Journal of Ai ML DL

1. Introduction

Supply chains are no longer viewed simply as linear systems that move materials, products, and information from one point to another. In recent years, they have become complex, data-driven networks that must respond quickly to fluctuating demand, cost pressures, environmental expectations, and unexpected disruptions. This shift has been especially visible in the context of Industry 4.0, where digital technologies are gradually reshaping how firms monitor operations, coordinate partners, and make decisions across the supply chain. Although large enterprises often receive most attention in this discussion, small and medium-sized enterprises (SMEs) are increasingly facing the same pressures, often with fewer resources and less technological capacity (Moeuf et al., 2018; Bosona & Gebresenbet, 2023).

Among the technologies associated with this transformation, the Internet of Things (IoT) has become particularly important because it allows firms to collect real-time information from physical assets, production systems, warehouses, transportation routes, and environmental conditions. Through connected sensors and devices, IoT can improve visibility across supply chain activities, helping managers detect delays, monitor inventory, track energy use, and respond more quickly to operational problems. In this sense, IoT does not only support automation; it also provides the informational foundation required for more responsive and transparent supply chain management (Frank et al., 2019; Taj et al., 2023). For SMEs, this can be valuable, since even modest improvements in visibility may reduce waste, improve planning, and support better coordination with suppliers and customers.

Artificial intelligence (AI) adds another important layer to this digital transformation. While IoT generates large amounts of operational data, AI can help interpret those data and convert them into practical decisions. Machine learning, predictive analytics, and optimization algorithms can support demand forecasting, inventory planning, supplier evaluation, route optimization, and risk detection (Lu, 2019; Aljohani, 2023). In supply chain settings, this means that firms may move away from reactive decision-making and toward more anticipatory management. For example, AI-supported systems can identify patterns in demand variation, predict possible disruptions, or recommend more efficient resource allocation. However, the value of AI depends heavily on data quality, organizational readiness, and the ability of firms to integrate AI outputs into day-to-day managerial decisions (Hangl et al., 2022).

Lean systems remain equally relevant in this discussion. Although lean management was developed before the current wave of digitalization, its core aim—reducing waste while improving value creation—fits closely with the goals of smart supply chains. Lean practices help organizations simplify workflows, remove non-value-adding activities, reduce excess inventory, and improve process discipline. When combined with IoT and AI, lean systems may become more dynamic because inefficiencies can be detected in real time and improvement actions can be guided by data rather than observation alone (Rahardjo et al., 2023; Vlachos et al., 2023). This integration is especially meaningful for SMEs, where resource limitations often make waste reduction and process efficiency essential for survival.

At the same time, sustainability has become a central concern in supply chain management. Firms are increasingly expected to reduce carbon emissions, improve waste management, support ethical sourcing, and adopt circular economy practices. Technologies such as IoT, AI, blockchain, and advanced data analytics may support these goals by improving traceability, transparency, and resource efficiency (Feng et al., 2020; Charles et al., 2023; Riad et al., 2024). Blockchain, for instance, has been discussed as a tool for strengthening traceability and trust across supply networks, particularly where product origin, supplier responsibility, or transaction transparency matters (Rejeb et al., 2019; Bosona & Gebresenbet, 2023). Yet, despite these potential benefits, SMEs often face barriers such as high implementation costs, limited technical expertise, fragmented infrastructure, and uncertainty about return on investment (Raja Santhi & Muthuswamy, 2022).

Against this background, the present study examines how the adoption of IoT, AI, blockchain, and lean-oriented practices relates to supply chain efficiency, sustainability performance, and financial outcomes among SMEs. Using quantitative data from 300 SMEs, the study investigates whether higher levels of technology adoption are associated with improvements in operational performance, carbon footprint reduction, waste management efficiency, and annual revenue. By focusing on SMEs, the study contributes to a more grounded understanding of smart supply chain transformation in organizations that are important to economic development but often underrepresented in Industry 4.0 research. More practically, the study aims to clarify how digital and lean integration may help SMEs build supply chains that are not only more efficient, but also more adaptive, sustainable, and competitive.

2. Materials and Methods

2.1 Study Design and Reporting Approach

This study adopted a quantitative, cross-sectional analytical design to examine the relationship between Industry 4.0 technology adoption and supply chain performance among small and medium-sized enterprises (SMEs). The methodological approach was developed to assess whether the integration of Internet of Things (IoT), artificial intelligence (AI), blockchain-enabled traceability, and lean-oriented operational practices was associated with measurable improvements in supply chain efficiency, sustainability performance, and financial outcomes. Because the study involved firm-level observational data rather than an intervention, the analysis was structured around association testing, predictive modelling, and robustness checking rather than causal inference.

The study design was informed by prior work showing that digital technologies can improve supply chain visibility, traceability, responsiveness, and operational coordination when they are appropriately embedded into business processes (Frank et al., 2019; Rejeb et al., 2019; Taj et al., 2023). Lean systems were also considered because waste reduction, process discipline, and resource optimization remain central to supply chain improvement, particularly when SMEs operate under financial and infrastructural constraints (Moeuf et al., 2018; Rahardjo et al., 2023; Vlachos et al., 2023). The methods were therefore organized to capture both digital adoption and sustainability-related performance indicators in a way that could be replicated by future researchers.

2.2 Study Setting and Population

The target population consisted of SMEs operating in supply-chain-dependent sectors, including manufacturing, distribution, retail, logistics, agri-food, and technology-enabled service industries. SMEs were considered eligible when they had an identifiable supply chain function, maintained records related to operational or sustainability performance, and had at least partial exposure to digital or lean management practices. Firms were excluded when they lacked supply chain activities, did not provide sufficient data for the selected variables, or submitted incomplete responses that could not be corrected during data cleaning.

A total of 300 SMEs were included in the final dataset. The sample size was considered adequate for multivariable regression and predictive modelling because it provided sufficient observations relative to the number of independent variables included in the analysis. Although the sample was not intended to represent every SME sector equally, it was designed to capture variation in technology adoption, sustainability maturity, and financial performance across different organizational contexts.

2.3 Sampling Strategy and Data Collection

Data were collected for the period 2021–2022 using a structured firm-level survey and available performance records. The survey instrument was designed to capture the degree of adoption of AI, IoT, blockchain, and lean-related practices, while performance records were used to collect information on sustainability and financial indicators. Where documentary evidence was unavailable, respondents were asked to provide the most recent internally reported values.

A purposive sampling strategy was used to identify SMEs with some relevance to supply chain operations and Industry 4.0 adoption. This approach was selected because many SMEs have uneven digital maturity, and random sampling alone may fail to include firms with measurable implementation of IoT, AI, or blockchain systems. Still, to reduce selection bias, firms from multiple sectors were included, and no company was selected solely because it had high technology adoption. The objective was to obtain a dataset with sufficient variation across low, moderate, and high adopters.

The questionnaire contained sections on organizational profile, technology adoption, lean-oriented practices, sustainability indicators, and financial performance. Respondents were preferably senior managers, operations officers, supply chain managers, IT managers, or business owners familiar with the firm’s operational systems. This respondent profile was chosen because accurate reporting on technology use and supply chain performance usually requires knowledge of both managerial decisions and operational processes.

2.4 Variables and Measurement

The main independent variables were AI adoption level, IoT adoption level, and blockchain adoption level. Each adoption variable was measured as a normalized score ranging from 0 to 1, where values closer to 0 indicated minimal or no implementation and values closer to 1 indicated more advanced or integrated adoption. AI adoption included the use of predictive analytics, automated decision-support tools, machine learning applications, demand forecasting systems, and inventory optimization tools. IoT adoption included real-time monitoring, sensor-based tracking, connected warehouse systems, energy monitoring, and logistics visibility tools. Blockchain adoption captured traceability systems, distributed record-keeping, transaction transparency, and supplier verification mechanisms, consistent with prior discussions of blockchain-supported supply chain transparency (Feng et al., 2020; Charles et al., 2023).

Lean-oriented practices were assessed through indicators related to waste reduction, process standardization, resource optimization, workflow simplification, and continuous improvement. Although lean systems are not digital technologies in themselves, they were included because digital transformation is more meaningful when technological tools are connected to process improvement rather than used in isolation (Rahardjo et al., 2023; Vlachos et al., 2023).

The main sustainability indicators were carbon footprint reduction, waste management efficiency, ethical sourcing score, and circular economy practices. Carbon footprint reduction reflected the reported percentage or index-based improvement in emissions-related performance. Waste management efficiency measured the extent to which firms reduced, reused, recycled, or more efficiently managed operational waste. Ethical sourcing score captured supplier responsibility, procurement transparency, and compliance with sustainability-oriented sourcing practices. Circular economy practices reflected reuse, recycling, product life-cycle awareness, and resource recovery activities.

Financial performance was represented by annual revenue, reported in monetary value and then checked for extreme outliers during data cleaning. Supply chain performance was treated as the principal outcome construct, reflecting improvements in efficiency, responsiveness, coordination, cost control, and sustainability-related operational performance.

2.5 Data Preparation and Quality Control

Before statistical analysis, the dataset was screened for missing values, duplicate entries, inconsistent responses, and implausible outliers. Missing values were first examined to determine whether they were random or concentrated within specific variables. Records with substantial missing information across core variables were excluded. For variables with limited missingness, mean or median imputation was considered depending on the distribution of the variable. Continuous variables were assessed for normality, dispersion, and extreme values.

Technology adoption variables were normalized to a 0–1 scale to allow comparison across AI, IoT, and blockchain indicators. Sustainability and financial variables were inspected using descriptive statistics, including mean, standard deviation, median, minimum, maximum, and interquartile range. Outliers were not removed automatically; instead, they were reviewed to determine whether they reflected plausible firm-level variation or data-entry error. This step was important because SME performance data can vary widely by sector, size, and market maturity.

To improve reproducibility, all data transformations were documented before model fitting. The analytical workflow included data cleaning, variable normalization, descriptive analysis, regression modelling, feature selection, cross-validation, and interpretation of model performance. Analyses were performed using SPSS and Python. Python-based procedures were used for data preprocessing, cross-validation, and model performance testing, while SPSS was used to verify descriptive and regression outputs.

2.6 Statistical Analysis

Descriptive statistics were first calculated to summarize the overall distribution of technology adoption, sustainability indicators, and financial outcomes. Mean and standard deviation were reported for approximately continuous variables, while median and interquartile range were considered where distributions were skewed. These summaries provided an initial understanding of digital maturity and sustainability performance among the 300 SMEs.

Multiple linear regression analysis was then conducted to examine the association between technology adoption and supply chain performance. The core regression model included AI adoption level, IoT adoption level, and blockchain adoption level as independent variables. Sustainability indicators and annual revenue were examined as performance-related outcomes and, where appropriate, as explanatory or control variables. The general model structure was expressed as follows:

Supply chain performance = β0 + β1(AI adoption) + β2(IoT adoption) + β3(Blockchain adoption) + β4(Sustainability indicators) + ε

Regression coefficients, standard errors, t-values, p-values, and 95% confidence intervals were reported. Statistical significance was evaluated at p < 0.05. However, interpretation was not based on p-values alone. The direction and magnitude of coefficients, confidence intervals, and practical relevance of the findings were also considered. This was important because statistically significant findings may not always represent meaningful managerial improvement, especially in heterogeneous SME datasets.

Model assumptions were examined before final interpretation. Linearity was assessed through residual plots, while multicollinearity was evaluated using variance inflation factors. Homoscedasticity was inspected through residual distribution, and normality of residuals was assessed using graphical and statistical checks. Where assumptions were not fully satisfied, results were interpreted cautiously, and the limitations were reported.

2.7 Feature Selection and Cross-Validation

To assess model robustness and reduce the risk of overfitting, feature selection and cross-validation were performed. Different combinations of predictors were tested to determine whether a smaller number of variables could explain supply chain performance with comparable or improved predictive accuracy. This step was useful because adding more variables does not always improve model quality; in some cases, additional predictors introduce noise, instability, or redundancy.

Cross-validation was conducted by repeatedly partitioning the dataset into training and validation subsets. Model accuracy was then compared across different feature sets. The best-performing model was identified based on cross-validation score, interpretability, and stability across validation folds. This procedure helped determine whether AI, IoT, blockchain, and sustainability-related indicators contributed consistently to supply chain performance prediction.

2.8 Ethical Considerations

The study used firm-level operational and performance data. No personal medical, biological, or patient-level information was collected. Participation was voluntary, and respondents were informed that the data would be analyzed in aggregate form for academic research purposes. Company identifiers were removed before analysis to protect confidentiality. Because the research involved organizational survey data rather than human clinical data, formal biomedical ethics approval was not required; however, confidentiality, informed participation, and responsible data handling principles were followed.

2.9 Reproducibility and Data Transparency

To support reproducibility, the study defined the sampling frame, eligibility criteria, variable structure, data-cleaning steps, and statistical procedures in sufficient detail for future replication. Future researchers can reproduce the analysis by collecting firm-level SME data using equivalent normalized adoption scores for AI, IoT, blockchain, lean practices, sustainability indicators, and financial outcomes. For full transparency, the final dataset, questionnaire items, coding scheme, and Python/SPSS syntax should be made available as supplementary materials where journal policy permits. This would allow independent verification of the regression outputs and cross-validation results, strengthening the reliability of the study.

3. Results

3.1 Descriptive Profile of Technology Adoption and Sustainability Indicators

The descriptive analysis provided an initial view of how the 300 SMEs were positioned in relation to Industry 4.0 adoption, sustainability practices, and financial performance. Overall, the firms showed a moderate level of digital technology adoption, although the pattern was not uniform across the three technologies examined. The mean IoT adoption level was 0.510, followed closely by AI adoption at 0.495, while blockchain adoption was slightly lower, with a mean value of 0.471 [Table 1]. These values suggest that, within the sampled SMEs, IoT and AI had gained somewhat stronger practical traction than blockchain, perhaps because real-time monitoring and predictive decision-support tools are often more immediately applicable to daily supply chain operations than distributed ledger systems.

The variation across firms was also notable. AI adoption ranged from 0.005 to 0.990, IoT adoption from 0.010 to 0.999, and blockchain adoption from 0.004 to 0.996

Table 1. Descriptive statistics for technology adoption, sustainability indicators, and financial characteristics among the sampled SMEs. The table summarizes the distribution of key variables across 300 small and medium-sized enterprises (SMEs), including artificial intelligence (AI), Internet of Things (IoT), and blockchain adoption levels; carbon footprint reduction; ethical sourcing score; waste management efficiency; circular economy practices; annual revenue; and duration of sustainability initiatives. Values are presented as count, mean, standard deviation, minimum, 25th percentile, median, 75th percentile, and maximum. Technology adoption indicators were normalized on a scale from 0 to 1, with higher values indicating more advanced implementation.

[Table 1]. This wide spread indicates that the sample included firms at very different stages of digital maturity. Some SMEs appeared to have only minimal engagement with advanced technologies, whereas others had moved toward more extensive integration. Such variation is useful analytically because it allows the study to examine how different levels of adoption relate to supply chain and sustainability outcomes.

The sustainability indicators also showed meaningful variation. The average carbon footprint reduction score was 54.755, while the mean waste management efficiency score was 54.994 [Table 1]. Circular economy practices showed a similar pattern, with a mean value of 55.433. These findings suggest that many firms had already adopted at least some sustainability-oriented practices, although the broad standard deviations imply that implementation intensity differed considerably among SMEs. This is consistent with the wider view that digital supply chain tools may support sustainability through improved monitoring, resource optimization, and traceability, but that the benefits depend strongly on organizational readiness and implementation quality (Frank et al., 2019; Taj et al., 2023; Riad et al., 2024).

The average annual revenue of the sampled firms was 171.61 million, with values ranging from 39.73 to 702.72 [Table 1]. This relatively wide range indicates substantial financial heterogeneity among the SMEs included in the study. The mean duration of sustainability initiatives was 5.653 years, suggesting that many firms had moved beyond very early-stage sustainability engagement. Even so, government policy support, organizational size, and industry sector were not fully quantified in the descriptive table. This is an important limitation because these factors may influence both the adoption of Industry 4.0 technologies and the outcomes associated with them.

3.2 Association Between Technology Adoption and Supply Chain Performance

Regression analysis was conducted to examine the relationship between technology adoption and supply chain performance. The model included AI adoption, IoT adoption, and blockchain adoption as key predictors. The constant term was positive, indicating a baseline level of supply chain performance even before accounting for the specific contribution of digital adoption variables [Table 2].

AI adoption showed a positive regression coefficient (B = 3.931), suggesting that firms with higher AI adoption tended to report better supply chain performance [Table 2]. This finding is conceptually consistent with earlier work showing that AI can support predictive analytics, demand forecasting, inventory optimization, and more responsive decision-making in complex operational settings (Lu, 2019; Aljohani, 2023). In practical terms, AI may help SMEs shift from reactive supply chain management toward more anticipatory planning, especially when AI tools are connected to real-time or historical operational data.

IoT adoption also demonstrated a positive coefficient (B = 3.461), indicating a favorable association with supply chain performance [Table 2]. This result is not unexpected, as IoT technologies can improve operational visibility through sensor-based tracking, real-time monitoring, and data capture across logistics, inventory, production, and energy-use systems. Prior studies similarly emphasize that IoT-enabled supply chains can become more transparent, responsive, and data-driven when firms are able to convert operational signals into managerial action (Malek et al., 2017; Wang et al., 2018; Taj et al., 2023).

Blockchain adoption had a smaller but still positive coefficient (B = 1.458) [Table 2]. Compared with AI and IoT, the effect size was more modest, suggesting that blockchain may contribute to supply chain performance in a more specialized way, particularly through traceability, verification, and transparency rather than direct operational optimization. This aligns with studies that describe blockchain as useful for strengthening trust, data integrity, and supply chain traceability, especially in multi-actor networks where verification is important (Rejeb et al., 2019; Feng et al., 2020; Charles et al., 2023).

However, the regression results should be interpreted carefully. Although the reported coefficients were positive, the confidence intervals shown in the table were relatively wide and, for some predictors, included negative lower bounds [Table 2]. This suggests that the statistical output should be checked again before making strong claims about significance. A more cautious interpretation is therefore appropriate: the results indicate a positive pattern between Industry 4.0 technology adoption and supply chain performance, but the strength and statistical certainty of these relationships require careful validation.

3.3 Comparative Influence of AI, IoT, and Blockchain

When the three technology variables were compared, AI

Table 2. Regression coefficients for the association between Industry 4.0 technology adoption and supply chain performance. The table presents the estimated regression coefficients, standard errors, t-values, p-values, and 95% confidence intervals for AI, IoT, and blockchain adoption levels. Positive coefficients indicate a favorable directional association between technology adoption and supply chain performance. The statistical outputs, particularly the reported p-values and confidence intervals, should be verified using the final regression model before publication.

Figure 1. Conceptual overview of IoT–AI integration in smart supply chain management. The figure illustrates how IoT-enabled sensors and connected devices collect real-time operational data across supply chain activities. These data are processed through AI-supported analytics and decision-making systems to improve forecasting, inventory planning, logistics coordination, operational visibility, and resource optimization. The framework provides the conceptual basis for evaluating how IoT and AI adoption may contribute to supply chain performance in SMEs. Adapted from Aliahmadi et al. (2022).

and IoT appeared to have stronger associations with supply chain performance than blockchain [Table 2]. This pattern may reflect the more direct role of AI and IoT in everyday supply chain decision-making. IoT provides the real-time data layer, while AI supports interpretation, prediction, and optimization. Together, these technologies can help firms monitor operations, reduce delays, forecast demand, and allocate resources more effectively (Frank et al., 2019; Lu, 2019; Aljohani, 2023).

Blockchain, by contrast, may produce benefits that are less immediately visible in general performance models. Its value is often linked to traceability, auditability, supplier verification, and transactional transparency. These benefits are important, but they may not always translate directly into short-term efficiency or revenue gains, particularly among SMEs with limited technological infrastructure or lower supply chain complexity (Bosona & Gebresenbet, 2023; Charles et al., 2023). This may help explain why blockchain adoption showed a smaller coefficient than AI and IoT in the present analysis.

3.4 Sustainability-Linked Performance Outcomes

The descriptive results suggested that sustainability performance was an important dimension of smart supply chain development. Carbon footprint reduction, waste management efficiency, and circular economy practices all showed moderate-to-high mean values across the sampled firms [Table 1]. These findings imply that many SMEs were not adopting digital tools only for productivity reasons, but also in relation to environmental and operational sustainability goals.

IoT and AI may be especially relevant in this area. IoT systems can monitor energy consumption, emissions, resource use, and waste generation, while AI can identify inefficiencies, predict operational bottlenecks, and recommend corrective actions. When these tools are integrated into supply chain processes, firms may be better positioned to reduce waste and improve environmental performance without necessarily compromising operational efficiency (Wang et al., 2018; Rahardjo et al., 2023; Riad et al., 2024). The present findings therefore support the broader argument that smart supply chains should be evaluated not only by speed or cost reduction, but also by their contribution to sustainability-oriented performance.

At the same time, the results also indicate that sustainability improvement is unlikely to depend on technology alone. The average duration of sustainability initiatives was more than five years, suggesting that organizational commitment and accumulated experience may also influence outcomes [Table 1]. In this regard, lean systems remain important because they provide a managerial logic for reducing waste, simplifying processes, and improving resource use. Digital technologies may strengthen lean implementation, but they cannot replace the need for disciplined process improvement (Moeuf et al., 2018; Vlachos et al., 2023).

3.5 Feature Selection and Model Performance

The cross-validation analysis examined how model performance changed as different numbers of features were included. The results showed that predictive accuracy did not improve steadily with the addition of more variables. Instead, the best cross-validation score appeared when a smaller number of features was selected, with accuracy reaching approximately 0.62 when two features were included [Figure 2]. After this point, performance fluctuated and, in some models, declined as additional features were added.

This pattern suggests that a more compact model may explain supply chain performance more effectively than a model with too many predictors. In practical terms, this may mean that a few carefully selected indicators—most likely those linked to AI and IoT adoption—carry stronger predictive value than a broader set of weakly related variables. The finding is consistent with the principle that predictive models can become less stable when additional features introduce noise, redundancy, or multicollinearity (Alomar, 2022).

The cross-validation results also reinforce the need for careful feature selection in smart supply chain research. Although Industry 4.0 systems involve many technologies, not every variable contributes equally to model performance. For SMEs, where data availability and technical resources may be limited, a simpler and more interpretable model may sometimes be more useful than a highly complex one. This is particularly important when the goal is to support managerial decisions rather than merely maximize statistical sophistication.

3.6 Conceptual Interpretation of IoT-Enabled and AIoT-Based Supply Chain Models

The conceptual figures included in the manuscript help situate the empirical findings within broader smart supply

Figure 2. Cross-validation performance across different numbers of selected predictive features. The figure shows changes in model accuracy as the number of selected features increases. The highest cross-validation score is observed when a smaller number of variables is included, suggesting that a compact model may provide stronger predictive performance than a more complex model with additional features. The findings highlight the importance of feature selection when developing interpretable supply chain performance models. Adapted from Alomar (2022).

Figure 3. Cognitive design principles for IoT-enabled smart supply chains. The figure presents interconnected capabilities that support adaptive and intelligent supply chain systems, including self-maintaining, self-aware, self-predicting, self-optimizing, self-configuring, self-organizing, self-comparing, and self-adaptive functions. Together, these capabilities illustrate how IoT-enabled supply chains can respond dynamically to operational changes, reduce inefficiencies, and support continuous improvement. Adapted from Radanliev et al. (2020).

Figure 4. AIoT-based framework for sustainable smart supply chain management. The figure illustrates an integrated artificial intelligence of things (AIoT) framework linking AI, IoT, RFID, cloud technologies, sensors, robotics, big data analytics, additive manufacturing, supplier collaboration portals, and advanced trace-and-tracking systems. The framework emphasizes how these technologies can support digital planning, sourcing, logistics, environmental management, health and safety, social responsibility, and broader sustainability objectives. Adapted from Aliahmadi et al. (2022).

chain theory. The IoT–AI integration model illustrates how data from connected devices can move through platform-level processing and AI-based decision systems to generate actionable supply chain outcomes [Figure 1]. This supports the logic of the empirical findings, where IoT and AI showed stronger associations with supply chain performance than blockchain. In practice, IoT supplies the operational visibility, while AI helps convert that visibility into forecasting, optimization, and decision support (Aliahmadi et al., 2022).

The cognitive design framework further expands this interpretation by emphasizing self-maintaining, self-optimizing, self-configuring, self-predicting, and self-adaptive capabilities within digital supply chains [Figure 3]. These capabilities are particularly relevant to SMEs seeking to improve responsiveness without adding excessive administrative complexity. The framework suggests that smart supply chains are not only technology-enabled but also cognitively adaptive, meaning they can learn from operational conditions and adjust processes over time (Radanliev et al., 2020).

Similarly, the AIoT-based sustainable supply chain framework highlights the connection between AI, IoT, big data analytics, RFID, cloud technologies, trace-and-tracking systems, and sustainability-oriented supply chain management [Figure 4]. This model is useful for interpreting the observed relationship between technology adoption and sustainability indicators. It suggests that environmental management, health and safety, social responsibility, logistics, sourcing, and digital planning are interconnected rather than isolated supply chain functions (Aliahmadi et al., 2022).

3.7 Summary of Key Findings

In summary, the results indicate that SMEs with higher levels of AI and IoT adoption tended to show better supply chain performance, while blockchain adoption showed a smaller but still positive association [Table 2]. Sustainability indicators, including carbon footprint reduction, waste management efficiency, and circular economy practices, were moderately developed across the sample and appeared to be meaningfully connected to the broader smart supply chain transformation process [Table 1]. Cross-validation further suggested that a limited number of well-selected features may provide stronger predictive performance than larger, more complex models [Figure 2].

Taken together, the findings support the argument that smart supply chain performance in SMEs is shaped by a combination of digital adoption, lean-oriented thinking, and sustainability practice. Still, the results should be interpreted with caution until the regression outputs, particularly the p-values and confidence intervals, are verified. The evidence points in a promising direction, but stronger methodological transparency and statistical validation would make the findings more robust and publication-ready.

4. Discussion

4.1 Principal Findings and Overall Interpretation

This study examined how the adoption of IoT, artificial intelligence, blockchain, and lean-oriented systems relates to supply chain performance, sustainability outcomes, and financial indicators among SMEs. Taken together, the findings suggest that Industry 4.0 technologies are not merely technical add-ons to supply chain operations; rather, when implemented with some degree of organizational readiness, they may become part of a broader performance-improvement structure. The descriptive findings showed moderate adoption of IoT and AI across the sampled firms, with blockchain adoption slightly lower [Table 1]. This pattern is understandable, as IoT and AI often provide more immediate operational benefits, such as real-time monitoring, predictive analytics, inventory planning, and logistics optimization, while blockchain may require more complex inter-organizational coordination before its benefits become visible (Frank et al., 2019; Rejeb et al., 2019; Charles et al., 2023).

The regression results further indicated positive associations between technology adoption and supply chain performance, particularly for AI and IoT [Table 2]. Although these findings are promising, they should be interpreted with some caution because the reported confidence intervals and statistical values need careful verification before strong inferential claims are made. Still, the direction of the coefficients is consistent with the broader literature, which suggests that data-driven technologies can support more responsive, transparent, and efficient supply chains (Wang et al., 2018; Taj et al., 2023; Riad et al., 2024). In this sense, the study contributes to the growing evidence that SMEs can benefit from digital transformation, provided that adoption is connected to practical operational needs rather than pursued as a symbolic modernization effort.

4.2 Role of IoT in Supply Chain Visibility and Responsiveness

One of the clearest implications of the study is the importance of IoT as a visibility-enhancing technology. The descriptive results showed that IoT adoption was slightly higher than AI and blockchain adoption [Table 1], which may indicate that SMEs are increasingly recognizing the value of real-time operational data. IoT-enabled systems can collect information from sensors, warehouses, transportation routes, machines, and energy-use points, allowing managers to observe supply chain conditions more directly. This visibility can be especially valuable in SMEs, where supply chain disruptions, inventory errors, and resource inefficiencies may have a proportionally larger effect on profitability and customer satisfaction.

The positive association between IoT adoption and supply chain performance [Table 2] is consistent with previous studies showing that IoT can improve real-time monitoring, energy efficiency, and operational coordination (Malek et al., 2017; Wang et al., 2018). In practical terms, IoT does not improve supply chains simply because devices are connected. Its value emerges when the collected data are translated into decisions, such as adjusting stock levels, identifying delayed shipments, tracking equipment conditions, or reducing energy waste. This distinction is important. For SMEs, technology adoption may be less about installing advanced systems and more about using accessible data to make better day-to-day decisions.

The IoT–AI integration model included in the manuscript supports this interpretation by showing how connected devices can generate operational data that feed into analytical and decision-support processes [Figure 1]. In this way, IoT acts as the sensory layer of the smart supply chain. It helps firms detect what is happening, but it often requires complementary analytical tools, managerial routines, and lean practices to turn visibility into measurable performance improvement.

4.3 Contribution of AI to Prediction, Optimization, and Decision-Making

AI adoption also showed a positive relationship with supply chain performance [Table 2]. This finding is meaningful because AI is increasingly used to support demand forecasting, inventory control, supplier evaluation, risk prediction, warehouse automation, and transportation optimization. Prior literature has similarly emphasized the potential of AI and predictive analytics to improve agility and reduce uncertainty in supply chain systems (Lu, 2019; Aljohani, 2023; Riad et al., 2024).

However, AI should not be viewed as a standalone solution. Its performance depends on the quality, completeness, and timeliness of available data. In SMEs, data may be fragmented, manually recorded, or stored across disconnected systems. Therefore, the benefits of AI may be strongest when it is combined with IoT-based data capture, standardized processes, and managerial capacity to act on model outputs. The cross-validation results support this more restrained interpretation, as model performance appeared to improve most when a limited number of relevant features were selected rather than when more variables were added indiscriminately [Figure 2]. This suggests that, in SME contexts, simpler and better-targeted digital models may sometimes be more useful than complex systems with many weakly informative inputs.

This finding also has methodological importance. It implies that future studies should not assume that all Industry 4.0 indicators contribute equally to supply chain performance. Instead, researchers should examine which combinations of technologies provide the strongest explanatory or predictive value. In the present study, AI and IoT appear to carry stronger performance relevance than blockchain, at least within the structure of the available dataset [Table 2].

4.4 Blockchain Adoption and Traceability-Oriented Benefits

Blockchain adoption had a smaller positive coefficient compared with AI and IoT [Table 2]. This does not necessarily mean that blockchain is unimportant. Rather, its contribution may be more specific and less directly reflected in general supply chain performance measures. Blockchain is often most useful where traceability, data integrity, supplier verification, transaction transparency, and product origin authentication are critical (Rejeb et al., 2019; Feng et al., 2020; Bosona & Gebresenbet, 2023). These benefits may be particularly relevant in food, agriculture, pharmaceuticals, logistics, and other sectors where trust and compliance are central.

For many SMEs, however, blockchain adoption may remain at an early or partial stage. Implementation may require technical expertise, investment, partner cooperation, and standardized data-sharing arrangements. Without these conditions, blockchain may not produce immediate gains in efficiency or revenue. This may explain why its coefficient was positive but smaller than that of AI and IoT [Table 2]. The finding is consistent with the idea that blockchain strengthens the trust and transparency layer of supply chains, while IoT and AI more directly influence monitoring, prediction, and operational optimization (Charles et al., 2023).

4.5 Lean Systems as the Process Foundation of Smart Supply Chains

Although the statistical model focused mainly on AI, IoT, and blockchain, lean systems remain conceptually central to the interpretation of the findings. Lean practices provide the operational discipline required to convert digital information into process improvement. Without lean thinking, firms may collect large volumes of data without reducing waste, shortening lead times, or improving resource use. This is particularly relevant for SMEs, where limited financial and human resources make efficiency gains essential (Moeuf et al., 2018; Rahardjo et al., 2023).

The manuscript’s broader framework suggests that lean systems can work alongside IoT and AI by identifying waste, simplifying workflows, and supporting continuous improvement. This is consistent with the view that Industry 4.0 and lean manufacturing are not opposing models, but potentially complementary systems when implemented thoughtfully (Vlachos et al., 2023). IoT can reveal where inefficiencies occur, AI can help predict or prioritize corrective actions, and lean practices can guide how those corrections are implemented. In this sense, the smart supply chain should not be understood only as a digital system; it is also a managerial system that depends on process discipline and organizational learning.

4.6 Sustainability Implications of Digital and Lean Integration

The sustainability indicators in the study showed moderate-to-high average values for carbon footprint reduction, waste management efficiency, and circular economy practices [Table 1]. These results suggest that SMEs in the sample were engaging with sustainability not only as a compliance issue, but also as part of operational improvement. This is important because supply chain sustainability increasingly depends on the ability to monitor resources, reduce waste, improve traceability, and coordinate across multiple actors.

IoT and AI may support these goals in different but complementary ways. IoT can track energy use, emissions, inventory movement, and waste generation in real time, while AI can identify inefficiencies and recommend more efficient planning decisions. When combined with lean systems, these technologies may help firms reduce unnecessary movement, overproduction, excess inventory, and material waste. Previous studies have similarly argued that Industry 4.0 technologies can support more sustainable and resilient supply chains when integrated with operational strategy rather than implemented in isolation (Raja Santhi & Muthuswamy, 2022; Rahardjo et al., 2023; Riad et al., 2024).

The AIoT-based sustainable supply chain framework further reinforces this interpretation by linking AI, IoT, RFID, big data analytics, cloud systems, trace-and-tracking technologies, and sustainability dimensions such as environmental management, health and safety, and social responsibility [Figure 4]. The findings of the present study fit reasonably well within this framework, as firms with stronger digital adoption also appeared to report better sustainability-related indicators [Table 1]. Still, the relationship should not be interpreted as purely technological. Sustainability performance is also shaped by leadership commitment, sectoral pressure, regulatory expectations, supplier behavior, and the maturity of internal management systems.

4.7 Cognitive and Adaptive Capabilities in Smart Supply Chains

The cognitive design framework included in the manuscript provides another useful lens for interpreting the results [Figure 3]. Smart supply chains increasingly require self-monitoring, self-prediction, self-optimization, and self-adaptation. These capabilities are not fully achieved by technology adoption alone, but they become more feasible when firms combine IoT-based data capture, AI-supported analytics, and lean-driven process improvement (Radanliev et al., 2020).

For SMEs, cognitive supply chain capabilities may develop gradually. A firm may first adopt IoT for tracking inventory or energy use, then apply AI tools for forecasting or route optimization, and later integrate these insights into broader decision-making routines. The cross-validation results, which suggested that a smaller number of strong features may produce better model performance [Figure 2], are also consistent with this gradual approach. SMEs may not need to adopt every advanced technology at once. Instead, they may benefit more from identifying the most relevant digital tools for their specific operational bottlenecks.

4.8 Financial and Managerial Implications

The descriptive findings showed considerable variation in annual revenue across the sampled firms [Table 1]. This variation suggests that financial performance among SMEs may be influenced by many factors beyond technology adoption alone, including firm size, sector, market access, managerial capability, and supply chain complexity. Nevertheless, the positive coefficients for AI, IoT, and blockchain indicate that digital adoption may contribute to performance improvement when these tools are properly aligned with operational goals [Table 2].

From a managerial perspective, the findings suggest that SMEs should approach digital transformation selectively and strategically. AI and IoT may be prioritized when the immediate goal is to improve visibility, forecasting, efficiency, and responsiveness. Blockchain may be more relevant where traceability, auditability, and supplier trust are central concerns. Lean systems should remain a foundation across all stages because they help ensure that technology adoption leads to measurable process improvement rather than unnecessary complexity.

The study also implies that SMEs should avoid treating Industry 4.0 adoption as a one-time investment. Instead, digital transformation should be understood as a staged process involving data readiness, employee capability, process redesign, and continuous performance monitoring. This is particularly important because SMEs often operate with limited budgets and cannot afford poorly targeted technology implementation.

4.9 Limitations of the Study

Several limitations should be acknowledged. First, the study used a cross-sectional design, meaning that it can identify associations but cannot prove causality. Although higher technology adoption was positively associated with supply chain performance, it is also possible that better-performing firms had more resources to invest in AI, IoT, and blockchain systems. A longitudinal design would be needed to examine whether technology adoption leads to sustained performance improvement over time.

Second, the dataset included 300 SMEs, but the manuscript does not fully quantify the role of industry sector, organizational size, government policy support, or digital readiness [Table 1]. These factors may strongly influence both technology adoption and supply chain outcomes. For example, an SME in logistics or manufacturing may benefit from IoT differently than a service-oriented SME. Similarly, firms operating under strong regulatory or sustainability pressure may adopt traceability tools more quickly than firms in less regulated sectors.

Third, the regression results require verification because some reported p-values, t-values, and confidence intervals appear inconsistent [Table 2]. Therefore, strong claims of statistical significance should be avoided until the regression output is recalculated and confirmed. The current evidence is best interpreted as indicating a positive directional relationship rather than definitive proof of significant causal effects.

Fourth, the measurement of lean systems requires further clarification. While lean principles are discussed throughout the manuscript, the empirical model does not clearly show a distinct lean adoption variable. Future revisions should either define lean as a measurable construct or adjust the title and claims to reflect the variables actually included in the statistical model.

4.10 Future Research Directions

Future research should build on these findings by using longitudinal designs, larger and more sector-specific samples, and more detailed measures of digital maturity. A future study could examine whether firms that adopt IoT and AI over several years show sustained improvements in efficiency, sustainability, and profitability. It would also be useful to test whether lean practices mediate or moderate the relationship between technology adoption and supply chain performance.

Additional research should also examine the role of government policy, infrastructure readiness, employee digital skills, and financial constraints in SME technology adoption. Prior studies have already shown that SMEs face unique barriers when implementing Industry 4.0 systems, including cost, technical expertise, and integration challenges (Moeuf et al., 2018; Hangl et al., 2022). Understanding these barriers more precisely would help policymakers and managers design more realistic digital transformation pathways.

Finally, future work should investigate how AI, IoT, blockchain, and lean systems interact rather than evaluating them only as separate predictors. The evidence from this study suggests that AI and IoT may be especially influential, but their combined effect with lean process improvement and sustainability strategy may be more important than any single technology alone.

4.11 Concluding Interpretation

Overall, the study suggests that IoT, AI, blockchain, and lean-oriented systems may contribute to smarter, more sustainable, and more responsive supply chains among SMEs. The strongest practical message is that digital transformation works best when technologies are connected to real operational problems: poor visibility, inefficient resource use, weak forecasting, waste generation, and limited traceability. AI and IoT appear particularly relevant for performance improvement, while blockchain may offer more specialized benefits related to trust and transparency [Table 2]. Sustainability outcomes also appear closely connected to smart supply chain development, especially when digital tools are supported by lean thinking and long-term organizational commitment [Table 1; Figure 4].

Still, the findings should be presented with appropriate caution. The study provides useful evidence of positive associations, but stronger methodological transparency, clearer variable definitions, and verified statistical outputs are needed before firm causal conclusions can be drawn. With these improvements, the manuscript can make a meaningful contribution to the literature on SME digital transformation, sustainable supply chains, and Industry 4.0-enabled operational performance.

5. Conclusion

This study highlights the potential value of integrating IoT, artificial intelligence, blockchain, and lean-oriented practices into SME supply chain operations. The findings suggest that AI and IoT may have particularly meaningful roles in improving visibility, prediction, coordination, and operational responsiveness, while blockchain may contribute more specifically to traceability and transparency. Sustainability outcomes, including carbon footprint reduction, waste management efficiency, and circular economy practices, also appear to be closely connected with smart supply chain development. Still, the results should be interpreted with care. The study identifies positive associations rather than confirmed causal effects, and the regression outputs require careful statistical verification before strong claims are made. Overall, the manuscript supports the idea that digital transformation in SMEs is most effective when technology adoption is not treated as a standalone investment but is combined with lean thinking, sustainability planning, and practical managerial capacity.

Author Contributions

M.F.A.B. conceptualized the study, conducted data analysis, and drafted the manuscript. M.A.R. contributed to methodology development, literature review, data interpretation, and manuscript revision. Both authors reviewed and approved the final manuscript.

Acknowledgements

The authors sincerely thank the SME owners, managers, and supply chain practitioners who participated in this study and generously provided firm-level data. Their cooperation was essential to the completion of this research. The authors also acknowledge the institutional support received throughout the data collection and analysis phases of this work.

Conflict of Interest

The authors declare no conflict of interest.

Financial Disclosure

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. The authors confirm that no financial relationship or sponsorship of any kind influenced the study design, data collection, analysis, interpretation, or decision to publish.

References