Exploring the implementation and use of smart maintenance technologies in the manufacturing industry: insights from industrial cases

The increasing competition and development within the manufacturing industry have led manufacturing companies to strive for new production systems, strategies, processes, machinery types and maintenance practices (Frost et al., 2019; Stefanini et al., 2023). The importance of the maintenance function has grown alongside increasing demands on production system productivity, availability, safety, quality and customer satisfaction (Arnaiz et al., 2010; Manenzhe et al., 2023).

From a high-level perspective, maintenance has moved from Corrective Maintenance in Industry 1.0 to Predetermined Maintenance in Industry 2.0, and Condition-Based Maintenance (CBM) in Industry 3.0 (Moubray, 1997). Industry 4.0 deals with Information and Communication Technologies (ICT), intelligent factories and the development of Internet and embedded system technologies (Liu and Xu, 2017). It has been described in many ways, such as: (1) a combination of Information Technology (IT), manufacturing processes, and the internet, (2) digital transformation, (3) the fourth industrial revolution and (4) smart factories (Matt et al., 2021). Nine core technologies are presented in Industry 4.0: (1) Industrial Internet of Things (IIoT), (2) Big Data and Analytics, (3) Augmented Reality (AR), (4) Simulation, (5) Autonomous Robots, (6) Additive Manufacturing (AM), (7) Cyber Security, (8) Cloud Computing and (9) System Integration (Vaidya et al., 2018). Furthermore, in Industry 4.0, the physical world of manufacturing processes, including humans, machinery and sensors, is integrated with the cyber world, including advanced software for prediction, control and maintenance (Bajic et al., 2021).

In Industry 4.0, a new generation of maintenance is presented (Tortora et al., 2025). Earlier published research studies present many ways of approaching the new generation of maintenance, including Smart Maintenance (Bokrantz et al., 2020a, b), Self-Maintenance (Singh et al., 2013), Maintenance 4.0 (Cachada et al., 2018), E-maintenance by (Arnaiz et al., 2010) and eMaintenance (Kajko-Mattsson et al., 2010; Karim et al., 2018). Additionally, the new generation of maintenance can be described by the following levels: (1) detection of failure: is something wrong?, (2) diagnosis of failure: what is wrong?, (3) health assessment: how wrong is it?, (4) prognosis: when is it expected to go wrong? and (5) prescriptive maintenance: what action to do when in order to mitigate it from failing (Tinga, 2025; Giacotto et al., 2025).

Moreover, the new generation of maintenance has been described through four dimensions: (1) data-driven decision-making, making decision based on maintenance data, (2) human capital resources, the competence development of maintenance employees, such as analytical skills and ICT skills, (3) internal integration, the integration of maintenance function with other internal functions, such as production and IT and (4) external integration, integration of maintenance function with external functions, such as machine and sensors suppliers and builders (Bokrantz et al., 2020a). Furthermore, the nine core technologies of Industry 4.0, as well as the technologies Cyber-Physical System (CPS) and Artificial Intelligence (AI), play an essential role in the development of smart maintenance technologies and in improving traditional maintenance (Silvestri et al., 2020; Bengtsson and Lundström, 2018; Al-Najjar et al., 2018). For instance, AI has been shown to be important for predictive maintenance of machine tool systems based on machine condition data (Lee et al., 2019). Different types of AI models, such as the Isolation Forest (Liu et al., 2012), can be used to detect anomalies in an unsupervised maintenance dataset (Giliyana, 2025). CPS have been used to develop a smart maintenance model (Al-Najjar et al., 2018). However, traditional maintenance competences will, for the foreseeable future, be necessary to carry out the active maintenance actions identified by the measurements. Measurement and analysis alone will never be enough. Maintenance actions are what will finally reduce breakdowns and other wastes (Bengtsson and Lundström, 2018).

Industry 4.0 was introduced at the Hanover Fair in 2011, yet the manufacturing industry still faces many challenges in implementing and using smart maintenance technologies (Matt et al., 2020). These challenges include: (1) start-up cost and uncertain return on investment, (2) Cyber Security issues when connecting machines using IIoT through different types of communication protocols and Application Programming Interfaces (APIs), (3) implementation resources, (4) competence when implementing and using such technologies and (5) the knowledge of what to monitor, and what type of data to collect in order to make proper decisions (Giliyana et al., 2022, 2023). Challenges arising from the lack of a fully developed strategy for implementing Industry 4.0 technologies have also been discussed (Flores et al., 2018). Technician skill challenges, IIoT implementation challenges and challenges related to Big Data and Analytics have also been highlighted (James et al., 2022; Silvestri et al., 2020).

From the industry's perspective, a clear view of the steps an organization should follow to implement smart maintenance technologies is lacking (Silvestri et al., 2020). In addition, a review of 82 published papers within maintenance found that only three had direct practical links to industry (Fraser et al., 2015). Many implementation challenges, including technological challenges in System Integration, have also been investigated (Lundgren et al., 2021; Kans and Galar, 2017). System integration is one of the most important Industry 4.0 technologies, since Industry 4.0 represents a fusion of diverse technologies, and implementing these technologies involves integrating various systems (Pozzi et al., 2023). To enable more efficient data-driven decision-making, it is crucial that these systems are seamlessly integrated and can communicate effectively across diverse protocols and APIs (Rikalovic et al., 2021; Giliyana et al., 2024a).

Although several approaches for smart maintenance have been investigated, a gap remains between earlier research and industrial practice (Fraser et al., 2015; Silvestri et al., 2020). As a result, manufacturing companies receive limited practical guidance on how to proceed. Therefore, this research explores how smart maintenance technologies are implemented and used in large manufacturing companies, based on empirical findings from 11 case companies. The aim is twofold: (1) to increase the understanding of smart maintenance technologies in an industrial setting and (2) to provide practical recommendations for manufacturing practitioners. Rather than outlining an implementation process, this research provides industrial insights that bridge the gap between research and practice.

This section elaborates on how IIoT, Big Data and Analytics, Cloud Computing, System Integration, AI and CPS can be applied in maintenance contexts, as well as the challenges, enablers, and benefits associated with such technologies. Challenges and enablers are interpreted along three themes: organizational, human and technical. Implementing smart maintenance technologies requires attention to all three, since obstacles and solutions are not only technical but also organizational and human. Enablers are practices reported in the literature that support implementation and address challenges within the same themes. Benefits are the operational and strategic improvements that follow from implementing such technologies, as discussed in Section 2.1. Tables 1 and 2 summarize the challenges and enablers, organized along these themes.

IIoT enables connectivity and maintenance data collection across APIs and standard communication protocols, such as Open Platform Communications Unified Architecture (OPC UA) and Message Queuing Telemetry Transport (MQTT), which have facilitated the development of Big Data and Analytics to support AI-driven maintenance decisions (Silvestri et al., 2020; Silva et al., 2021; Lee et al., 2019). These technologies support maintenance trend analysis, condition monitoring, inspection, and fault analysis (Silvestri et al., 2020; Forcina et al., 2021). Furthermore, such technologies can overcome the limitations of Predetermined Maintenance, such as varying Mean Time Between Failures (MTBF), limited Original Equipment Manufacturer (OEM) maintenance experience and potential bias toward spare part sales, by facilitating data-driven maintenance planning (Ahmad and Kamaruddin, 2012; Labib, 2004; Silvestri et al., 2020; Forcina et al., 2021).

A CPS-based smart maintenance model has been developed that enables automated data collection, condition-based maintenance recommendations, automatic maintenance actions and reporting of required manual tasks to the maintenance department (Al-Najjar et al., 2018). AI models further support this by enabling Predictive Maintenance through machine condition data (Enshaei et al., 2024; Lee et al., 2019).

Beyond automated condition monitoring, making maintenance data accessible across an organization is equally important. In this case, Cloud Computing enables maintenance data sharing across departments and supports maintenance tasks, such as providing operators with real-time machine monitoring, issue notifications and direct contact with maintenance personnel (Campos et al., 2021). A cloud-based Computerized Maintenance Management System (CMMS) provides accessible maintenance data for engineers, managers and technicians via laptops and smartphones (Wienker et al., 2016; Chang et al., 2016).

For these technologies to function as an interconnected system, seamless communication between data-collection hardware and software via APIs and standard communication protocols is essential. One example is integrating a vibration measurement system with a CMMS to automatically generate maintenance work orders (Giliyana et al., 2024a; Agerskans et al., 2025).

Many challenges associated with implementing and using smart maintenance technologies have been identified, organized into three themes: organizational, human and technical (Giliyana et al., 2025). The organizational theme presents the categories cross-functional team, and time and resources, such as alignment between organizations, defining responsibilities, getting support from the IT department, resources, and finance. The human theme includes the categories competence and acceptance, such as finding people with the right skills, and resisting knowledge upgrades and the use of new technologies. Within the technical theme, challenges concern data management and communication, including data storage, analysis and visualization. Many other challenges have also been presented in earlier research, such as: (1) the ambition level of an organization and the available amount and quality of data and knowledge, (2) lack of useful and relevant data in industrial practice, (3) lack of labelling of registered failures in the CMMS due to lack of knowledge, (4) lack of condition measurements, which results in poor assessment of machine health and prognostics and (5) no threshold value indicating when to take maintenance actions (Tinga, 2025). Human-related challenges have also been highlighted, such as (1) lack of knowledge, (2) limited teamwork between departments, (3) low interest in learning and knowledge upgrade, (4) lack of technical skills and (5) weak communication between managers and operators (Nasirinejad et al., 2024). Furthermore, about 20–50% of breakdowns in the studied companies have been found to be caused by human errors (Salonen, 2023). Several technical-related challenges have also been emphasized, including limited sensor availability for maintenance data collection, data privacy issues and failure detection (Nasirinejad et al., 2024). Table 1 summarizes the challenges presented in previous research, organized into the themes of organizational, human and technical.

Several enablers for the implementation and use of smart maintenance technologies are presented in earlier research. These include, for example, building cross-functional teams, good communication between departments, education and training, and pilot projects (Giliyana, 2023; Bengtsson, 2008). Further enablers for smart maintenance technologies have been highlighted, such as clear vision and communication, Industry 4.0 training that shows benefits for both company and employees, and aligning strategic goals across departments, as implementing smart maintenance involves multiple departments (Salonen, 2009; Bajic et al., 2021; Lundgren et al., 2022). Other enablers include using external sensors independent of the machine's Programmable Logic Control (PLC) to collect maintenance data from legacy machines, and adhering to one standard communication protocol to ease System Integration (Giliyana et al., 2025). Moreover, maintenance research often focuses on condition monitoring, remote services, wear modelling, Remaining Useful Life and failure prediction, but implementation also requires focus on human and organizational support (Lundgren et al., 2023). Enablers for AI-based data analysis have also been listed, such as understanding the problem and applying industrial AI accordingly, understanding the system to collect correct and high-quality data, understanding the physical meaning of parameters and their relation to system behavior, understanding how parameters vary across machines and also focusing on areas not purely technical (Lee et al., 2018). The importance of domain knowledge when applying AI to maintenance has also been emphasized (Bengtsson et al., 2024). Table 2 summarizes the enablers, organized into themes of organizational, human and technical.

This article is based on an exploratory case study (Yin, 2018; Karlsson, 2023). Semi-structured interviews were used to collect empirical data, which were analyzed through a qualitative data analysis process presented by Miles et al. (2019) and Saldaña (2021).

The study was conducted in the context of large manufacturing companies, and the unit of analysis is the implementation and use of smart maintenance technologies. Case study research is suitable for exploration, theory building, elaboration and refinement, and can be used to investigate issues within a single company or across multiple companies (Karlsson, 2023). Case study research has also been described as an empirical method for investigating a phenomenon within its real-world context (Yin, 2018). For the aim of this research, this approach was considered appropriate, as the implementation and use of smart maintenance technologies needed to be studied within the context of manufacturing companies. Based on the empirical findings, this research focuses on the following smart maintenance technologies: IIoT, Big Data and Analytics, Cloud Computing, System Integration, AI and CPS.

The case selection followed a purposeful sampling approach, in which participants were selected for their relevance to the phenomenon under study (Maxwell, 1996). An overview of the participating companies and respondents is presented in Table 3. Eleven manufacturing companies, all engaged in discrete-part production, were included in the study to explore experiences in implementing and using smart maintenance technologies across different organizational settings. These companies were selected for their practical experience with these technologies. Respondents were chosen based on their role, years of experience and involvement in maintenance, digitalization, and production development activities. The interviews focused on technologies or initiatives that were currently in use or being implemented within the companies to provide in-depth perspectives.

In this research, semi-structured interviews were used to collect empirical data. When conducting semi-structured interviews, guides are used that address the research aim while allowing respondents to elaborate on issues they consider important (Adeoye-Olatunde and Olenik, 2021). This made semi-structured interviews particularly suitable for this research, as the implementation and use of smart maintenance technologies required flexible conversations where respondents could share their practical experiences. For this study, interviews were conducted using an interview guide consisting of 13 questions (see  Appendix). A total of 12 semi-structured interviews were conducted across 11 large manufacturing companies. The empirical data from case companies A–D were collected in 2024, and the remaining interviews were conducted in 2026 to broaden the empirical basis and include additional company perspectives. The interviews focused on technologies or initiatives currently used or being introduced within the companies to provide in-depth perspectives.

Saturation was assessed during data collection through the coding process described in Section 3.4. After each interview was transcribed and coded, the new codes were compared with the existing code structure. The early interviews introduced a considerable number of new codes, whereas the later interviews mainly reinforced codes that were already established. Furthermore, the 13 categories and 3 themes had stabilized before the final interview. On this basis, 12 interviews were considered appropriate for the aim of this study (Wutich et al., 2024).

To analyze the empirical data, the qualitative data analysis process presented by Miles et al. (2019) and Saldaña (2021) was followed. The recordings from the semi-structured interviews were transcribed and coded using descriptive coding, assigning labels to data that summarize in a single word or short phrase. The coding process was carried out through first-, second- and third-cycle coding.

In the first coding cycle, which summarized data segments, the respondents' answers were coded in NVivo, a computer-based qualitative data management program. In total, 335 codes were identified. These codes were later narrowed to unique 77 codes based on the identified similarities and patterns. In the second coding cycle, pattern coding, which groups codes into fewer categories, 13 categories based on patterns in the data were identified. In the third coding cycle, inspired by Hubka and Eder (1988) and Veile et al. (2020), the categories were organized into three themes: (1) organizational, (2) human and (3) technical, based on patterns identified in the data.

Finally, the codes, categories and themes were visualized in two figures. The first visualizes the empirical data structure for how smart maintenance technologies were implemented at the case companies (Figure 1), and the other visualizes the empirical data structure for the enablers, challenges and benefits associated with such technologies (Figure 2). Some codes appear in both figures, as they describe both the implementation of such technologies at the case companies and the enablers, challenges or benefits associated with them.

The quality of this study is assessed using the concept of trustworthiness, comprising the dimensions of credibility, transferability, dependability and confirmability (Lincoln and Guba, 1985). Credibility was supported by selecting respondents based on their role, years of experience and involvement in maintenance, digitalization and production development activities, following a purposeful sampling approach (Maxwell, 1996) to ensure that those best informed about the phenomenon were included (Karlsson, 2023). In case company A, two respondents were interviewed, while in case companies B–K, the same individuals were responsible for both traditional and smart maintenance. The empirical findings were systematically compared with the theoretical background, and peer debriefing was conducted iteratively throughout the analysis with co-authors, department members and two additional researchers to review and refine interpretations and reduce individual bias (Corley and Gioia, 2004). Transferability was addressed through a detailed description of the research context provided in Table 3, including company size, maintenance department size, type, respondent roles and years of experience. The 11 case companies vary in size, ranging from 150 to 3,500 site employees, and represent different organizational settings within the manufacturing industry. Dependability was supported by the structured three-cycle coding procedure described in Section 3.4 (Miles et al., 2019; Saldaña, 2021), conducted in NVivo, progressing from 335 initial codes to 77 unique codes, then to 13 categories and finally to three overarching themes, as visualized in Figures 1 and 2. Confirmability was addressed through intercoder agreement to handle potential differences in coding the same unit of text (Campbell et al., 2013). The initial coding was performed by one primary coder, and the coding structure was subsequently reviewed with co-authors, department members and two additional researchers to ensure clarity and replicability. In addition, direct respondent quotes further support the transparency and traceability of the analytical process (Rockmann and Vough, 2024).

The empirical findings present that the case companies in this research had implemented various technologies and were engaged in ongoing activities.

For instance, at case company A, the respondents explained: “… a smart device is implemented, which tracks fitness data of industrial machines, powered by experience-driven AI.” The device is installed on two multi-operational machines. Condition monitoring was not conducted continuously. Instead, a test program was run once a week to collect maintenance data during linear and circular motions. Then, the sensor supplier analyzes the data to assess the machines' health using various trend graphs. Related to what was explained by the respondent at case company A, the respondent at case company J noted: “We monitor the oven's status in real time, such as the temperature. We also use temperature thresholds in the oven. In addition, we have approved an AI service that is currently under development.”

A more internally oriented AI approach was described by the respondent at case company D: “… we have also begun to explore generative AI and machine learning for data-driven decision-making.” In this case, machine connectivity, machine learning models and generative AI solutions were integrated with the CMMS to automatically generate maintenance work orders based on anomaly-detection models. Similarly, the respondent at case company I stated that connectivity and maintenance data-driven decision-making were used. Furthermore, the respondent reported: “Maintenance data collection from eight critical machines, such as air compressors, to follow up the consumption of compressed air.”

Connectivity was also mentioned by the respondent at case company C, where OPC UA was used as the standard communication protocol for data collection from welding robots to read the status of all available variables, such as the power sources used in the robots. The respondent also described: “The technical specification for machine acquisition must describe how the equipment should be connected via OPC UA and which signals will be available.” To enable this, they purchased all the necessary options to connect the robots to the network and gain access to available sensor data. In addition to connectivity and data collection, at case company B, the respondent noted that the technology they implemented, which was immature when first used, involves advanced vibration measurements of machine spindles. While vibration measurement itself is not new, the respondent explained that the specific conditions required for machine tools operating at high rotational speeds had not previously existed at the case company.

The respondents mentioned different types of maintenance management system solutions at case companies E and F. At case company E, the respondent reported implementing an Enterprise Resource Planning (ERP) solution, “… a module for Asset Management in the ERP system”. The case company had implemented an ERP system that included an Asset Management module and was also running a project to introduce a real-time Overall Equipment Effectiveness (OEE) follow-up system. A similar approach was mentioned by the respondent at Company F: “… a module for maintenance as a part of our ERP system.” In addition to the ERP module, the case company F had begun implementing condition-monitoring tools. The vision was to install sensors to monitor vibrations, enabling maintenance based on actual production conditions rather than on time-based intervals recommended by the machine builder. It was also planned to use temperature sensors to provide further insight into equipment performance. To support this transition, the company had participated in training programs to build the necessary knowledge and competence for a more data-driven maintenance.

In addition to the maintenance management system, the respondents at case companies G, H and K focused on CMMS implementation. At case company G, the respondent reported that they have developed their own CMMS and have begun discussions about moving toward maintenance data-driven decision-making. At case companies H and K, the respondent noted that they have implemented a CMMS ready for System Integration, IIoT for maintenance data collection and AI for analysis.

These examples of initiatives and practical use cases show that the case companies have implemented various technical solutions for smart maintenance, including condition-monitoring technologies, connectivity solutions, ERP modules, CMMS systems and AI-driven initiatives.

This section presents the empirical data structure for the codes and categories, organized into three themes: organizational, human and technical, developed from participants' descriptions. The empirical data structure is visualized in two figures. The first visualizes the empirical data structure for how smart maintenance technologies were implemented at the case companies (Figure 1), and the other visualizes the empirical data structure for the challenges, enablers and benefits of smart maintenance technologies (Figure 2).

In the sections below, each category will be elaborated on according to its codes.

Pre-study describes how respondents approached the initial steps before implementing smart maintenance technologies.

At case company B, the respondent explained: “It starts with an idea, and then the implementation project is defined and included in the technology roadmap.” The same respondent added: “There are priorities on which ones are more valuable.” This shows that the initial steps focused on clarifying ideas and deciding which activities to move forward with.

The respondents highlighted the importance of pilot testing before implementation. For instance, the respondent at case company A explained: “The technology was interesting, so a pilot project was initiated.”

The evaluation of financial aspects was also highlighted. The respondent at case company B stated: “Validation of the project … Is it profitable? Can we proceed with the implementation phase?” Similarly, the respondent at case company E described using an “… investment plan …” to demonstrate the expected benefits of the technologies. These examples show that profitability and clear justification are important before full implementation.

Furthermore, this category includes defining requirements and measurement strategy and preparing business cases. Respondents also described assessing the existing maintenance maturity level, including current maintenance plans. Related to this, the respondent at case company D explained: “Not all departments share the same perspective …” This shows that differences in maturity and understanding within the organization can affect the company's readiness for new technology.

The findings indicate that this category plays an important role in reducing uncertainty, prioritizing the right initiatives and ensuring that the organization is prepared before implementation begins.

Machine selection focuses on how respondents explained their initial selection of a machine when implementing smart maintenance technologies. The focus was on identifying a suitable starting point by selecting a machine that would provide relevant data and clear results.

Several considerations influenced the selection process, including the number of prior maintenance actions, historical maintenance data from the CMMS and the machine's criticality and classification. At case company A, the respondent explained: “Historical data was retrieved from the CMMS with all types of faults …” This shows that past failure data was used to support a more informed decision. Regarding the machine's criticality and classification, the respondent at case company B noted: “The machines are classified as A, B, and C, where A is the most critical.” In addition, respondents highlighted the importance of data availability for the selected machine, such as access to OEE and process data.

This category indicates that machine selection was not random, but based on data, risk level and the potential to generate valuable insights.

Cross-functional team covers collaboration between departments during the implementation of smart maintenance technologies. Respondents highlighted the involvement of production, production technology, maintenance and IT departments, indicating that smart maintenance is not limited to any single function but requires collaboration across organizations.

At case company I, the respondent explained: “… work cross-functionally across IT and OT [Operational Technology], making joint decisions.” This highlights the importance of shared responsibility and joint decision-making. A similar approach was mentioned by the respondent at case company C: “IT is often involved, since it is largely about IT,” indicating a strong dependence on the IT department in technical activities. Furthermore, the respondent at case company H explained: “… several people are involved, mainly the production department. Those who use the machine daily are involved, as well as production engineers who have knowledge of, or more in-depth expertise in the machines.” This shows the importance of involving the production department, as they are the company's internal customer.

In addition to collaboration, respondents highlighted the importance of assigning clear roles and responsibilities, including defining who is responsible for each task and ensuring effective communication among the involved departments.

The findings indicate that cross-functional cooperation and role clarity are important for successful implementation, as smart maintenance technologies are characterized as cross-border technologies.

The goal and expectation of the technology were a recurring topic. Many respondents across the case companies described similar primary goals. A common aim among them was to enable early fault detection and improve health monitoring.

At case company A, the respondents stated that obtaining early information about machine health was a primary goal and explained: “… to obtain information about the machine's status before there are issues with the machine.” The goal was to receive a warning that something needed to be addressed before operators or maintenance personnel detected the problem themselves. A similar goal was described by the respondent at case company B: “… to indicate when the machine spindle is worn mechanically on a degradation curve, to signal that we have an impending failure or that we will encounter quality problems in this machine spindle if we keep it in operation.”

Likewise, the respondent at case company I stated: “Our primary goal is to ensure high technical availability and gradually transition to more Condition-Based Maintenance. We focus on early fault detection to reduce reactive maintenance. We monitor KPIs [Key Performance Indicators], such as MTTR [Mean Time To Repair], and MTBF. We work with structured reporting processes and handle approximately 4,500 preventive maintenance work orders annually, with a strong emphasis on preventive maintenance.” Similarly, the respondent at case company D mentioned threshold monitoring, failure prediction, expanding the toolbox with Industry 4.0 technologies, acquiring additional tools for value-driven maintenance, and working more with auto-generated maintenance work orders in the CMMS. At case company J, the respondent stated a similar goal: “The goal is to predict the required maintenance. We want to monitor and optimize operations to get the most out of the oven.”

Together, these statements show that early detection, a higher availability and a shift toward more preventive and CBM were common expectations.

Another recurring goal was improved understanding of processes and data-driven decision-making. The respondent at case company C stated that it was necessary to gain a better understanding of what is happening in the process and how different considerations interact. The respondent highlighted this by stating: “If we look back historically, there's been a lot of guesswork rather than facts,” underlining the need for reliable information to improve maintenance, operational reliability and data-driven decision-making. At case company E, the respondent similarly explained: “To measure, control, and improve operations, it is necessary to first understand what losses occur and what causes them.” The respondent further noted: “This requires the collection of measurement data, as the processes are complex and influenced by many factors.” These examples show that companies aim to replace assumptions with structured data and measurable insights.

Structured reporting and traceability were also highlighted. The respondent at case company F also stated that the goal is to increase production productivity. When management asks for explanations of performance variations, it is often not possible to provide clear answers due to a lack of structured data. To address this, “… the company aims to introduce predefined codes and systematic data collection, enabling better traceability and analysis.” The respondent at the case company I also mentioned predefined codes to present concrete information to management and to identify the essential causes of inefficiencies.

A further goal concerned efficiency, maintenance management control, and transparency. The primary goals mentioned by the respondent at case company G were to gain an overview of the processes for Preventive Maintenance and Corrective Maintenance. At case company H, the respondent stated that the aim was to increase efficiency and improve communication between maintenance and production. The maintenance management systems should facilitate task assignment, extract data and generate statistics on working hours and costs, support KPI tracking and improve spare parts management. Finally, the respondent at case company K highlighted that the primary goal is to have a maintenance management system ready for the future of maintenance.

These findings highlight the importance of early detection of mechanical wear, monitoring machine condition using degradation curves and data collection to signal failures and prevent quality issues. The main goals across companies were to get earlier signals, improve efficiency, support data-driven maintenance and seek greater control, predictability and reliability.

Decisions regarding which technologies to implement were influenced by financial constraints, compatibility requirements, previous experience, cross-functional collaboration and strategic priorities. Rather than following a standardized procedure, the case companies based their decisions on a combination of technical, financial and organizational considerations.

Financial and strategic planning played an important role in several cases. Regarding case company B, the respondent stated that the decision-making process was strongly linked to funding and strategic planning, “There are different approaches, and a lot depends on funding as well. What financing opportunities one has … We have something we call a technology roadmap, which includes developing new technology. We describe what we want to do and we apply for funding.” This demonstrates how access to financing and a technology roadmap influenced the choice of technologies. A similar financial approach was described by the respondent at case company E, where they use an “… investment plan …” to demonstrate the benefits of the technologies to be implemented and then test them in various pilot projects.

Technical requirements and compatibility were also important decision factors. At case company C, implementing new technologies requires using OPC UA as the standard communication protocol for data collection and seamless System Integration and ensuring Cyber Security. This shows that compatibility requirements shape decision-making. At case company D, approaches to connectivity, compatibility and data collection were also in focus, and the respondent reported that they decided to implement technologies to improve spare parts management, improve Corrective Maintenance through auto-generated maintenance work orders, support data-driven decision-making and use generative AI.

Previous experience and comparison of alternative solutions also guided decisions. In the case company A, the decision was based on comparing alternative technical approaches. The respondents explained that they had previously explored CBM and tested vibration sensors. However, they later found a smart device that appeared to offer a more comprehensive understanding of the entire machine unit, rather than focusing solely on a specific component of the machine, “We previously looked at CBM, and we have some sensors in other systems where we have tested vibration sensors. But then we came across this smart device, and it seemed to capture more information about the entire unit, rather than focusing on a specific component of it.”

Internal needs also influenced technology selections. At case company G, the respondent stated: “We decided to develop our own CMMS to gain a clearer overview of maintenance processes.” This highlights how internal functional needs shaped the decision. In contrast, case companies H and K selected an existing CMMS because it had already been validated at other sites and was familiar to the IT department, indicating that prior experience and familiarity with the system influenced the selection. Likewise, at case company J, the respondent explained: “We chose data collection solutions from the machine suppliers.”

Cross-functional collaboration was another factor, such as collaborating with universities on pilot projects to verify hypotheses, as mentioned by the respondent at case company B. Concerning case company I, the respondent stated that they “… work cross-functionally across IT and OT, making joint decisions.” Each organization has its own requirements and KPIs based on its technical specification, and they collaborate from an early stage.

Finally, case company F reported a lack of formal process, and the respondent explained: “A clear process for decision-making is not available today.”

These show that technology decisions were shaped by a combination of technical requirements, financial conditions, strategic goals and operational and organizational experience, rather than by a single procedure.

Installation focuses on the practical activities involved in implementing smart maintenance technologies in production. This category involves coordination across departments and requires both technical preparation and organizational support.

The Installation category describes how respondents discussed the validation and practical implementation of the technologies through cross-functional collaborations. For example, the respondent at case company B stated: “… sensors are often wired, so they place work orders with line repairers to install the sensors and cables. Then there are the boxes to set up. IP [Internet Protocol] addresses are needed, so the IT department is involved.” The findings also show that collaboration with the suppliers, such as sensors and machine suppliers, can provide support during installation. Related to this, the respondent at case company A explained: “… preparations for supply voltage, calibration programs, and the mechanical setup were made so that the sensor manufacturer could install the sensor.” In addition, this category includes installing hardware and software, connecting machines, integrating with CMMS or ERP systems, and determining the optimal placement of sensors.

The findings indicate that installation is not only a technical activity but also a coordinated organizational effort. Successful implementation depends on collaboration between maintenance, production, IT and external suppliers, as well as careful preparation of both technical infrastructure and System Integration.

Evaluation covers how respondents highlighted follow-up activities after implementation. It includes evaluations of technical performance, perceived profitability, learning from failure and effects on daily maintenance work, such as employees' perceptions. Regarding employees' perceptions, the respondent at case company B explained that, due to the absence of vibration-related quality problems, “It's much calmer in the factory” and “There's no stress in the factory.” This category also includes lessons learned from the implementation and data-driven follow-up, including monitoring trends and KPIs. Regarding data-driven follow-up, the respondent at case company F stated: “… the company aims to introduce predefined codes and systematic data collection, enabling better traceability and analysis.”

The findings indicate that evaluation involved both technical assessment and organizational reflection, combining the performance measurement system with employee experience.

Figure 2 presents the empirical data structure related to challenges, enablers and benefits associated with the smart maintenance technologies. In the sections below, each category will be elaborated on according to its codes.

Organizational and human challenges highlight challenges for successful implementation. Financial issues were often highlighted, particularly the difficulty of demonstrating clear technical and economic benefits. As the respondent from case company K noted: “ … it is important to clearly demonstrate the benefits and the savings that the initiative can generate.” This indicates that clear economic and technical value is vital for gaining managerial support and ensuring long-term commitment. In addition to demonstrating the technology's value, respondents at case company A noted that an implemented monitoring system had not yet generated any alerts, making it difficult to demonstrate profitability to the management team. This was also highlighted by respondents at case company B, who reported challenges in demonstrating financial value: “… managers wonder why we should keep the vibration sensors that cost a lot of money.” Although no emergency issues with machine spindles were reported, the technology was seen as expensive and there was uncertainty about its long-term benefits. Similar statements about understanding the technology's clear benefit were raised by the respondent at case company I. The respondent explained: “There is no resistance, but many may not fully understand the value of the technology.” In addition, at case company C, the respondent stated that showing concrete results from the technology was challenging. These results indicate that unclear benefits and a lack of reliable data on costs and financial benefits may make it harder for people to accept its use.

User resistance was a common pattern across several case companies. For instance, in the case company A, skepticism about the value of new technologies was highlighted. Respondents explained that mechanics questioned the purpose of the new technologies by explaining: “This won't yield anything. We already know what faults will occur in the machines.” The respondents clarified that the goal is not to know what faults will occur but to detect them early. Additionally, mechanics felt that the new technology wouldn't add value because they already had sufficient knowledge of the machinery and its potential failures, which increased their reluctance to adopt it. Resistance was especially noticeable when the benefits weren't immediately shown, increasing the skepticism among mechanics.

Resistance to new routines was especially noticeable in system usage and reporting practices. At case company H, the respondent explained: “Many employees prefer to report faults via email, but the organization has been strict that all reporting must be done through the CMMS rather than by email.” This shows how traditional work routines might conflict with new technologies. Furthermore, challenges at case company G included data quality issues and delayed reporting, demonstrating that technical systems alone are not enough without regular user compliance.

Moreover, organizational capabilities were also identified as major challenges. At case company E, the respondent reported difficulties defining roles within the maintenance module, planning maintenance activities, conducting root cause analysis to prevent recurring failures and managing a broader shift from subjective to objective decision-making. The respondent at case company F highlighted human aspects and adherence to routines as key challenges, while maintaining competence over time was identified as a challenge at case companies B and C. Furthermore, the respondent at case company D highlighted variations in organizational maturity.

Organizational and human enablers highlight several enablers that may support successful implementation, including showing success stories and results to motivate employees, performing risk analysis to avoid unanticipated risks, cross-functional collaboration, such as building trust between the IT and maintenance departments, as much as it depends on a well-designed IT architecture, highlighting a shift toward data-driven maintenance and setting clear maintenance goals. The respondent at case company B described communicating positive outcomes as an important enabler: “Maybe we should have an information campaign and perhaps show more success stories to ensure a more solid acceptance.” This shows that demonstrations of positive outcomes may reduce resistance and strengthen organizational commitment to the initiative.

Organizational and human benefits describe the benefits of the implemented smart maintenance technologies. These benefits include a focus on strategic issues, such as proper purchasing from the start, fewer large-scale quality investigations, fewer daily quality issues, an improved work environment, etc.

Regarding correct purchasing from the start, the respondent at case company B explained: “We have improved the process by installing these sensors right from when we purchase the machine. We select the sensors and send them to the machine builders, and we agree on where they should be placed on the spindle.” This shows that the company now includes smart maintenance technologies in the purchasing process, improving long-term planning. Furthermore, the respondent noted: “What I find exciting is that the technology has become so recognized among line engineers that they talk about it and want it. The engineers who purchase the machines take responsibility for this and ensure that it is included in the new machines.” This shows that the technology has become more accepted and integrated into long-term routines.

Positive user attitudes and increased interest in technology were reported at several companies. At case company E, the respondent stated: “There is great appreciation for the ability to follow up activities in the maintenance module in ERP. Without this module, the situation would have been highly unstructured and difficult to manage.” Similarly, respondents at case company G reported no challenges with their in-house-developed CMMS, which can also be seen as a positive attitude. At case company J, the respondent highlighted a positive outcome among operators: “We only see opportunities. Operators are asking for smart systems to monitor the health of the ovens.”

In addition to financial and strategic benefits, respondents also reported improvements in the work environment due to the absence of vibration-related quality problems and large quality investigations. The respondent at case company B stated: “Production is more reliable, and that reduces stress.” These statements indicate that the technology has led to fewer urgent spindle replacements, more stable production, less stress and better control.

Finally, the respondent at case company J explained: “It gives us a faster way to identify upcoming maintenance needs well in advance and plan our resources accordingly. We also want to involve production, since they are affected when the ovens are down, so they have the opportunity to plan production activities more effectively.” This shows that the technology supports earlier planning and better cooperation between maintenance and production.

The findings indicate that managerial considerations strongly influence the implementation of smart maintenance technologies. Unclear benefits, resistance, and competence gaps may hinder the implementation, and clear communication, structured goals and collaboration can support acceptance and value.

Technical challenges describe several challenges to consider. These include challenges related to dependence on departments other than maintenance, technical challenges during hardware installation, NC programming and complicated PLC communication, challenges with large datasets, collecting data from legacy machines by using external sensors, which complicates the System Integration, etc.

Considering the dependence on the IT department, the respondent at case company C explained: “IT is often involved, since it is largely about IT,” indicating a strong dependence on the IT department in technical activities. In addition, the respondent at case company B explained: “… sensors are often wired, so they place work orders with line repairers to install the sensors and cables. Then there are the boxes to set up. IP addresses are needed, so the IT department is involved,” indicating technical challenge and dependence on other department than maintenance during hardware installation.

Regarding NC programming and PLC communication, the respondent at case company B reported: “The biggest challenge is the NC programming …”

Regarding data management and large datasets to manage, for case companies H and K, the primary challenge was unstructured data. The respondent at case company H explained: “Much of the information is entered as free text, which makes it difficult to evaluate failures and identify root causes,” and the respondent at case company K explained: “Spare parts are stored in different systems and have unstructured naming. The legacy data is not compatible with the ERP system. It is complicated to transfer the data and achieve proper integration.”

Cyber Security and data protection were highlighted as critical concerns at case companies C, E, I and J. This indicates that technical integration is not only a system issue but also a security concern. Additionally, the respondent at case company E highlighted the need to manage data in legacy systems.

In the case company A, additional challenges related to expanding understanding of the technologies were noted, including deepening their understanding of the technology and aligning data collection with reality. The respondents explained: “… to expand understanding of the technology itself and to align data collection with reality.” The respondent explained that it will take time to learn how to interpret the graphs and trends and to assess how well they correspond with reality, which will necessitate adjustments. Moreover, the respondents stated that the data is stored in a cloud owned by the sensor supplier, and they don't have access to detailed data.

In contrast, the respondent at case company G reported few challenges, as they had developed their own CMMS customized to their maintenance processes.

Technical enablers highlight several enablers that support the technology's implementation. One example is the use of 4G routers, which can facilitate the installation of sensors that operate independently of the company's regular network. This makes the installation process easier and reduces dependence on internal IT systems. Other technical enablers include access to detailed sensor data, proper sensor placement on machines and ensuring that NC programming is performed by the machine builder, who has greater expertise in this area. Regarding NC programming, the respondent at case company A stated, “… one who knows NC programming the best is the machine builder.” This shows that involving the machine builder can reduce technical limitations and improve the quality of the implementation. Additionally, this category covers enablers related to exploring CBM options before selecting smart maintenance technologies, ensuring Cyber Security when connecting machines, verifying hypotheses through pilot projects by collaborating with universities and the suppliers of the machines and sensors.

Technical benefits describe the benefits of smart maintenance technologies. These benefits primarily include increased production reliability, early failure detection, improved process understanding and more data-driven decision-making.

Increased production reliability and the ability to prevent unplanned downtime were reported across several case companies as technical benefits. At case company I, the respondent explained: “Where sensors are installed, we have been able to predict failures. The strength of the technology lies in detecting early signs of potential issues, enabling us to prevent unplanned downtime.” A similarly high level of production reliability was reported at case company E. At case company A, monitoring, observation and tracking machine status were described as important benefits. Regarding case company C, the respondent highlighted the possibility of tracking trends before a production stop occurs, such as auto-generated maintenance work orders, as mentioned by the respondent at case company D. Together, these findings show that the technology improves stability, enables early detection and reduces the risk of unexpected production disturbances.

At case companies F and G, respondents highlighted the importance of having a maintenance plan for each machine. Additionally, at case company H, the respondent stated: “Historical data can be reviewed to support analysis and decision-making.” This shows that maintenance and performance data support better planning, continuous improvement and deeper insight into equipment condition and efficiency. Related to this, the respondent at case company K mentioned better control over the start and completion of the maintenance work orders. Together, these findings indicate that structured maintenance planning and access to historical data contribute to better control, improved follow-up and more informed decision-making.

Improved process understanding and more data-driven decision-making were also highlighted. At case company C, the respondent described gaining a clearer understanding of the process and making correct purchasing decisions from the beginning. Similarly, the respondent at case company D highlighted the use of various KPIs, which led to more cost-effective sub-processes and improved process knowledge. These examples show that the technology supports better operational decisions.

Finally, positive effects were also visible at the operator level. At case company C, operators appreciated the visual insight into process performance, and demand for the technology was growing.

The findings demonstrate that technical considerations are essential for successful implementation. While integration, data quality and security issues can pose challenges, proper system setup and structured data management enhance reliability and support more data-driven decision-making.

The empirical findings are discussed below in relation to three identified themes: organizational, human and technical. For each theme, the observations across the 11 case companies are interpreted in relation to the theoretical background and discussed for the understanding of smart maintenance technologies in industrial practice.

Organizational: At the organizational level, the empirical findings indicate that the implementation of smart maintenance technologies is shaped more strongly by organizational conditions than by the selection of technology itself. Several challenges influenced how technologies were introduced, such as large investments and unclear returns (Zhang et al., 2017), often leading to gradual implementation through pilot projects (Bengtsson, 2008; Mittal et al., 2018). Cross-functional collaboration between maintenance, production and IT departments was consistently required. This confirms the internal integration dimension identified by Bokrantz et al. (2020a) and reflects the need to align departments and secure IT support, as described by Giliyana et al. (2025). Defining clear roles and responsibilities within such collaboration aligns with Salonen (2009), who identifies the need to clarify strategic goals across departments. At case company F, the absence of a structured decision-making process appeared to negatively affect implementation. This is consistent with Flores et al. (2018), who indicate the importance of a clear strategy when implementing new technologies. Overall, the empirical findings support prior research, showing that organizational readiness, including maturity level, goal clarity and financial justification, is a crucial consideration in successful implementation (Lundgren et al., 2022; Bajic et al., 2021; Zhang et al., 2017; Salonen, 2009). Collectively, these findings indicate that organizational readiness, including clear strategy, defined roles and cross-functional collaboration, constitutes a fundamental precondition for the successful implementation of smart maintenance technologies, regardless of the specific technology introduced.

Human: At the human level, the empirical findings indicate that the acceptance and effective use of smart maintenance technologies depend more strongly on human conditions than on the capability of the technologies themselves. Although early failure detection and condition monitoring were common goals, respondents also mentioned better coordination, clearer information and more structured processes, aligning with previous research indicating that the focus should be not only technical but also organizational and human (Lundgren et al., 2023). User resistance was observed at several companies, but was mainly driven by unclear benefits rather than opposition to the technology itself, consistent with acceptance of the new technologies and resistance to knowledge upgrades reported by Nasirinejad et al. (2024) and Bajic et al. (2021). When results were made visible, acceptance increased, supporting education and training, and demonstrating the benefits of the technologies for both the company and employees (Bajic et al., 2021; Bengtsson, 2008). Furthermore, competence gaps were evident, which aligns with the human capital dimension identified by Bokrantz et al. (2020a) and the difficulty in finding Industry 4.0-skilled operators noted by Silvestri et al. (2020). The findings also confirm that implementation affects employees beyond their technical roles, as improvements in production reliability contributed to a calmer and less stressful work environment, supporting the argument that implementation should maintain a human-centered focus and be evaluated from an employee perspective (Bengtsson, 2008; Lundgren et al., 2023). Collectively, these findings indicate that user acceptance depends on whether employees can see the aim and benefits of the technology, rather than on the technology itself, positioning the human theme as a central condition for successful implementation.

Technical: At the technical level, the empirical findings indicate that smart maintenance is, in practice, used to improve rather than replace traditional maintenance, and that the most fundamental barriers concern the conditions for using the technology rather than the technology itself. A range of technologies was implemented at the case companies, including IIoT, AI-driven analysis, CMMS and ERP-integrated maintenance modules. Technology selection was driven by traditional maintenance routine, financial considerations and integration requirements rather than technological trends. A consistent pattern was the use of new technologies to enhance traditional maintenance practices rather than replace them, aligning with Bengtsson and Lundström (2018) and Al-Najjar et al. (2018), who emphasize that measurement and analysis alone are insufficient without the active maintenance actions that ultimately reduce breakdowns. Furthermore, the empirical findings present that the case companies primarily operated at the detection and condition-monitoring levels, aligning with Tinga (2025), with some beginning to move toward prognosis and prescriptive maintenance (Giacotto et al., 2025). Challenges limiting this progression included poor labelling of free-text data, the absence of threshold values for maintenance actions and limited access to detailed sensor data, all of which correspond to challenges identified by Tinga (2025).

Data quality emerged as the most fundamental technical challenge. For instance, unstructured data, legacy systems and poor data structure limited the value of even well-integrated technical systems, confirming the findings of Tinga (2025) and Bajic et al. (2021) by suggesting that data quality may be a greater barrier than the technology itself. This also aligns with Lee et al. (2018), who emphasize the importance of collecting accurate, high-quality data by understanding the system and its parameters. Cyber Security and System Integration were also recurring concerns, consistent with previous research (Nasirinejad et al., 2024; Lundgren et al., 2021), in which adhering to a single communication protocol to ensure seamless System Integration and Cyber Security aligns with Giliyana et al. (2025). An important technical enabler, as indicated by the empirical findings, was the involvement of the machine builder in NC programming, aligning with the external integration dimension of Bokrantz et al. (2020a). The reported technical benefits, primarily early failure detection, trend monitoring and improved process understanding, were mainly related to better visibility and control rather than full automation, reflecting that the case companies are still in early stages of a broader technological transition. Collectively, these findings indicate that the value of smart maintenance technologies depends less on how advanced the technology is and more on the quality of the underlying data and the integration of the technology into traditional maintenance practices.

Across the three themes, a common pattern emerges: the success of smart maintenance technologies depends more on the conditions surrounding the technology than on the technology itself. Organizational readiness, employee acceptance and data quality together determine whether a technology can deliver value in practice, regardless of how advanced it is.

Based on the empirical findings, this study provides recommendations for manufacturing companies seeking to implement and use smart maintenance technologies. As shown in Figure 3, the recommendations are organized into three interconnected themes: organizational, human and technical. These themes do not operate independently, and challenges in one area consistently affect the others.

Organizational: Involve maintenance, production and IT from the start, as smart maintenance technologies are characterized as cross-border technologies and cross traditional departmental boundaries. Assign clear roles and establish joint decision-making between these functions early to reduce delays and conflicts. Secure management support by demonstrating concrete results through pilot projects, tracking relevant KPIs and presenting outcomes in business terms. Maintain organizational commitment over time by continuously communicating progress and value. When selecting a machine to start with, base the decision on historical failure data and criticality classification, as this reduces implementation risk and increases the likelihood of generating meaningful results early.

Human: Make the benefits of new technologies visible at an early stage, as user resistance is most caused by unclear value rather than opposition to technology itself. When employees can see concrete results, attitudes shift and engagement increases. Use success stories and demonstrations to build acceptance across the organization. Assign responsibilities clearly, as accountability motivates employees to engage with and take ownership of new systems. Involve operators and production personnel early, as their practical knowledge contributes to more informed implementation decisions. Finally, invest not only in initial training but also in continuous learning, as maintaining competence over time is essential for sustained implementation success.

Technical: Prioritize data quality before investing in advanced analytics or AI-driven tools. Unstructured data, free-text entries and incompatible legacy systems limit the value of even well-integrated technical solutions and should be addressed early. Establish structured data collection practices as a foundation for more data-driven maintenance. Additionally, ensure Cyber Security in connectivity, adhere to standardized communication protocols for seamless System Integration, create a roadmap for new technologies and collaborate with suppliers and universities on pilot projects.

This study has explored how smart maintenance technologies are implemented and used in 11 manufacturing companies. This section presents the academic and industrial contributions, the limitations and suggestions for future research.

This study contributes to the academic literature in four ways. First, it adds empirical insights from 11 manufacturing companies to a research field where there is a gap in studies directly linked to industrial practice (Fraser et al., 2015; Silvestri et al., 2020). Second, the findings show that the main barriers to implementation are not the technologies themselves, but the organizational and human considerations. This adds to research mainly focused on the technical (Lundgren et al., 2023), and supports the dimensions human capital resources, internal and external integrations by Bokrantz et al. (2020a), by indicating that these considerations act as preconditions for implementation. Third, the study adds to research on user acceptance (Nasirinejad et al., 2024; Bajic et al., 2021) by presenting that resistance is mainly driven by unclear benefits rather than the technology itself. This means that the focus should be on making the benefits of the technology visible to employees. Fourth, the new generation of maintenance is often described as a progression toward prognosis and prescriptive maintenance (Tinga, 2025; Giacotto et al., 2025). However, the findings show that the case companies mainly operated at the detection and condition-monitoring levels and used smart maintenance technologies to improve traditional maintenance rather than replace it. This adds to the understanding of the gap between what smart maintenance technologies promise and how they are used in industrial settings.

This study is intended to support manufacturing companies in implementing and using smart maintenance technologies. The practical contribution lies in two areas. First, while earlier studies often present recommendations within a single area, such as technology selection (Silvestri et al., 2020), human capital (Bokrantz et al., 2020a) or strategy (Flores et al., 2018), this study brings organizational, human, and technical recommendations together and presents how they depend on each other. Practitioners are therefore guided to address all three themes at the same time, rather than focusing on one at a time. Second, the recommendations are specific and based on insights from the case companies, rather than general principles. Examples include basing machine selection on historical failure data and criticality classification, prioritizing data quality before investing in advanced analytics or AI-driven tools, and making the benefits of new technologies visible to employees from an early stage. These practical actions provide more concrete guidance than the broader recommendations often found in the existing literature.

There are some limitations to consider when interpreting the findings. The case companies are all Swedish manufacturing companies engaged in discrete-part production. The findings therefore reflect this specific context, and future research could explore how the identified themes apply in continuous-process industries, such as paper and pulp or chemicals, and in manufacturing contexts outside Sweden. The study also focuses on large manufacturing companies. Small and medium-sized companies may have different resources and organizational structures, and future research could examine how the recommendations apply, or may need to be adapted, in such settings. Finally, as the implementation of smart maintenance technologies develops over time and involves several roles within the organization, longitudinal studies and studies that include a wider set of respondents per company could provide a broader view of how implementation unfolds in practice.

Three research directions are suggested. First, longitudinal studies are needed to examine how the implementation of smart maintenance technologies develops over time and how the identified themes relate to implementation outcomes. Such studies could follow a small number of manufacturing companies using a combination of repeated interviews, document analysis and observations, to capture how challenges and enablers evolve as implementation matures. Second, further research is needed on the benefits of smart maintenance technologies and how these benefits should be defined, measured and compared. Quantitative studies could be conducted in collaboration with industry to develop and test indicators for outcomes such as reduced downtime, improved process understanding and changes in employee work environment. Third, the financial implications and return on investment of smart maintenance technologies need further attention. Future research could combine case studies of completed implementations with quantitative analysis of investment decisions, costs and benefits over time, to provide more concrete guidance to manufacturing companies on how to evaluate smart maintenance investments.

1. What smart maintenance technology do you consider you have implemented?

2. Did you have any primary goal with the implementation of the smart maintenance technology?

3. How did you make the decision regarding what smart maintenance technology to implement?

4. Did you define the requirements for the smart maintenance technology to be implemented?

5. Did you follow any implementation process? If yes, can you describe the process? What steps does it consist of?

6. Did you involve different departments in the implementation process? If yes, how and why?

7. Did you face any challenges during the implementation process?

8. How has the implementation of smart maintenance technology affected the company's maintenance processes and company's goals?

9. What improvements have you observed since the implementation, if any?

10. Have there been any unexpected effects or results that you have encountered after the implementation?

11. Any challenges in using the smart maintenance technology?

12. How has user acceptance and adaptation to the new technology been?

13. Anything else you think I should know?