Digitalization and Automation (D&A) of the production process enable the maintenance of real-time decision-making and implementation, from planning to realization (Osmolski and Kolinski, 2020), where minor changes in predictive performance led by automation can result in significant improvements that have different impact (Radovanović et al, 2020). Investments in technologies for D&A can reduce labor costs (OECD, 2017) and enable the efficiency of manufacturing companies (Rajković et al, 2023). Industry 4.0 (I4.0) is characterized by advanced D&A, with the use of electronics and various technologies in manufacturing processes (Lu, 2017), while the application of advanced technologies enables digital production (Rakićević et al, 2023). I4.0 can enhance the company’s process performances and drive innovation (Glogovac et al, 2022), where “smart” innovation management presents a prerequisite for better results and better acceptance of innovations, leading to better economic business results (Likar et al, 2023). Authors Pfohl et al (2017) believe that the digitalization of products and processes is a key driver of I4.0.

Industrial development of the Republic of Serbia impacts the rise of living standards and economic progress of the country, as well as competitiveness on the market (Đaković, 2024). The Digital Transformation Centre at the Serbian Chamber of Commerce and Industry of Commerce is responsible for the digital transformation of the economy in the Republic of Serbia, as well as for supporting businesses. So far, it has already implemented three programs for companies in the Republic of Serbia to improve their businesses through digital means: “Support Program for the Digital Transformation of Small and Medium-sized Enterprises (MSMEs) 2019/2020” (Chamber of Commerce and Industry of Serbia, 2019), as well as programs aimed at mitigating the consequences caused by the COVID-19 pandemic, SPEED 1.0 and SPEED 2.0 (Chamber of Commerce and Industry of Serbia, 2020). One of the Republic of Serbia’s developmental objectives is to improve the competitive potential of the industry by implementing reform processes and transition that impact the development of the industrial potential, higher investments and faster development of domestic and private sectors (Jakopin and Bajec, 2009). Digital transformation in the Republic of Serbia is of high importance in creating conditions for accelerated economic development, as well as preserving creative potential (Stefanović et al, 2021). According to Đaković (2024), in the Republic of Serbia, the connection between the traditional industry and the Information-Communication Technologies (ICT) sector is not strong, while companies invest five times less in ICT than the world average. Results from the latest available report from the Statistical Office of the Republic of Serbia regarding the application of ICT in the Republic of Serbia for 2023 (Statistical Office of the Republic of Serbia, 2023) show that 100% of the companies have a broadband Internet connection. One-fifth of the companies employ ICT experts and use the software for Enterprise Resource Planning (ERP), 10.7% of the companies use the software for Customer Relationship Management (CRM), while 5.1% use the Business Intelligence software. Cloud services are used by 37% of the companies. This indicates the company’s readiness and openness towards digital transformation.

The Ministry of Economy has developed a new strategic framework aimed at achieving improved competitiveness of the Serbian industry and strengthening its position in the international market. The framework focuses on improving coordination of industrial development policy and increasing the efficiency of instruments for its implementation (Đaković, 2024). It requires the application and design of a balanced industrial policy that will enable or improve comparative advantage with lower prices and costs for standard products (Đuričin and Vuksanović Herceg, 2018).

According to the International Monetary Fund (2023), the Republic of Serbia is one of the emerging economies and it is classified a slower-middle-income economies by the World Bank (n.d.). The main characteristics of the emerging economies are a weak commercial and industrial base, low gross national income per capita, poor infrastructure and low living standards (Roztocki and Weistroffer, 2011), as well as higher poverty prevalence than in developed economies and limited access to essential services (Hossain et al, 2023). Studying the past successes and failures can help avoid future mistakes and guide more effective technology implementation. After the dissolution of Yugoslavia, the Republic of Serbia has been undergoing a prolonged post-transition process. The consequences of unfinished and ineffectively implemented reforms impact the shaping of its institutional and economic system. Key features of this legacy include fragmented policies, institutional gaps, limited infrastructural capacities and relatively weak administrative efficiency (Stamenović, 2019). In such a context, the possibility of implementing digital transformation depends primarily on factors such as facilitating conditions and the existing level of digitalization, while individual user perceptions (e.g., social influence or effort expectancy) have a secondary role (Al-Debei and Al-Lozi, 2012; Astuti and Ayinde, 2025; Sabiteka et al, 2025). This pattern is also characteristic of other post-transition countries of the Balkans and Eastern Europe, where institutional vacuums, lack of resources and poorly developed infrastructure are the main obstacles to digital transformation (World Bank Group, 2022; Zhang et al, 2023). However, due to limited academic research on technologies and their application in the developing and emerging countries, practical experience offers valuable lessons. The research presented in this paper fills the research gap regarding the application of Industry 4.0 technology in the manufacturing companies in the Republic of Serbia. On account of all these factors, it is important to define the main determinants that impact the adoption of Industry 4.0 technology in these companies as they may reflect trends observed in other emerging economies that are similar in terms of industrialization and technological development level with the Republic of Serbia, especially in Eastern Europe and the Balkan region (such as Bosnia and Herzegovina, Montenegro, North Macedonia, a part of Croatia, Slovenia, Greece, Romania and others). Understanding the determinants that impact technology adoption in the emerging economies is essential, as they differ from developed countries due to variations in digital literacy, infrastructure, social impacts and trust levels (Dwivedi et al, 2020).

The main technologies and concepts that are the most important for D&A are, among others: Internet of Things (IoT), Artificial Intelligence (AI) and Big Data (BD) (Abdallah et al, 2021). All these technologies increase the level of manufacturing companies’ D&A, where these technologies are mostly applied in finance and accounting (60.5%), planning and control (56.5%) and production (50.8%) (Čater et al, 2021). The production capacity and ability to store and share information are enabled by the application of IoT, AI and other technologies (OECD, 2017). IoT presents a technology that connects physical devices and different industrial objects to exchange data with other devices, widely applied in different industries and is considered one of the most applied technologies under I4.0 (Khang et al, 2024). In recent years, AI has been recognized as a strategically important technology of I4.0 that enhances and increases productivity (Zhang and Peng, 2025) and has been adopted by many industries, mostly in manufacturing, for business performance improvement (Dinmohammadi, 2023). The integration of AI and IoT enables “smart” applications that automate actions, monitor activities, manage tasks, predict outcomes and enhance decision-making in manufacturing and other sectors (Khang et al, 2024).

I4.0 involves equipment and machines that generate large and complex data. BD encompasses rapidly growing datasets that are difficult to manage with conventional database systems and tools while considering data volume, velocity and variety (Obitko and Jirkovský, 2015). The growing demand for safe and efficient “smart” manufacturing, along with new technological advancements, drives the adoption of IoT, AI and BD in companies, leading to the introduction of these technologies into “smart” manufacturing companies (Jagatheesaperumal et al, 2021). The benefits of implementing these technologies in manufacturing are undeniable. However, a question emerges: What are the factors that impact the adoption of these technologies in manufacturing companies? Due to the aforementioned reasons, this paper deals with these three technologies (IoT, AI and BD) for which an expanded Unified Theory of Acceptance and Use of Technology (UTAUT) model is proposed and tested. The UTAUT model is expanded by adding more determinants that more precisely quantify the technology adoption. Having in mind the specificities of each of these technologies, the expanded UTAUT model is tested on each of the three selected technologies to provide technology-specific insights. All three models are validated and compared. The ensuing conclusions are drawn, followed by the recommendations for the implementation of each technology. The data on which the models were tested was gathered from the manufacturing companies in the Republic of Serbia. Therefore, the main research question (RQ) of this paper is: Which determinant(s) impact technology (IoT, AI and BD) adoption and application in the manufacturing companies in the Republic of Serbia? Various stakeholders would gain benefits from the provided insights into technology adoption.

The paper consists of seven sections. After the introductory section, the second section outlines the devised and widely applied models for technology acceptance, focusing on their shortcomings. The third section presents the hypotheses development. The fourth section presents the research background and the instrument development. The fifth section presents the research results, survey and hypotheses results, for each technology. The sixth section presents the discussion of results, practical and theoretical contributions of the paper, hypotheses results, per hypothesis, and the response to the RQ. The seventh section outlines the conclusion, limitations and implications for future research.

To achieve success and reach defined objectives, companies need to use modern technologies and information systems. These technologies impact employees (Xu and Suzuki, 2025). To analyze technology adoption and factors that impact their implementation and acceptance, many different models and theories have been defined. Some of them are focused on factors that impact acceptance and technology adoption, while authors Venkatesh et al (2003) have created a specific UTAUT model that integrates elements of eight models for user acceptance. This model includes Performance Expectancy (PE), Effort Expectancy (EE) and Social Influence (SI)—three direct determinants of the intention to use that determine Behavioral Intention (BI), which are essential in the determination of user acceptance and user behavior, and two direct determinants of usage behavior, such as BI and Facilitating Conditions (FC) (Venkatesh et al, 2003).

Following the main objective of the paper and focusing on the impacts of technology acceptance to define key determinants, the authors had to determine which model is more appropriate for the analysis—the Technology Acceptance Model (TAM) or UTAUT. Both models were developed to better understand the reasons behind users’ acceptance or rejection of technology and to predict the adoption or rejection of new technology (Ammenwerth, 2019). TAM was designed to explain user acceptance in the context of organizational and task-oriented information technology, while its application for non-organizational and non-task-oriented circumstances is questionable, leading to the creation of the UTAUT, a new, improved model for technology acceptance (Kim, 2014). Even though the primary goal of TAM is to understand users’ perceptions and adoption of technology (Buabeng-Andoh and Baah, 2020), TAM was criticized for simplicity (Bagozzi, 2007), inability to incorporate external variables (Dishaw and Strong, 1999) and the omission of human and social factors (Davis, 1989b). Compared to TAM, UTAUT has superior predictive ability and a more comprehensive theory (Okumus et al, 2018). TAM was designed in 1986 (Davis, 1989a), while UTAUT was created in 2003 (Venkatesh et al, 2003), based on eight already existing models for technology acceptance. The UTAUT is widely used for the evaluation of technology usage, explaining 70% of usage (Tosuntaş et al, 2015). Additionally, the authors Ling et al (2011) conducted research through comparative analysis of TAM and UTAUT models and concluded that the UTAUT provides more accurate predictions compared to TAM.

However, the main limitations of both models should be mentioned. TAM is considered a simplified model that focuses on perceived usefulness and perceived ease of use, leading to the omission of complex variables that impact technology adoption (Lee et al, 2025), as well as the omission of determinants such as education or age, that could impact the technology acceptance (Zahid et al, 2013). The model is inadequate for practical use in companies with defined rules and regulations (Ajibade, 2018). The UTAUT model is considered to be too complex because it lacks individual characteristics, i.e., the characteristics that participate in the behavior and describe the dispositions of users that could be significant in explaining their behavior; moreover, that determinants of the original UTAUT model may lack universal applicability (Dwivedi et al, 2017). The UTAUT model can consist of some irrelevant determinants (Bervell and Umar, 2017) and some original UTAUT determinants become less significant by introducing new determinants (Blut et al, 2022).

Considering both advantages and limitations of these models, the authors of this paper concluded that the UTAUT model, i.e., its expanded version with the incorporated new determinants to predict behavioral intention and technology application (Venkatesh et al, 2016), is more suitable for the analysis in this paper than TAM. The UTAUT model is a widely accepted model that can be used as a supporting tool for strategy creation and decision-making (Al-Saedi et al, 2020).

Due to the limited academic research on information technology (IT) in developing and emerging countries, practical experience offers valuable lessons, while studying past successes and failures can help avoid future mistakes and guide more effective IT implementation (Roztocki and Weistroffer, 2011). The following will present the hypotheses (H) for all determinants that impact the intention to use the selected technologies (IoT, AI and BD) in the manufacturing companies in the Republic of Serbia. Additional determinants used for the expansion of the UTAUT model have been identified in other research as significant determinants in the process of technology adoption. For this reason, their inclusion in the model is empirically and theoretically justified. Because the results of the three UTAUT models will be presented, distinguished by technology, “name” in the defined hypotheses will be changed into the name of technology, i.e., IoT, AI or BD.

3.8 Barriers (B)

There are many barriers that negatively impact the adoption of technologies for D&A in production processes, such as financial costs and regulatory uncertainty (Fu et al, 2018), disapproval by employees, lack of recognition for construction, limited involvement of employees in decision-making, the cost of investments in machines and equipment (Ediriweera and Wiewiora, 2021), the lack of user knowledge, engagement, participation, training, technical skills and expertise, as well as dissatisfaction and resistance to change (Chouki et al, 2020). Therefore, H11 can be defined: H11. Barriers (B) have a negative impact on the BI to use technologies for D&A in manufacturing companies.

The expanded UTAUT model for this research, with hypotheses, is presented in Fig. 1, where “plus” (+) presents a positive, while “minus” (–) presents a negative impact of the determinant on the BI. Also, since all three technologies are digital technologies that can be applied to the D&A of production processes, hypotheses H5–H11 refer to all these technologies. To verify the proposed conceptual model, Structural Equation Modelling (SEM) will be employed. SEM is the widely used and acknowledged analysis for conceptual model verification (Kline, 2023). Having in mind the sample size and model complexity, Covariance-Based (CB-SEM) modelling was employed using AMOS 22 (IBM AMOS, Version 22.0, IBM Corporation, Armonk, NY, USA). Descriptive statistics and scale reliability were explored using SPSS 25 (IBM SPSS Statistics, Version 25.0, IBM Corporation, Armonk, NY, USA).

Fig. 1.

Research model - expanded UTAUT model with hypotheses for technology adoption in manufacturing companies. UTAUT, Unified Theory of Acceptance and Use of Technology; PE, Performance Expectancy; EE, Effort Expectancy; SI, Social Influence; FC, Facilitating Conditions; ADA, Attitudes towards Digitalization and Automation; AS, Awareness and Support; CLDA, Current Level of Digitalization and Automation; CSWDAA, Customer, Supplier and Warehouse Digitalized and Automated Activities; SAC, Support and Abilities to Change; EIF, External and Internal Factors; IF, Internal Factors; EF, External Factors; B, Barriers; BI, Behavioral Intention. “plus” (+) presents positive impact of the determinant on the BI; “minus” (–) presents a negative impact of the determinant on the BI.

This section presents a research background regarding manufacturing companies and D&A in the Republic of Serbia, as well as the development of the research instrument used in this research—a survey.

This section presents the results of the survey conducted in the manufacturing companies in the Republic of Serbia, descriptive statistics and sample analysis, as well as the results from the expanded UTAUT model applied to the three selected technologies. The sample size was 330 people, i.e., 330 companies, given that one person per company was contacted. A total of 228 out of 330 people, i.e., companies, participated in the anonymous online survey, that is a 69.09% response rate. This can be considered a high response rate, since “the average online survey response rate is 44.1%” (Wu et al, 2022). The published articles on the topic of the application of digital technologies in manufacturing companies had a smaller number of surveyed manufacturing companies, with 100 manufacturing companies and scientific research institutes in Poland (Woźniak et al, 2021), 103 manufacturing companies in Italy (Zheng et al, 2020), 219 manufacturing companies from automotive and allied industries in South Africa (Bag et al, 2021), and others. The results of the research presented in this paper can be considered relevant, given that the representativeness of the sample is high, above the average response rate.

This study, based on the three expanded UTAUT models, provides a deeper understanding of how various determinants impact the adoption of three technologies for D&A of production processes (IoT, AI and BD) in the manufacturing companies in the Republic of Serbia. The results from 228 surveyed manufacturing companies in the Republic of Serbia show that PE, FC and CLDA had a significantly positive impact on BI to use all these technologies. The research findings are discussed in detail, per hypothesis. The research results in this paper have provided practical and theoretical contributions. Regarding the theoretical contributions, the first is a presentation of the theoretical background of D&A, I4.0, their connection and related technologies. The second is an explanation of models for technology acceptance, pointing out differences between the UTAUT and TAM models, as the two most applied models that are mostly used for the measurement of technology acceptance. Regarding practical contribution, the first is that this paper presents the results of the survey conducted in 228 manufacturing companies in the Republic of Serbia, with a very high response rate (almost 70%) that indicates the significance of the results and the current situation in the manufacturing companies in this country. The second is that, based on the survey, the additional determinants of the expanded UTAUT model for the adoption of three technologies are defined. The third is that this paper points out the main determinants that positively and negatively impact the intention to use these technologies. The following are presented hypotheses results and suggestions for improving the impact of determinants on BI, for all three technologies. H1. PE has a positive impact on the BI to use “name” technology in manufacturing companies. The first hypothesis for all three technologies is supported, i.e., PE has a significant positive impact on BI to use IoT, AI and BD, while it has the strongest impact on BI of BD. This supports the results in previous studies, i.e., for IoT (Xu and Suzuki, 2025), AI (Gansser and Reich, 2021) and BD (Cabrera-Sánchez and Villarejo-Ramos, 2020) supporting H1. The higher the PE, i.e., performance efficiency of technologies, the higher the BI to use these technologies. This result shows that the respondents believe that the implementation of these technologies will provide strong performance and that they are aware that focusing on the implementation of IoT, AI and BD will improve performance in production processes. H2. EE has a positive impact on the BI to use “name” technology in manufacturing companies. The impact of the determinant EE on BI was not supported in all three models. The higher the EE, i.e., ease of using BD, the lower the BI to use such technology, that supports the results in previous studies, IoT (Li et al, 2024), AI (Gohil, 2023) and BD (Brünink, 2016). This does not support H2, showing that respondents consider the application of these technologies complicated and complex. Therefore, the ease of using these technologies should be highlighted, enabling simplified instructions, as well as improved training, so the employees in the manufacturing companies in the Republic of Serbia could understand that without big effort they can use IoT, AI and BD in production processes. H3. SI has a positive impact on the BI to use “name” technology in manufacturing companies. The results supported hypothesis H3 for AI and BD, but not for IoT. The results show that SI has a positive impact on BI to use AI and BD, that supports the results of previous studies, i.e., for AI (Gohil, 2023) and BD (Cabrera-Sánchez and Villarejo-Ramos, 2020). For IoT this supports the results from (Li et al, 2024). The higher the SI, i.e., social influence, the higher the BI to use AI and BD. For technologies AI and BD, this supports H3, showing that the respondents would implement these technologies if they had a recommendation or influence from different social groups, such as colleagues, friends or other companies. Word of mouth or previous experience from other people, colleagues or companies, cooperation within an industry or with other manufacturing companies and group motivation for technology application could increase BI to use such technologies (Ronaghi and Forouharfar, 2020). However, this does not apply to the IoT. By increasing transparency, organizing training, enabling implementation support and exchanging experiences with IoT users, it will positively impact on increase of BI to use such technology. H4. FC has a positive impact on the BI to use “name” technology in manufacturing companies. The FC is the second determinant whose impact on BI was confirmed in all three models. The results show that FC has a positive impact on BI to use IoT, AI and BD, which is consistent with the results of the previous studies, i.e., IoT (Ronaghi and Forouharfar, 2020), AI (Li et al, 2024) and BD (Brünink, 2016), while it has the strongest positive impact on BI of AI. This supports H4, showing that the respondents believe that providing facilitating conditions enables the application of these technologies. The higher the FC, i.e., facilitating conditions to use these technologies, the higher the BI to use them. Therefore, companies should provide all the necessary conditions, resources and support to use such technologies. H5. Attitudes to Digitalization and Automation (ADA) of production processes have a positive impact on BI to use technologies for D&A in manufacturing companies. The first determinant whose impact was not detected in any of the models is ADA. The attitudes towards D&A do not have an impact on the BI to use IoT, AI and BD, showing that employees might not be aware of the possibilities that these technologies could bring. They should be more open to introducing new technologies in the production processes, as well as be aware of the improvements they bring. Improved and more frequent information, education, knowledge sharing and management engagement about the opportunities that D&A can bring to manufacturing companies could positively impact BI to use these technologies. A possible reason for this result might be that the majority of the surveyed companies in the Republic of Serbia are not the users of digital technologies, therefore, they lack knowledge of their possible benefits (Kutlača et al, 2020). Companies can believe that they do not need any more D&A because their current level of D&A is sufficient. Also, the fear of making systems more complex, replacing employees with technology, increasing the complexity of processes, as well as focusing on current objectives and strategies, can decrease the intention to use these technologies. D&A levels of the manufacturing companies in the Republic of Serbia should be increased to enhance awareness and an open attitude toward their application. H6. Awareness and Support (AS) of the management of the company have a positive impact on the BI to use technologies for D&A in manufacturing companies. Determinant whose impact was not supported by models is AS. The higher AS, i.e., awareness and support of the management of the company, the lower BI to use IoT, AI and BD and the measured impact is not statistically significant. Such a result was not expected. The companies whose management is unaware of the potential improvements that the application of technologies for D&A could bring are unlikely to support their implementation. Expanding on the reasons mentioned in the previous determinant, some of the additional reasons could be the fear of high implementation costs, the lack of state support, infrastructure and connection problems, the fear of nonacceptance by employees and others. Better information gathering, information exchange with market competitors and better market insights should lead to improved AS, that could impact an increase in BI to use these technologies. H7. The Current Level of Digitalization and Automation (CLDA) in production processes in a manufacturing company has a positive impact on BI to use technologies for D&A in manufacturing companies. The third determinant, whose impact was supported in all three models, is CLDA. The results show that CLDA has a positive impact on BI to use IoT, AI and BD, while it has the strongest impact on BI of IoT. The higher the CLDA, i.e., the current level of D&A in manufacturing companies, the higher the BI to use them. This supports H7. Therefore, companies that have already turned towards D&A and have processes that can be digitalized and automated are more open to applying technologies and are aware of the improvements these technologies bring. H8. Customer, Supplier and Warehouse Digitalized and Automated Activities (CSWDAA) have a positive impact on the BI to use technologies for D&A in manufacturing companies. The results show that the impact of CSWDAA was only supported in the model for AI. The higher CSWDAA, i.e., customer, supplier and warehouse digitalized and automated activities, the higher the BI to use AI. A positive result was expected since the main stakeholders with everyday communication and collaboration can push innovations in the manufacturing companies and processes. Automated activities in warehouses can decrease mistakes, reduce time, improve inventory management, decrease the number of employees responsible for warehouse monitoring and organizing and enable quality standards. Customers can request changes and improved products that lead to the implementation of technologies, as well as suppliers that have already implemented technologies that lead to faster order fulfilment, showing positive technology implementation results. Customers and suppliers, as part of the supply chain, can have a great impact on manufacturing companies, pushing for changes and improvements with enhanced information sharing and exchange (Pfohl et al, 2017). The reasons for not supporting H8 for IoT and BD are linked to the reasons for determinants ADA and AS. Companies should strive to have automated warehouse management activities and digitalized and automated communication with customers and suppliers, to prevent or reduce mistakes, failures and time and enable data integration and real-time data monitoring with accurate and precise data. H9. External factors (EF) have a positive impact on the BI to use technologies for D&A in manufacturing companies. The hypothesis H9 is not confirmed, as there was no statistically significant impact of EF on BI on any of the models. External factors such as market conditions, competitors, state encouragement and support, and the impact of customers and suppliers, should lead to the initiatives to introduce innovations in production processes. The non-significant impact on all three models can be explained by the lack of technology applications at customers and suppliers, that cannot impact the company to use such technology. The absence of subsidies cannot impact the company from introducing and using such technology. There is also a possibility that competitors do not use such technologies or do not consider them useful for a manufacturing company, so they do not push their implementation in other companies. H10. Internal factors (IF) have a positive impact on the BI to use technologies for D&A in manufacturing companies. The impact of IF was only detected in the model of BD. The higher IF, i.e., internal factors in the company, the lower BI to use BD, that conflicts with (Damanpour et al, 2018), where IF had a positive impact. The observed results can be explained as follows. The management of the company can fear changes or cannot motivate employees enough to use such technologies, while the employees are usually afraid of the unknown and new, as well as of being replaced, leading to them not having a willingness to learn new technologies, especially in situations when they are used to their standard way of working (Prentice, 2022). Employees are afraid of losing control or even losing their jobs, they are resistant to changes, considering that technology implementation is highly costly, complex and takes a long implementation time (Ivanov et al, 2020). These are the reasons why the internal factors do not impact positively on BI, i.e., H10 is not supported. Changing some of them should positively impact the BI to use these technologies. IF such as employees, management of the company, innovations and changes in production products should lead to innovations in production processes. H11. Barriers (B) have a negative impact on the BI to use technologies for D&A in manufacturing companies. The final hypothesis H11 is not confirmed, as there was no statistically significant impact of B on BI in any of the models. It was not expected, since barriers such as high implementation costs, the lack of financial support and standardization, the problems with the Internet and electricity supply, employees’ resistance to change, missing support from the state and government, the problems with data security or leaking data outside the organization and others, usually prevent the use and implementation of technology and have a negative impact on technology application (Jones et al, 2021). Different types of challenges require different methods of prevention, such as (Albukhitan, 2020): traditional processes—introducing a modern digital solution that will convert paper-based processes into digital; resistance to change—companies should reduce stress and time-consuming processes, improve efficiency, enable remote work and transparent communication about technology to all employees; budget restrictions—planning of investment processes; absence of relevant knowledge—if existing knowledge is insufficient, engagement of new employees or external consultants, not limited knowledge only to some employees, but to the whole organization; security—protection layers and mechanisms must be set to protect and secure the leaks of data, recognition and document vulnerability issues and others. By applying some prevention methods, it could be possible to decrease or completely remove some of the barriers that impact the adoption of these technologies.

Overall, the supported hypotheses for all three technologies are: H1, H4 and H7, while H2, H5, H6, H9, H10 and H11 are not supported. The hypothesis H3 can be considered as supported, since it is supported for two technologies, while H8 can be considered as not supported, because it is supported only for one out of the three technologies. These insights are presented in Table 7. Observing the determinants and their importance on technology adoption, the determinant PE is significant for all three technologies, while EE is insignificant for those technologies, SI is significant for the adoption of AI and BD technologies, but not for IoT; FC is significant for all three technologies, ADA and AS are not significant for any technology, CLDA is significant for all three technologies, CSWDAA is significant for AI, but not for IoT and BD, while the determinants EF, IF and B are not significant for the adoption of any of these three technologies.

Based on the results presented in this paper, the response to RQ Which determinant(s) impact technology (IoT, AI and BD) adoption and application in the manufacturing companies in the Republic of Serbia? can be given. According to the results from the expanded UTAUT models, applied for the three technologies (IoT, AI and BD) separately, the positive impact on these three technologies’ adoption and application was found from three determinants: PE, FC and CLDA. The highest positive impact on BI to use all three technologies had the FC determinant, which indicates that the company should provide the necessary resources (knowledge, technology, tools, devices) to use these technologies. Regarding the negative impacts, two determinants differ by technology, EE and IF. Regarding IoT technology, none of the determinants had a negative impact. When it comes to the AI, the most negative impact on BI to use this technology has determinant IF. This can be caused because of uncertainty in working with technologies, insufficient familiarity with technologies and readiness to use them, fear that technologies will replace people in production processes, as well as that they will impact the reduction of human labor. To decrease the negative impact of this determinant some of the recommendations could be that the management work towards motivating employees, enabling training and support within the introduction and even implementation of these technologies, as well as the gradual introduction of technology, being realistic and open-minded regarding the complexity of use, but also focused on positive and improved business results that these technologies could contribute to. The most negative impact on BI of BD to use this technology had determinant EE. This can be caused by technological complexity, the lack of required knowledge and skills, missing introduction of how to apply this technology in everyday work, and technical barriers. To decrease the negative impact, similar recommendations as for the previous determinant can be applied, as well as the introduction of simplified solutions, technologies and software that would lead to an easier and more accepted introduction of BD in the production process.

Based on the presented results, different strategies for technology adoption can be defined. For all three technologies, strategies for technology adoption should be an investment in the infrastructure of the company, enabling support and resources (equipment, hardware, software, devices and others), introducing Key Performance Indicators (KPIs) and benchmarking to follow and improve performance, enabling training, demonstrations and real experiences for employees to become better acquainted with technologies as well as to understand the numerous and significant improvements these technologies can introduce into daily operations. Additionally, for the IoT technology, a strategy should be leaning toward more digitalized and automated processes and solutions that enable the connection with IoT. Regarding BD and AI, the strategies should focus on connecting employees and their networking, in person or online, organizing mentoring sessions, presentations and experience sharing, where more experienced colleagues can present their knowledge and advice regarding the application of these technologies, presentations of real case studies. Additionally, strategies should be focused on alleviating employees’ resistance to change and not creating pressure to use such technologies from other employees or superiors. They should have transparent communication and support towards changes and be involved in changes. To ensure a more successful adoption of BD technology, employees should be more motivated, interested and included in technology implementation, so the strategies should be developed in that direction.

The main objective of this paper is to identify key determinants that influence the use of widely adopted technologies for D&A (I4.0) in manufacturing companies, through the expanded UTAUT model. The herein proposed extended UTAUT model, with the specific determinants, to the best of our knowledge, was not previously applied, leading to the originality of the research and its results. Each of the chosen three technologies has its specificities, where different determinants can have positive or negative impact on their acceptance in manufacturing and digitalization. This paper presents empirical evidence confirming the validity of the expanded UTAUT model in explaining the adoption of three technologies within manufacturing companies and pointing out PE, FC and CLDA, as the determinants that positively impact the BI to use these three technologies, where determinant FC has the highest impact. Since the institutional vacuums, poorly developed infrastructure and problems regarding the lack of resources are the main obstacles to digital transformation in post-transition countries (World Bank Group, 2022; Zhang et al, 2023), the findings of this paper can be generalized and applied to other post-transition countries of the Balkans and Eastern Europe, because of the similar infrastructural, institutional and digital challenges with the Republic of Serbia. This paper is of high value for decision-makers in manufacturing companies and provides information on the factors that mostly impact technology acceptance for IoT, AI and BD. Additionally, it presents the current level of application and implementation of these technologies in the Republic of Serbia. The main results of this research can be used by managers, decision-makers and practitioners in the manufacturing companies in the Republic of Serbia. These findings highlight the key determinants that impact technology adoption, helping them make decisions about which technologies to implement and how to design the most effective implementation strategy based on the determinants that impact the intention to use IoT, AI and BD technologies.

However, there are some limitations of this study. The first limitation is that the survey was conducted in the Republic of Serbia, so the results can be compared only with the results from the countries that have similar economic and social status and market conditions, i.e., developing countries. The second limitation which might arise is the number of determinants with which the UTAUT model is extended. As several added determinants (e.g., ADA, AS, EF, IF, B) consistently show no statistical significance, one might consider their removal in the future model alterations. It is possible to add other determinants that can more effectively impact the BI or explain the adoption of these technologies or even simplify the model by excluding the non-significant determinants or combine or regroup the existing ones. This would enable the model to focus more on core determinants to improve the interoperability of the results. Having in mind that no similar studies have been conducted in the region, a complex model has been proposed as a starting point that can be modified and simplified as needed. The third limitation is that some of the determinants were not statistically significant in the model. Even if some determinants were initially statistically insignificant, they may indicate directions for future expansion of the model. The future research directions would be to compare the results from this study with similar studies. This would lead to the conclusion of which technology would have the highest adoption and acceptance rate, pointing out which determinants impact adoption and acceptance the most. Another future research direction is to conduct a survey in the surrounding countries, to explore the application of technologies and technology acceptance in the countries with similar social, cultural and economic factors, such as countries in the Balkan region and to generalize the model to be used in other countries. The third direction for future research is to focus on the empirical testing and further development of additional determinants that could improve the explanatory power of the UTAUT model in the context of production management. The fourth direction for future research is to focus more on the organizational level, leading to a smaller sample and a less complex model.

The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

TR designed the research study, performed the research and wrote the initial version of the manuscript. TR and SR analyzed the data and did a discussion structure. DLC and MS contributed to the analysis and interpretation of the results. All authors have read, critically reviewed and approved the final manuscript. All authors contributed to editorial changes in the manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

This research received no external funding.

The authors declare no conflict of interest.