doi.org/10.3389/frai.2023.1264372
Csiszar A, Hein P, Wachter M et al. (2020) Towards a user-centered development process of machine learning applications for manufacturing domain experts. In: 2020 third international conference on artificial intelligence for industries (AI4I). IEEE, pp 36–39
Merhi MI (2023) Harfouche A (2023) Enablers of artificial intelligence adoption and implementation in production systems. Int J Prod Res. https://doi.org/10.1080/00207543.2023.2167014
Demlehner Q, Laumer S (2024) How the terminator might affect the car manufacturing industry: examining the role of pre-announcement bias for AI-based IS adoptions. Inf Manag 61:103881. https://doi.org/10.1016/j.im.2023.103881
Ghobakhloo M, Ching NT (2019) Adoption of digital technologies of smart manufacturing in SMEs. J Ind Inf Integr 16:100107. https://doi.org/10.1016/j.jii.2019.100107
Binsaeed RH, Yousaf Z, Grigorescu A et al (2023) Knowledge sharing key issue for digital technology and artificial intelligence adoption. Systems 11:316. https://doi.org/10.3390/systems11070316
Papadopoulos T, Sivarajah U, Spanaki K et al (2022) Editorial: artificial Intelligence (AI) and data sharing in manufacturing, production and operations management research. Int J Prod Res 60:4361–4364. https://doi.org/10.1080/00207543.2021.2010979
Chirumalla K (2021) Building digitally-enabled process innovation in the process industries: a dynamic capabilities approach. Technovation 105:102256. https://doi.org/10.1016/j.technovation.2021.102256
Fragapane G, Ivanov D, Peron M et al (2022) Increasing flexibility and productivity in Industry 4.0 production networks with autonomous mobile robots and smart intralogistics. Ann Oper Res 308:125–143. https://doi.org/10.1007/s10479-020-03526-7
Shahbazi Z, Byun Y-C (2021) Integration of Blockchain, IoT and machine learning for multistage quality control and enhancing security in smart manufacturing. Sensors (Basel). https://doi.org/10.3390/s21041467
Javaid M, Haleem A, Singh RP et al (2021) Significance of sensors for industry 4.0: roles, capabilities, and applications. Sensors Int 2:100110. https://doi.org/10.1016/j.sintl.2021.100110
Download references
Open Access funding enabled and organized by Projekt DEAL.
Authors and affiliations.
Business Unit Industrial Change and New Business Models, Competence Center Innovation and Knowledge Economy, Fraunhofer Institute for Systems and Innovation Research ISI, Breslauer Straße 48, 76139, Karlsruhe, Germany
Heidi Heimberger, Djerdj Horvat & Frank Schultmann
Karlsruhe Institute for Technology KIT, Institute for Industrial Production (IIP) - Chair of Business Administration, Production and Operations Management, Hertzstraße 16, 76187, Karlsruhe, Germany
Heidi Heimberger
You can also search for this author in PubMed Google Scholar
Correspondence to Heidi Heimberger .
Conflict of interest.
The authors report no conflict of interest.
Publisher's note.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .
Reprints and permissions
Heimberger, H., Horvat, D. & Schultmann, F. Exploring the factors driving AI adoption in production: a systematic literature review and future research agenda. Inf Technol Manag (2024). https://doi.org/10.1007/s10799-024-00436-z
Download citation
Accepted : 10 August 2024
Published : 23 August 2024
DOI : https://doi.org/10.1007/s10799-024-00436-z
Anyone you share the following link with will be able to read this content:
Sorry, a shareable link is not currently available for this article.
Provided by the Springer Nature SharedIt content-sharing initiative
Advertisement
Supported by
The former president is focusing his most vicious attacks on domestic political opponents, setting off fresh worries among autocracy experts.
By Michael C. Bender and Michael Gold
Donald J. Trump rose to power with political campaigns that largely attacked external targets, including immigration from predominantly Muslim countries and from south of the United States-Mexico border.
But now, in his third presidential bid, some of his most vicious and debasing attacks have been leveled at domestic opponents.
During a Veterans Day speech, Mr. Trump used language that echoed authoritarian leaders who rose to power in Germany and Italy in the 1930s, degrading his political adversaries as “vermin” who needed to be “rooted out.”
“The threat from outside forces,” Mr. Trump said, “is far less sinister, dangerous and grave than the threat from within.”
This turn inward has sounded new alarms among experts on autocracy who have long worried about Mr. Trump’s praise for foreign dictators and disdain for democratic ideals. They said the former president’s increasingly intensive focus on perceived internal enemies was a hallmark of dangerous totalitarian leaders.
We are having trouble retrieving the article content.
Please enable JavaScript in your browser settings.
Thank you for your patience while we verify access. If you are in Reader mode please exit and log into your Times account, or subscribe for all of The Times.
Thank you for your patience while we verify access.
Already a subscriber? Log in .
Want all of The Times? Subscribe .
Background Large Vessel Occlusion (LVO) is a serious condition that causes approximately 24-46% of acute ischemic strokes (AIS). LVO strokes tend to have higher mortality rates and result in more severe longterm disabilities compared to non LVO ischemic strokes. Early intervention with endovascular therapy (EVT) is recommended; however, EVT is limited to tertiary care hospitals with specialized facilities. Therefore, identifying patients with a high probability of LVO in prehospital settings and ensuring their rapid transfer to appropriate hospitals is crucial. While LVO diagnosis typically requires advanced imaging like MRI or CT scans, various scoring systems based on neurological symptoms have been developed for prehospital use. Although previous systematic reviews have addressed some of these scales, recent studies have introduced new scales and additional data on their accuracy. This systematic review and meta-analysis aim to summarize the current evidence on the diagnostic accuracy of these prehospital LVO screening scales.
Methods This systematic review and meta-analysis will be conducted in accordance with the PRISMA-DTA Statement and the Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy. We will include observational studies and randomized controlled trials that assess the utility of LVO scales in suspected stroke patients in prehospital settings. Eligible studies must provide sufficient data to calculate sensitivity and specificity, and those lacking such data or being case reports will be excluded. The literature search will cover CENTRAL, MEDLINE, and Ichushi databases, including studies in English and Japanese. Bias will be assessed using QUADAS-2, and meta-analysis will be conducted using a random effects model, with subgroup and sensitivity analyses to explore heterogeneity.
The authors have declared no competing interest.
This study did not receive any funding
I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.
The details of the IRB/oversight body that provided approval or exemption for the research described are given below:
We will search the following databases CENTRAL, MEDLINE, and Ichushi.
I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.
I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).
I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable.
All data produced in the present study are available upon reasonable request to the authors.
View the discussion thread.
Thank you for your interest in spreading the word about medRxiv.
NOTE: Your email address is requested solely to identify you as the sender of this article.
IMAGES
COMMENTS
David L Streiner, PhD1. Path analysis is an extension of multiple regression. It goes beyond regression in that it allows for the analysis of more complicated models. In particular, it can examine situations in which there are several final dependent variables and those in which there are "chains" of influence, in that variable A influences ...
Abstract. Résumé. Path analysis is an extension of multiple regression. It goes beyond regression in that it allows for the analysis of more complicated models. In particular, it can examine situations in which there are several final dependent variables and those in which there are "chains" of influence, in that variable A influences ...
• based on literature review •"test" model on a new sample • problem is compounded in path analysis (relative to a single regression model) because testing of contributions within a single regression is not a test of the contribution of that path to the model • it is possible to find that deleting one or variables that do not
Structural path analysis (SPA) is an important method to study the transfer influence and path relationship between different factors in the production supply chain. This article mainly summarizes and analyzes the literature on the application of SPA in economy, environment and energy. First, introducing briefly the concept, model and ...
Structural path analysis (SPA) is an important method to study the transfer influence and path relationship between different factors in the production supply chain. ... Liming Chen. Structural path analysis and its applications: literature review[J]. National Accounting Review, 2020, 2(1): 83-94. doi: 10.3934/NAR.2020005. Related Papers: Abstract.
Literature reviews establish the foundation of academic inquires. However, in the planning field, we lack rigorous systematic reviews. In this article, through a systematic search on the methodology of literature review, we categorize a typology of literature reviews, discuss steps in conducting a systematic literature review, and provide suggestions on how to enhance rigor in literature ...
We then examine how path analysis is used in the accounting literature. We identify two problems with the way path analysis is used. First, path analysis is often used as a tool to disentangle direct and indirect causal effects; however, studies that use path analysis for this purpose often assume away potential endogeneity by imposing the ...
The Prevalence of Path Analyses: Literature Review Path analyses with fallible variables are relatively common in the behavioral sciences. To document the extent to which path analyses are used across a wide range of psychological subdisci-plines, we reviewed the recent issues of seven major American
This paper presents a structured literature review on Trust in AI, using only quantitative methods to select the most important papers in the area. • This paper presents a Main Path Analysis of the literature on Trust in AI, proposing a main path that highlights the most influential works and how the literature has evolved. •
To address this gap in the literature, we conducted a systematic review following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines using five databases: EBSCO host ...
As mentioned previously, there are a number of existing guidelines for literature reviews. Depending on the methodology needed to achieve the purpose of the review, all types can be helpful and appropriate to reach a specific goal (for examples, please see Table 1).These approaches can be qualitative, quantitative, or have a mixed design depending on the phase of the review.
Main path analysis. Main path analysis is a powerful tool that can identify chains of significant links in an acyclic directed network, thereby extracting the skeleton of a large and complicated directed network (Liu and Kuan 2016). By simplifying the network, it reveals the important knowledge flows in the citation network and tracks the ...
Affective Design Analysis of Explainable Artificial Intelligence (XAI): A User-Centric Perspective. The results show that affective is a viable trust calibration route for XAI and anxiety about AI, incidental emotion, perceived AI reliability, and experience using the system are significant moderators of the trust calibration process for an end ...
Types of Literature Review are as follows: Narrative literature review: This type of review involves a comprehensive summary and critical analysis of the available literature on a particular topic or research question. It is often used as an introductory section of a research paper. Systematic literature review: This is a rigorous and ...
Preliminary analysis. The descriptive statistics confirmed the assumptions of normality and multicollinearity for all study data. As shown in Table 3, there were significant correlations between all potential mediating variables and the dependent fatigue variables.In contrast, none of the independent variables (age, education, gender, years since diagnosis and disease course) were ...
The results of the literature review and screening processes are best managed by various tools and software. You can also use a simple form or table to log the relevant information from each study. ... Covidence (requires a subscription) - full suite of systematic review tools including meta-analysis; Combining Software (EndNote, Google Forms ...
Xie, J. and Tan, L. (2018) The Impact of Rising House Prices on Business Development: Literature Review and Path Analysis. Open Journal of Accounting, 7, 73-81. doi: 10.4236/ojacct.2018.71005 . 1. Introduction. Since the house reform in 1998, China's real estate market has undergone a revolutionary change.
Abstract. Path analysis is a method for explicitly formulating theory, and attaching quantitative estimates to causal effects thought to exist on a priori grounds. There are four basic kinds of path models: recursive, block, block-recursive, and nonrecursive. Because some questions can be answered only under certain path analytic structures ...
Path analysis. In this study, path analysis was used to test the hypotheses based on the results obtained from PLS analysis. The study utilised the two main criteria under the PLS model to validate and confirm each hypothesis. ... Information sharing in reverse logistics supply chain of demolition waste: A systematic literature review. Journal ...
We uncover the key aspects and activities that underpin sustainable energy development through a meticulous and systematic review of existing literature. Our research journey begins with analysing publication trends and tracing the chronological trajectory of scholarly output on sustainable energy development.
Path of Science. 2024. Vol. 10. ... Section "Technics" 1014 A Comprehensive Analysis of Sustainable Energy Development: A Review of Existing Research Christian Chukwuemeka Nzeanorue 1, Moses Sodiq Sobajo 2, Benedict Chibuikem Okpala 3, ... Development—A Systematic Literature Review. Energies, 15 (21), 8284. doi: 10.3390/en15218284 2 ...
Systematic literature review was chosen due to the abundance of research papers related to A*. With the use of systematic literature review we are able to determine the scope of our study and enable us to analyze and synthetize our research through the use of a focused research questions that needs to be answered. The literature will be ...
The construction sector is notorious for its high rate of fatalities globally. Previous research has established that near-miss incidents act as precursors to accidents. This study aims to identify research gaps in the literature on near-miss events in construction and to define potential directions for future research. The Scopus database serves as the knowledge source for this study. To ...
Trump's most direct path to 270 electoral votes would be to keep all the states he won in 2020 and flip Georgia and Pennsylvania (two states he won in 2016) back to his column.
• based on literature review •"test" model on a new sample • problem is compounded in path analysis (relative to a single regression model) because testing of contributions within a single regression is not a test of the contribution of that path to the model • it is possible to find that deleting one or variables that do not
By drawing on a systematic literature review (SLR), we examine existing studies on AI adoption in production and explore the main issues regarding adoption that are covered in the analyzed articles. Building on these findings, we conduct a comprehensive analysis of the existing studies with the aim of systematically investigating the key ...
This article is a practical guide to conducting data analysis in general literature reviews. The general literature review is a synthesis and analysis of published research on a relevant clinical issue, and is a common format for academic theses at the bachelor's and master's levels in nursing, physiotherapy, occupational therapy, public health and other related fields.
The former president is focusing his most vicious attacks on domestic political opponents, setting off fresh worries among autocracy experts.
The literature search will cover CENTRAL, MEDLINE, and Ichushi databases, including studies in English and Japanese. Bias will be assessed using QUADAS-2, and meta-analysis will be conducted using a random effects model, with subgroup and sensitivity analyses to explore heterogeneity. ... This systematic review and meta-analysis aim to ...
We used main path analysis to review the IT outsourcing literature from 1992 to 2013. ... used bibliometric analysis methods to analyze and visualize the citation network characterizing the rich body of ITO literature. The main path analysis precisely identified and visualized the major knowledge flow in the evolution of ITO research and major ...