Data Analytics Framework Boosts Smart Manufacturing Potential

Researchers at Pusan National University have identified 29 distinct issues hindering manufacturing data analytics (MDA) adoption, detailed in a comprehensive framework published in the October 2025 issue of the Journal of Manufacturing Systems. Led by Assistant Professor Ki-Hun Kim, the team systematically reviewed 35 prior studies and validated their findings through case studies in the rubber manufacturing industry. This research addresses a critical gap, as fewer than one in five MDA projects currently reach full implementation, costing manufacturers potential efficiency gains and hindering the transition to smart manufacturing practices.

Smart Manufacturing and Data Analytics

Despite the demonstrable benefits of manufacturing data analytics (MDA) in transitioning to smart manufacturing, adoption rates remain low, with fewer than one in five projects achieving full implementation. Researchers at Pusan National University have addressed this challenge by developing a comprehensive issue set for MDA implementation (CISM), designed to proactively identify and resolve obstacles encountered during the process. CISM recognises that successful MDA deployment requires consideration of not only technical hurdles, but also organizational and environmental factors – a holistic approach often absent in prior studies.

The research team systematically reviewed 35 relevant papers identified through the SCOPUS database, culminating in a framework encompassing 29 distinct issues categorised into nine groups. These issues are mapped to both the five core steps of the MDA process – data preparation, analysis, evaluation, interpretation, and implementation – and the broader technological, organizational, and environmental (TOE) contexts influencing success. Specifically, 26 issues relate to technological context, 11 to organizational context, and 4 to environmental context.

To validate CISM’s efficacy, the team applied it to three case studies within the rubber manufacturing industry, focusing on optimising recipe formulation and mixing processes. The framework successfully captured all implementation challenges encountered during these projects, demonstrating its practical applicability and comprehensiveness. The authors propose future research directions including ranking issue criticality, exploring context-specific relevance across different manufacturing sectors, and developing targeted mitigation strategies.

CISM offers manufacturers a structured methodology for identifying and prioritising issues hindering effective MDA implementation. Furthermore, it provides a valuable resource for developing educational materials and training programmes, ultimately facilitating the delivery of higher-quality products with increased efficiency and reliability. The framework represents a significant step towards wider adoption of data-driven smart manufacturing practices.

The Challenge of MDA Implementation

The granular detail within CISM addresses a critical gap in existing literature, which often overlooks the complex interplay between technical challenges and broader contextual factors. The framework’s mapping of identified issues to both the stages of the MDA process and the TOE contexts – technological, organizational, and environmental – provides a nuanced understanding of potential roadblocks. This allows manufacturers to move beyond simply identifying problems to understanding where in the implementation process – and why – those problems are likely to arise.

The categorization of issues—spanning data quality concerns, knowledge gaps between data scientists and domain experts, and alignment of analytical models with real-world manufacturing systems—highlights the multifaceted nature of successful MDA deployment. The research team’s identification of 26 technologically-rooted issues underscores the importance of robust data infrastructure and analytical capabilities. However, the 11 organizational issues—encompassing topics such as change management and inter-departmental collaboration—demonstrate that technical proficiency alone is insufficient. Similarly, the four environmentally-focused issues acknowledge the influence of external factors, such as regulatory compliance and market volatility, on MDA initiatives.

Validation through case studies in rubber manufacturing provided practical confirmation of CISM’s utility. By successfully capturing all implementation challenges encountered during optimisation of recipe formulation and mixing processes, the framework demonstrated its capacity to function as a comprehensive diagnostic tool. This practical validation is particularly significant, as it moves beyond theoretical frameworks to offer a tangible resource for manufacturers seeking to improve their MDA implementation strategies.

The authors’ proposed avenues for future research – specifically, issue prioritisation and context-specific relevance – represent logical next steps in refining CISM’s applicability. Determining the relative importance of each identified issue will allow manufacturers to focus their resources on the most critical areas. Furthermore, exploring how these issues manifest differently across various manufacturing sectors will enhance the framework’s adaptability and broaden its potential impact.

Introducing the Comprehensive Issue Set

The comprehensiveness of CISM extends beyond simply identifying potential obstacles; it facilitates a proactive approach to risk mitigation. By mapping each of the 29 identified issues to both the stage of the MDA process – from initial data preparation through to implementation – and the broader technological, organizational, and environmental (TOE) contexts, manufacturers gain a nuanced understanding of where and why specific challenges are likely to arise. This granular detail enables targeted interventions and resource allocation, moving beyond reactive problem-solving to preventative strategies.

The framework’s utility is further enhanced by its explicit recognition of non-technical factors. While 26 of the identified issues relate to technological context – encompassing data quality, infrastructure limitations, and analytical model selection – the remaining 11 organizational issues and 4 environmental issues underscore the critical importance of change management, inter-departmental collaboration, regulatory compliance, and market volatility. This holistic perspective acknowledges that successful MDA implementation requires not only technical proficiency but also a supportive organizational culture and an awareness of external forces.

Furthermore, the validation of CISM through case studies in the rubber manufacturing industry provides practical confirmation of its applicability in real-world settings. By successfully capturing all implementation challenges encountered during optimisation of recipe formulation and mixing processes, the framework demonstrated its capacity to function as a comprehensive diagnostic tool, capable of identifying potential roadblocks before they escalate into significant problems. This practical validation is particularly significant, as it moves beyond theoretical frameworks to offer a tangible resource for manufacturers seeking to improve their MDA implementation strategies.

Validating CISM Through Case Studies

The validation of CISM through real-world case studies underscores its practical utility beyond theoretical construction. Application to three projects within the rubber manufacturing industry – focused on optimising recipe formulation and mixing processes – demonstrated the framework’s ability to comprehensively capture all encountered implementation challenges. This confirms CISM’s capacity to function as a diagnostic tool, identifying potential obstacles before they escalate, and is particularly valuable for manufacturers seeking to enhance their strategies for deploying manufacturing data analytics.

The authors’ proposals for future research – specifically, prioritising issue criticality and exploring context-specific relevance – represent logical steps towards refining CISM’s applicability. Determining the relative importance of each identified issue will enable manufacturers to focus resources on the most impactful areas. Furthermore, investigating how these issues manifest differently across various manufacturing sectors will broaden the framework’s adaptability and potential impact, facilitating wider adoption of effective manufacturing data analytics practices.

Future Research and Practical Applications

Beyond its diagnostic capabilities, CISM provides a foundation for targeted educational initiatives. Manufacturers can leverage the framework to establish clear guidelines for identifying and proactively addressing issues hindering effective MDA implementation. This extends beyond internal training; CISM can inform the development of standardised educational resources, fostering a more skilled workforce capable of successfully deploying and managing data analytics solutions. Such initiatives are crucial for realising the full potential of smart manufacturing and ensuring a consistent level of competence across the industry.

The identified issues within CISM are not static; their relative importance is likely to evolve alongside technological advancements and changing market dynamics. Future research should therefore focus on developing a dynamic weighting system for each issue, reflecting its current criticality and potential impact. This would enable manufacturers to prioritise their efforts effectively, allocating resources to the most pressing challenges and maximising the return on their investment in MDA. Furthermore, exploring the interplay between different issues – identifying potential cascading effects or synergistic relationships – could provide valuable insights for developing more robust and resilient implementation strategies.

Expanding the validation of CISM beyond the rubber manufacturing industry is also crucial. While the initial case studies provide valuable confirmation of the framework’s utility, its applicability across diverse manufacturing contexts – such as automotive, aerospace, or pharmaceuticals – remains to be fully established. Conducting further case studies in these sectors would not only broaden the scope of validation but also reveal potential sector-specific nuances and challenges that require tailored mitigation strategies. This iterative process of validation and refinement will ensure that CISM remains a relevant and effective tool for manufacturers seeking to embrace the benefits of data-driven smart manufacturing.

Finally, the integration of CISM with existing quality management systems – such as Six Sigma or Lean Manufacturing – offers a promising avenue for future research. By incorporating data analytics-related issues into established quality control processes, manufacturers can proactively identify and address potential problems before they impact production efficiency or product quality. This synergistic approach would not only enhance the effectiveness of MDA implementation but also foster a culture of continuous improvement, driving innovation and competitiveness across the entire organisation.