Modern manufacturing is increasingly entrenched in complex and volatile environments, where traditional management methodologies have become insufficient. The challenges faced today demand far more agile, adaptive, and intelligent strategies that can respond in real-time to changing variables. As a result, manufacturers are turning their attention to digital solutions such as Manufacturing Data Analytics (MDA), which is rapidly gaining recognition as a transformative enabler of smart manufacturing. By harnessing the power of data, MDA enables companies to identify subtle patterns and trends within their internal operations and across external market landscapes. These insights help firms anticipate and address supply chain disruptions, geopolitical risks, and evolving consumer expectations with greater precision and speed.
Despite its potential, the uptake of MDA in manufacturing remains surprisingly low. Fewer than 20% of MDA-related projects ever reach full implementation. This slow rate of adoption is not due to a lack of interest, but rather the presence of multiple implementation challenges that arise at various stages of the MDA process. Implementing MDA involves a sequence of five interrelated stages: preparing data, analysing it, evaluating the outcomes, interpreting results, and incorporating these insights into production systems. Each stage presents its own set of technical difficulties, compounded by organisational inertia and environmental constraints. Although previous research has explored many of these obstacles, most studies are limited in scope, often focusing on just one or two phases and rarely integrating the broader technological, organisational, and environmental (TOE) dimensions that are essential to understanding MDA’s real-world challenges.
To fill this knowledge gap, a team of researchers from the Department of Industrial Engineering at Pusan National University in South Korea has developed a new framework designed to systematically capture the full spectrum of challenges that hinder MDA implementation. Led by Assistant Professor Ki-Hun Kim, with contributions from Mr. Sa-Eun Park and Mr. Sang-Jae Lee, the team introduced the Comprehensive Issue Set for MDA Implementation (CISM). “To speed up the adoption of MDA, manufacturers need to be able to proactively recognise and resolve the various technical, organisational, and environmental barriers,” explains Dr Kim. “CISM provides the structured lens through which these challenges can be understood, prioritised, and ultimately addressed.” Their research was published in Volume 82 of the Journal of Manufacturing Systems in October 2025, following its online release on 9 June 2025.
To create this framework, the researchers performed a systematic review of existing academic literature. Using the SCOPUS database, they identified 35 studies that discussed barriers to the implementation of MDA. Through detailed analysis and synthesis, they were able to identify 29 distinct issues and categorise them into nine thematic clusters, each corresponding to a specific TOE domain and stage in the MDA process. Among these, 26 issues were aligned with the technological context, 11 with organisational structures and practices, and 4 with environmental factors such as regulatory or market dynamics. The nine categories capture a wide range of concerns—from data accessibility and compatibility, to the communication gap between data scientists and domain experts, and the difficulty of adapting analytics models to reflect the realities of manufacturing operations.
To assess the framework’s applicability in real-world settings, the research team conducted three case studies within the rubber manufacturing sector. These studies focused on optimising complex production steps, including the formulation of rubber recipes and the consistency of the mixing process. By applying CISM to these scenarios, the researchers confirmed its practical value: the framework was able to capture and explain all the significant challenges that emerged during the projects. This not only validated the comprehensiveness of CISM but also demonstrated how it could serve as a diagnostic tool to guide manufacturers through the process of adopting MDA in a structured and informed manner.
The authors also suggest several avenues for future research, including ranking the importance of each identified issue and investigating their relevance across different manufacturing contexts, such as discrete manufacturing, continuous processing, or hybrid systems. Furthermore, the development of customised mitigation strategies tailored to specific industries or enterprise sizes could significantly enhance the effectiveness of CISM. Educational and training initiatives based on this framework could also play a crucial role in preparing the workforce to engage with MDA more effectively, thereby reducing the human and organisational resistance that often accompanies digital transformation efforts.
In sum, CISM represents a meaningful advancement in the pursuit of data-driven smart manufacturing. By providing a comprehensive, structured understanding of the challenges associated with MDA implementation, it equips both researchers and practitioners with the knowledge needed to overcome existing barriers. As industries worldwide grapple with the imperative to innovate and adapt in the face of rapid technological and geopolitical change, tools like CISM will be vital. Not only do they help ensure that MDA projects succeed, but they also lay the groundwork for a more responsive, resilient, and intelligent manufacturing ecosystem—one that is better equipped to serve both industry needs and consumer expectations in the digital age.
More information: Sa-Eun Park et al, Comprehensive issue identification for manufacturing data analytics implementation: Systematic literature review and case studies, Journal of Manufacturing Systems. DOI: 10.1016/j.jmsy.2025.05.006
Journal information: Journal of Manufacturing Systems Provided by Pusan National University