Sustainable logistics and supply chain performance with human–machine collaboration in the context of Industry 5.0: a holistic structural model analysis

In recent years, supply chain management (SCM) has become strategically important in many sectors, and significant developments have taken place in this field. However, environmental concerns such as waste generation associated with supply chain activities, increasing carbon emissions and the rapid depletion of natural resources have also come to the fore in the shadow of this growth. SCM, which encompasses corporate processes that interact directly with the environment, such as procurement, production and distribution, plays a critical role in achieving sustainable development goals. Today, the majority of businesses are part of at least one global supply network, and the way these networks are managed is decisive in terms of reducing environmental impacts and ensuring resource conservation (Shebeshe and Sharma, 2024). However, the increasing complexity, volatility and sustainability requirements of global supply chains render traditional decision-making approaches inadequate. Classical decision support systems, typically based on historical data and static, linear models, are becoming increasingly ineffective in managing the uncertainties created by factors such as demand volatility, geopolitical risks, emission constraints, and environmental, social and governance (ESG) responsibilities (Lin, 2025). This situation necessitates new strategic approaches towards the digitalisation, flexibilisation and resilience of supply chains.

Throughout industrial history, production and logistics systems have undergone fundamental transformations in different periods. The First Industrial Revolution, which began with the invention of the steam engine, represented the transition from craft production to mechanical production, while the Second Industrial Revolution enabled mass production with the introduction of electricity and assembly lines. The Third Industrial Revolution laid the foundation for automation through the integration of electronic and information technologies into production processes. The Fourth Industrial Revolution (Industry 4.0), on the other hand, represents a new wave of transformation in production through the convergence of nine key digital technologies: artificial intelligence, the Internet of Things (IoT), cyber-physical systems, cloud computing and big data (Locatelli et al., 2024). The fundamental principle of Industry 4.0 is to create a data-driven “smart manufacturing” structure where machines and devices communicate with each other throughout their entire life cycle and to minimise human intervention in the production process (Xu, 2020; Alohali et al., 2022; Piardi et al., 2024; Frank et al., 2019; Wu et al., 2024). Although this machine learning and automation-based approach has increased efficiency, it has relatively pushed the creative role of humans in the production process into the background. In this context, Industry 5.0 has emerged as a paradigm that reintegrates technological progress with human intelligence, creativity and ethical values (Maddikunta et al., 2022; Dwivedi et al., 2023). As a response to the technology-centric structure of Industry 4.0, this paradigm develops a production philosophy that prioritises not only automation but also collaboration between humans and machines, sustainability and social welfare. In this respect, Industry 5.0 approaches innovation not merely as a technical process but as a socio-technical transformation, redefining the nature of human–machine interaction in both production and logistics systems (Fani et al., 2024; Hasan et al., 2024). The fundamental goal of the paradigm is to increase both the operational efficiency (OE) of businesses and the well-being of employees through advanced technologies, while ensuring the responsible use of natural resources (Dacre et al., 2025). Industry 5.0 goes beyond digitalisation and automation, addressing sustainability, resilience and human-centredness within a holistic framework. Resilience refers to the capacity to ensure the continuity of industrial production by minimising disruptions and uncertainties during times of crisis. Sustainability focuses on environmental factors that encourage reduced energy consumption and the reuse of existing resources. The human-centric approach places the needs and expectations of individuals at the heart of the production process, enhancing workforce well-being through personalisation, ergonomics and human–machine interaction (Zeb et al., 2024). These three dimensions not only increase OE but also reinforce social welfare and environmental responsibility. Although production and logistics systems benefit significantly from Industry 4.0 technologies in terms of performance, the cognitive, physical and psychosocial effects of these technologies on humans are not yet fully understood. Although automation, robotics and digitalisation are becoming increasingly widespread, there is a general consensus that humans remain an indispensable element of operational systems (Neumann et al., 2021; Grosse et al., 2023). Accordingly, Industry 5.0 brings human intelligence back to the centre of production by combining advanced technologies with human capabilities to develop more flexible and user-centred production systems. This transformation brings about fundamental changes not only in production systems but also in SCM structures. AI-supported decision systems, big data analytics and automation-based solutions provide efficiency, visibility and agility in logistics processes (Rainer et al., 2025; Lin, 2025). The reactive, resource-intensive and coordination-deficient structure of traditional logistics systems is being addressed through real-time data-driven digital SCM approaches. These technologies offer innovative solutions in many areas, from demand forecasting to fleet management, increasing OE and aligning it with sustainability goals.

When considered together, sustainable SCM and Industry 5.0 approaches emerge as two complementary elements that transform both the environmental and economic performance of businesses. Technologies such as blockchain, collaborative robots (cobots) and augmented reality enable the traceability of products, ensure ethical sourcing practices and make human–machine interaction safer (Jamil et al., 2024). The information-sharing and quality-enhancing structure of Industry 4.0, combined with the human-centric approach of Industry 5.0, enables production systems to become not only intelligent but also responsive and sustainable (Fang et al., 2025). This transformation enables businesses to develop more agile, resilient and customer-focused solutions in the face of increasing personalisation demands, shorter delivery times and competitive pressure (Andres et al., 2024). However, in this era of accelerating digitalisation, macroeconomic changes, crises such as pandemics and geopolitical tensions have increased the fragility of supply chains. The COVID-19 pandemic demonstrated that no business is completely immune to such disruptions, necessitating the restructuring of production and logistics networks (Carayannis et al., 2023; Karmaker et al., 2023). This process has clearly highlighted the need for supply chains to be not only efficient but also resilient and sustainable. The need for businesses to develop new strategies focused on digitalisation, resilience and inclusivity in the post-pandemic era aligns with Industry 5.0's human-centric transformation vision (Castagnoli et al., 2024; Guo et al., 2023). Ultimately, the transition from Industry 4.0 to 5.0 represents an effort to balance technological progress with human values and sustainability principles. This integrated approach aims not only to provide a competitive advantage for production and supply chains but also to support economic resilience, environmental responsibility and social welfare. Therefore, achieving sustainable performance in today's industrial ecosystem requires addressing technological innovation alongside human–machine collaboration (HMC), ethical production and systemic resilience.

While the existing literature has examined the impact of Industry 4.0 technologies on OE, automation and resource utilisation, these relationships are typically analysed in isolation and with a predominantly technology-centric focus. In particular, the interaction between HMC, operational processes, RE and workforce capability (WC) within a unified sustainability framework remains relatively underexplored.

This study addresses this gap by developing and empirically testing an integrated structural model that simultaneously examines the relationships between HMC, OE, RE, WC, and sustainable logistics and supply chain performance (SLSCP) within the context of Industry 5.0. It further explores the indirect pathways through which HMC is associated with sustainable performance, particularly through the mediating roles of operational and RE. The analysis is based on survey data collected from 118 academics and industry professionals in Italy, providing complementary theoretical and practice-oriented perspectives. Accordingly, this study contributes to the literature by providing a more integrated understanding of the relationships among these constructs within the Industry 5.0 framework.

Research on Industry 5.0 in the field of sustainable logistics and supply chains goes beyond technological innovation and focuses on the design of human-centric, resilient and environmentally responsible production systems. The existing literature offers significant contributions, particularly in explaining the relationships between HMC, OE, RE, WC and SLSCP. In this context, Table 1 below summarises the fundamental studies conducted at the intersection of Industry 5.0 and sustainable logistics and SCM, presenting the research topics, methods used and main findings in a comparative manner. This table not only strengthens the theoretical basis of the current study but also contributes to identifying gaps in the literature and research opportunities.

When Table 1 is examined generally, it can be observed that Industry 5.0 paradigms have contributed to an integrated transformation in production and logistics systems in terms of HMC, RE and sustainability (Maddikunta et al., 2022; Grosse et al., 2023; Dacre et al., 2025). The literature generally suggests that technological tools such as AI, digital twins and cobots are associated with improvements in OE while also enhancing employee well-being, safety and job satisfaction (Stecke and Mokhtarzadeh, 2022; Pasparakis et al., 2023; Płaza et al., 2025). In this regard, human-centric design and ergonomics applications indicate that technological innovation may act as a catalyst not only for productivity but also for social sustainability (Nasir et al., 2025; Locatelli et al., 2024). Furthermore, digital technologies such as big data, AI and blockchain support environmental sustainability by improving visibility, agility and resource management levels in supply chains (Loske and Klumpp, 2021; Sharma and Gupta, 2024a; Wu et al., 2025). On the other hand, AI is often associated with improvements in efficiency, sustainability and flexibility across industrial processes, from production to supply chain coordination. However, the ethical, economic and governance implications accompanying this transformation also require careful consideration. Therefore, balancing technological progress with ethical principles, accountability and human-centredness remains an important aspect of the Industry 5.0 approach (Nazarian and Khan, 2024; Wu et al., 2024). Overall, this body of literature suggests that Industry 5.0 represents not only a technological evolution but also a multi-layered transformation paradigm that reshapes the balance between people, the environment and operational processes.

Although the studies summarised in Table 1 provide valuable insights into Industry 5.0, sustainable logistics and HMC, several important limitations can be identified. First, a significant portion of the existing research adopts conceptual, qualitative, or multi-criteria decision-making approaches, while empirical studies testing integrated structural relationships remain relatively limited. Second, prior studies tend to focus on specific dimensions such as technological drivers, circular economy practices, or human–robot interaction, often examining these elements independently rather than within a unified analytical framework. This fragmented approach makes it difficult to fully understand how human, operational and resource-based factors interact to shape sustainable supply chain performance. In addition, previous studies report inconsistent findings regarding the organisational outcomes of Industry 5.0, particularly with respect to WC and the translation of human-centric practices into sustainability performance. Third, although HMC is frequently highlighted as a key component of Industry 5.0, its role is generally discussed at a conceptual level or in relation to isolated outcomes such as productivity or job satisfaction. The mechanisms through which HMC influences sustainability-oriented performance outcomes, particularly through operational and RE, remain insufficiently explored in empirical research. Finally, the relationship between WC and sustainability performance presents mixed and underdeveloped findings in the literature, with existing studies reporting inconsistent evidence regarding its direct contribution to sustainability outcomes. While several studies emphasise the importance of human-centric approaches, the direct and indirect contributions of WC to sustainable supply chain performance have not been clearly established. Accordingly, this study extends the existing literature by integrating these interrelated dimensions within a single empirical framework, providing a more comprehensive understanding of how human, operational and resource-based capabilities jointly contribute to SLSCP in the context of Industry 5.0.

This section discusses the relationships between the variables of the structural model developed in line with Industry 5.0's human-centric, resilient and sustainable production approach, based on the findings of the existing literature and formulates research hypotheses. Within the literature, a prominent trend is that the variables of HMC, OE, RE and WC are among the key determinants of SLSCP. In this context, the theoretical background and possible interactions of each variable are discussed in detail in the following subsections.

From a theoretical perspective, the relationships proposed in this study can be understood through the integration of socio-technical systems theory, the resource-based view (RBV) and dynamic capabilities. Socio-technical systems theory emphasises the interdependence between human and technological components within organisational systems, suggesting that performance outcomes emerge from the effective alignment of social and technical subsystems (Sony and Naik, 2020; Farivar et al., 2026). In this context, HMC represents a core mechanism through which human capabilities and advanced technologies interact to shape operational processes (Xu and Gao, 2024). In addition, the RBV highlights the strategic importance of organisational resources and capabilities, including human capital and technological competencies, in achieving sustainable competitive advantage (Rivard et al., 2006; El Nemar et al., 2025). Within this framework, WC and RE can be considered as critical organisational assets that influence long-term performance outcomes (Darcy et al., 2014; Lubis, 2022). Furthermore, the dynamic capabilities perspective provides a complementary lens by explaining how organisations integrate, build, and reconfigure internal and external competencies to respond to changing environments. OE, in this sense, can be interpreted as a capability that enables the transformation of technological inputs and human resources into sustainability-oriented performance outcomes. Taken together, these theoretical perspectives provide a foundation for understanding how HMC, OE, RE and WC interact within the Industry 5.0 context to influence SLSCP.

Industry 5.0 paradigms represent a production approach in which human intelligence and advanced automation systems are integrated in a complementary manner. Within this context, HMC combines the flexibility, intuition and problem-solving capabilities of humans with the precision, speed and reliability of machines (Maddikunta et al., 2022; Dacre et al., 2025). Unlike traditional automation systems, collaborative robots (cobots) enable direct interaction between humans and machines, improving process efficiency, quality and safety while reducing cycle times and error rates (Stecke and Mokhtarzadeh, 2022; Keshvarparast et al., 2024). In addition, HMC contributes to employee well-being by reducing ergonomic risks, balancing cognitive and physical workloads, and supporting learning processes. These improvements are closely associated with enhanced OE and WC (Pasparakis et al., 2023). Moreover, HMC facilitates RE by enabling energy savings, material optimisation and waste reduction through the use of digital technologies such as AI, IoT and digital twins (Wu et al., 2024; Lin, 2025).

From a socio-technical systems perspective, these outcomes reflect the alignment between human capabilities and technological systems, which is expected to improve overall system performance. Socio-technical systems theory suggests that organisational performance is maximised when technological capabilities are aligned with human knowledge, skills and decision-making processes. Therefore, HMC is conceptualised as the primary socio-technical mechanism through which Industry 5.0 technologies generate operational and organisational benefits. Accordingly, HMC is expected to positively influence OE, RE and WC.

OE is widely recognised as a critical element in sustainable production and logistics systems, particularly in terms of reducing resource consumption and improving process performance. Higher levels of OE contribute to the reduction of energy and material waste, thereby supporting more efficient resource utilisation (Loske and Klumpp, 2021; Lin, 2025). In addition, digitalisation, AI-based planning systems and predictive maintenance applications enhance process transparency and flexibility, which further supports improvements in RE (Wu et al., 2024; Sharma and Gupta, 2024a).

OE also plays an important role in shaping WC by creating structured, secure and learning-oriented work environments. These conditions facilitate the development of employees' digital skills and adaptive capabilities, contributing to long-term organisational learning and performance (Schröder et al., 2023; Locatelli et al., 2024).

From a dynamic capabilities perspective, OE represents an organisational capability that enables firms to integrate, reconfigure, and leverage technological and human resources in response to changing operational requirements. Accordingly, OE can be seen as a multidimensional capability that influences both RE and WC.

SLSCP involves not only the reduction of environmental impacts but also the effective management of social and organisational dimensions. RE contributes to environmental sustainability through practices such as energy conservation, waste reduction and circular resource utilisation (Jayarathna et al., 2023; Kharayat and Gupta, 2025). Efficient resource use is also associated with lower emissions and cost reductions, making it a key component of sustainable operations (Dacre et al., 2025). WC, on the other hand, reflects the human dimension of sustainability, including employees' ability to adapt to digital transformation, maintain safety standards and contribute to organisational resilience. Human-centric work environments, continuous learning and skill development are important factors supporting long-term sustainability performance (Pasparakis et al., 2023; Nasir et al., 2025; Tian et al., 2025).

In line with the RBV, RE and WC can be considered valuable organisational resources that create sustainable competitive advantages when they are effectively transformed into performance outcomes. From this perspective, both constructs are expected to contribute to SLSCP. Accordingly, both RE and WC are expected to be associated with SLSCP.

OE is an important factor in achieving sustainable supply chain performance, as it directly influences process reliability, resource utilisation and environmental outcomes. Advanced technologies such as AI-based decision systems, demand forecasting tools and automated logistics operations enhance both agility and efficiency in supply chain processes (Lin, 2025). Higher levels of OE are associated not only with cost and time advantages but also with improvements in energy efficiency, process flexibility and environmental compliance (Bag et al., 2020; Loske and Klumpp, 2021; Rainer et al., 2025). In this sense, OE plays a role in supporting sustainability both directly and indirectly.

From both the socio-technical systems and dynamic capabilities perspectives, OE represents the effective alignment of technological resources, organisational processes and human capabilities, thereby facilitating improvements in sustainable supply chain performance. Therefore, OE is expected to positively influence SLSCP.

The conceptual model developed in line with these hypotheses presents the direct and indirect effects of HMC, OE, RE and WC variables on SLSCP within the Industry 5.0 framework using a holistic approach. The model assumes that HMC is a central element shaping sustainable performance through both operational and RE and WC. In this context, the proposed structural framework is as shown in Figure 1.

This section presents the methodological framework of the research. The background of the study, research design, questionnaire development process, sample characteristics and data collection stages are explained in detail. The research is structured to examine the relationships between HMC, operational and RE, and WC within the context of Industry 5.0 using a quantitative approach. The questionnaire developed for this purpose was adapted from scales found in the relevant literature; its content validity was ensured through expert opinions and it was finalised after a pilot application.

The research model was developed based on the previously proposed conceptual framework and empirical findings in the literature (Nazarian and Khan, 2024; Jamil et al., 2024; Dwivedi et al., 2023). The scale items were designed to cover the themes of Industry 5.0 applications, HMC, digitalisation, sustainability and logistics performance. The content of the questionnaire has been adapted from scales and models found in the literature (Waheed et al., 2022; Guo et al., 2023; Jamil et al., 2024; Locatelli et al., 2024; Nazarian and Khan, 2024). The survey consists of two main sections. The first section aims to determine the demographic characteristics of the participants, such as age, gender, education level, professional background and area of responsibility. The second section measures the participants' perceptions of digital transformation, OE and sustainability. A 5-point Likert scale (1 = Strongly Disagree, 5 = Strongly Agree) was used in this section. The questionnaire was evaluated for content validity by two academics and one industry expert; language adaptations were made to the statements in line with their recommendations. Based on feedback from 18 participants in the pilot study, some statements were revised and simplified.

The fieldwork for the research was conducted in Italy, and the participants consisted of professionals and academics working in the fields of logistics, SCM, engineering, technology and academia. Italy was selected as the empirical setting as it is one of the countries where manufacturing and logistics activities are prominent and where Industry 4.0 and emerging Industry 5.0-related practices can be observed. The inclusion of both academics and industry professionals was intended to capture both theoretical knowledge and practical experience, allowing a more comprehensive evaluation of the research variables. As the study examines perceptual constructs related to Industry 5.0 adoption, sustainability and organisational capabilities, both groups were considered appropriate respondents for evaluating the proposed relationships. Academics contributed theoretically informed evaluations, whereas industry professionals provided practice-based insights derived from their operational experience.

The data collection process was carried out between April and August 2025. The questionnaire was prepared and distributed online via Google Forms and shared with academic circles, research networks and industry professionals working in relevant sectors. Participation was entirely voluntary, and the confidentiality and anonymity of responses were ensured in accordance with ethical guidelines.

This section presents the findings derived from the questionnaire data. First, the demographic characteristics of the participants are summarised, followed by the results of the measurement model and structural model analyses. The analyses were conducted using SmartPLS 4 software; confirmatory factor analysis and structural equation modelling (SEM) were employed. SEM is an integrated analysis technique that allows the simultaneous testing of complex relationships between observed and latent variables (Hair et al., 2019; Kline, 2023). This method enables the evaluation of both the measurement model (validity and reliability) and the structural model (testing of hypotheses), thereby ensuring the comprehensive validation of the research model. The SEM was tested in two stages. In the first stage, the validity and reliability of the measurement model were examined; in the second stage, the relationships in the structural model were tested using the Bootstrapping method (5,000 repetitions). Factor loadings above 0.70 were considered sufficient for convergent validity (Hair et al., 2019). Furthermore, composite reliability (CR ≥ 0.70) and average variance extracted (AVE ≥0.50) values meeting the threshold values indicate that the model is appropriate in terms of reliability and convergent validity (Hair et al., 2022). Discriminant validity was verified using the Fornell–Larcker criterion (Fornell and Larcker, 1981).

A total of 118 valid responses were included in the analysis. This sample size is considered adequate for PLS-SEM analysis. According to the “10-times rule”, the minimum sample size should be at least ten times the maximum number of structural paths directed at any latent construct (Hair, 2014; Kock and Hadaya, 2018). In this study, the most complex construct (SLSCP) receives three direct paths, indicating a minimum requirement of 30 observations. Therefore, the sample size of 118 is well above this threshold. In addition, sample sizes above 100 are generally regarded as acceptable for models with moderate complexity (Hair et al., 2019). The participants' demographic characteristics are summarised in Table 2.

Table 2 shows that the participant group forming the sample of the study exhibited a balanced distribution in terms of age, gender, education and occupational diversity. Most participants are aged 31–50 (61%), and the gender distribution is balanced, with 45.8% female and 50.0% male. In terms of educational level, 68% of the sample are postgraduate (master's or doctoral) graduates. 71.1% of participants demonstrated moderate or advanced familiarity with Industry 5.0 concepts, indicating a high level of awareness of the subject among the sample.

The measurement model was evaluated based on convergent and discriminant validity, internal consistency and multicollinearity criteria (Cheung et al., 2024; Habibi et al., 2025). The general structure of the model is presented in Figure 2, and the factor loadings of the indicators related to the measurement model are presented in Table 3.

The factor loadings of the indicators show that all variables are above the acceptable threshold value (0.70) (Henseler et al., 2015; Purwanto and Sudargini, 2021). The loadings of the indicators for all constructs range from 0.75 to 0.92, confirming that each indicator strongly represents its respective construct. Therefore, it was concluded that the indicator reliability of the model was achieved. Subsequently, Cronbach's Alpha, rho_A, CR and average variance extracted (AVE) values were analysed to assess the internal consistency and convergent validity of the constructs included in the model. These indicators are fundamental criteria widely used in the literature to evaluate the internal consistency of each construct and the overall reliability of the measurement model (Hair et al., 2022; Pehlivan et al., 2024). The results obtained show that the constructs meet the reliability and validity criteria. The relevant values are presented in Table 4.

Upon examining Table 5, findings indicate that the measurement model is reliable and valid, thus allowing us to proceed to the analysis of the structural model. Another important step in assessing the validity of the model is to test whether the constructs are statistically distinguishable from each other. The Fornell–Larcker criterion was used for this purpose. This criterion states that the square root of each construct's AVE value must be higher than its correlations with other constructs (Fornell and Larcker, 1981; Hair et al., 2019). This determines whether the variables in the model represent different concepts without the risk of multicollinearity. The results of the Fornell–Larcker analysis are summarised in Table 5.

The Fornell–Larcker criterion results shown in Table 5 indicate that each construct's own AVE square root value (values on the diagonal) is higher than its correlation coefficients with other constructs. This finding confirms that the constructs in the model are meaningfully distinct from one another and that discriminant validity is achieved. To test whether multicollinearity existed among the variables in the measurement model, the variance inflation factor (VIF) values were examined. A VIF value below 5 indicates that there is no excessive correlation among the indicators and that the model produces stable estimates. The VIF values of the structural model created within the scope of the study are shown in Table 6.

Table 6 shows that all average VIF values are below the recommended threshold of 5 (ranging from 2.48 to 3.89), indicating that multicollinearity is not a concern in the measurement model. This result confirms that the indicators represent their respective constructs independently and that the estimated regression coefficients are stable and reliable. Together with the satisfactory reliability (Cronbach's alpha and CR), convergent validity (AVE) and discriminant validity (Fornell–Larcker criterion) results, these findings provide additional support for the robustness of the measurement model.

The overall fit level of the structural model developed in this study was assessed using the SRMR (Standardised Root Mean Square Residual), d_ULS (Unweighted Least Squares Discrepancy), d_G (Geodesic Discrepancy), Chi-square and NFI (Normed Fit Index) measures. These indicators are commonly used to assess the fit between the model's predicted structural relationships and the data (Hair et al., 2019; Rahman et al., 2024). The findings indicate that the model fits the data well and that the structural relationships are statistically supported. The relevant model fit indices are presented in Table 7.

Upon examining Table 7, the SRMR value of 0.078, which is below the 0.08 threshold, indicates that the model has a good fit. Similarly, the d_ULS (1.421) and d_G (1.289) values are below 3 and within the recommended ranges. The NFI = 0.729 value is also acceptable. These findings confirm that the model's overall fit is within the limits specified in the literature and that it is suitable for transitioning from the measurement model to the structural model.

After the measurement model met the validity and reliability criteria, the causal relationships between variables in the structural model were tested. The model explains the effect of HMC in the context of Industry 5.0 on SLSCP through OE, RE and WC. The model includes four endogenous variables (OE, RE, WC, SLSCP) and one exogenous variable (HMC). The hypothesis test results of the structural model are shown in Figure 3 and Table 8.

Upon examining the results, the model's overall explanatory power is R² = 0.70, indicating that the model has a high level of explanatory power. The R² values for the endogenous variables (OE = 0.636, RE = 0.465, WC = 0.826, SLSCP = 0.880) support the model's statistically robust and theoretically consistent structure. According to the findings, HMC has positive and significant effects on both OE and RE as well as WC. This result demonstrates that HMC not only increases process efficiency but also improves resource utilisation and employee competency levels. These effects of HMC are consistent with Industry 5.0's human-centric and learning-focused production vision; therefore, the adoption of digital technologies in businesses also strengthens knowledge sharing, flexibility and adaptability.

OE plays a critical role as an intermediate variable in the model. OE has significant effects on RE, WC and SLSCP, indicating that OE directly and indirectly affects both resource management and sustainable performance. This finding supports that OE yields sustainable results not only at the production level but also across the entire supply chain.

RE showed the strongest direct effect on SLSCP in the model (β = 0.738, p < 0.001). This result clearly demonstrates that RE is the most decisive factor in SLSCP. Optimising resource use and reducing waste directly aligns with Industry 5.0's environmental sustainability goals. In contrast, the effect of WC on SLSCP was not found to be significant (β = 0.085, p = 0.402); this indicates that workforce capabilities do not directly contribute to sustainable performance on their own, but create indirect effects through other structures.

When examining indirect effects, the chained mediation HMC → OE → RE → SLSCP emerged as the strongest indirect path in the model. This result indicates that the impact of HMC on sustainable performance is largely mediated through operational and RE mechanisms. Similarly, the significance of the HMC → OE → RE and OE → RE → SLSCP paths reinforces the mediating role of OE in the model. Furthermore, the significance of the HMC → OE → WC path reveals that human–machine interaction increases labour capacity through factors such as learning, standardisation and occupational safety.

When assessed at the total effect level, the total effect of HMC on SLSCP is 0.692 and the total effect of OE is 0.530, with most of these effects occurring through indirect pathways. This indicates that HMC contributes significantly to sustainable performance not through direct contributions but through efficiency-based indirect effect chains.

Overall, these findings confirm that HMC within the Industry 5.0 framework has a multi-layered impact on SLSCP. OE and RE, in particular, have emerged as key intermediary mechanisms in conveying this effect. The model's explanatory power and statistically significant relationships support the relevance of examining Industry 5.0's sustainability and human-centric objectives in an integrated manner.

The findings of this study indicate that HMC contributes to SLSCP primarily through its influence on operational processes rather than through a direct effect. This suggests that HMC functions as an enabling mechanism that improves coordination, process quality and system reliability. Such an interpretation is consistent with prior research highlighting the role of human–machine interaction in enhancing efficiency, safety and decision accuracy in logistics and production contexts (Loske and Klumpp, 2021; Stecke and Mokhtarzadeh, 2022; Pasparakis et al., 2023; Zia and Haleem, 2025). OE plays a central role in translating these improvements into sustainability outcomes. The results show that its contribution becomes meaningful when it leads to more effective resource utilisation. This supports the view that digitalisation and process optimisation enhance sustainability through their impact on resource management practices rather than acting as standalone drivers (Ghobakhloo et al., 2023; Hsu et al., 2024; Lin, 2025). This finding suggests that OE alone may not fully explain sustainability outcomes, differing from studies that consider efficiency a direct driver of sustainable performance.

Among all variables, RE emerges as the most influential determinant of sustainable performance. This highlights that sustainability in supply chains is closely linked to the ability to reduce waste, improve energy use and adopt circular practices. Similar conclusions have been reported in studies emphasising the importance of resource-oriented approaches in achieving environmental and economic outcomes (Jayarathna et al., 2023; Kharayat and Gupta, 2025; Dacre et al., 2025). In this respect, the findings reinforce the argument that resource management practices constitute the core mechanism through which sustainability objectives are achieved. In contrast, WC does not show a statistically significant direct effect on sustainable performance. While this may appear counterintuitive, it provides an important insight into the human-centric dimension of Industry 5.0. Previous studies often report a positive association between WC and performance outcomes (Neumann et al., 2021; Płaza et al., 2025; Yasari et al., 2025); however, the present findings suggest that its contribution may be more indirect. These findings indicate that WC supports SLSCP through its integration with operational and technological capabilities rather than through a direct effect. In this model, human competencies contribute by facilitating technology adoption and strengthening operational routines, suggesting that their impact on sustainability is realised indirectly.

Finally, the significance of indirect pathways, particularly those linking HMC to performance through operational and RE, underlines the interconnected nature of sustainability outcomes. These findings indicate that sustainable performance emerges from the interaction of multiple organisational and technological mechanisms rather than isolated factors. In this context, Industry 5.0 can be understood as a system-level transformation in which human, operational and resource-based elements operate in combination to shape SLSCP. The results also reflect the broader principles of Industry 5.0, particularly sustainability, resilience and human-centricity, suggesting that sustainable supply chain performance depends on the effective integration of technological, operational and human capabilities rather than on technological adoption alone.

From a theoretical perspective, this study contributes by clarifying the structural relationships between key Industry 5.0 variables. The findings show that the HMC → OE path is strong and significant, indicating that HMC primarily operates through operational processes. The OE → RE → SLSCP pathway suggests that sustainability outcomes are more closely associated with RE than with direct technological effects alone. In addition, the non-significant WC → SLSCP relationship suggests that WC contributes indirectly, rather than acting as a direct performance driver.

From a managerial perspective, these results imply that firms should prioritise improving OE as a bridge between digital technologies and sustainability outcomes. In practice, this means focusing on RE (e.g. energy use, waste reduction) and aligning digital investments with process-level improvements. For example, firms may prioritise investments in energy monitoring systems, waste tracking tools and AI-supported process optimisation to improve RE. The findings also indicate that workforce development should be designed to support operational and technological integration rather than being treated as an isolated objective. Managers should therefore adopt an integrated approach that aligns digital technologies, process improvement and employee development to support long-term sustainability objectives.

At the policy level, the results highlight the importance of supporting integrated approaches that link technology adoption with RE and skills development. For instance, policymakers may introduce targeted incentives for SME digitalisation, establish industry–academia collaboration platforms and define measurable sustainability standards to guide firms' transition processes.

This study has several limitations that should be considered when interpreting the findings. First, the analysis is based on a relatively small sample (n = 118) and is limited to participants in Italy, which may restrict the generalisability of the results across different contexts. Second, the use of a mixed sample including both academics and industry professionals may introduce heterogeneity in responses. Future research may examine these groups separately to determine whether theoretical and practice-based perspectives differ in evaluating Industry 5.0 implementation and sustainable supply chain performance.

In addition, as the data were collected using a single survey method, the possibility of common method bias cannot be fully excluded. Future research may address this limitation by using multiple data sources or longitudinal designs. Further studies are encouraged to test the proposed model in different countries and sectors, and to incorporate additional variables such as organisational maturity, ESG capabilities, circular economy practices, digital maturity, organisational culture, resilience and governance mechanisms. In particular, longitudinal approaches may provide deeper insights into the indirect and time-dependent effects of WC and HMC on sustainable supply chain performance.