Modern manufacturing requires more than traditional management strategies, specifically it needs manufacturing data analytics (MDA). However, a recent study finds that adoption of MDA is still low with several barriers to implementation.
Imagine the hum of a clean, streamlined factory with not a single bolt or movement happening out of place. Where manufacturing in the first industrial revolution was chaotic and dangerous, the ideal of production today is flexible, customized to the latest need, and efficient beyond measure. Or that’s the dream, anyway. Modern manufacturing does require more than traditional management strategies, specifically manufacturing data analytics (MDA), but one study finds that adoption of MDA is still low. It turns out that there are still barriers to implementation.
Why aren’t more manufacturers adopting MDA?
Bringing the ordered world of data into a messy human process like manufacturing requires a shift in thinking and scaffolding as people and processes adapt. MDA implementation typically involves a few different steps (varying in number depending on which process you follow) with all the usual suspects there: preparing data, defining the question, gathering all data sources, collecting and cleaning data, analyzing it, visualizing it, and implementing it into larger processes. In each step, there’s an opportunity to go wrong.
Even more, broader issues within the technological, organizational, and environmental context have been investigated, but a recent study from Pusan National University published in the Journal of Manufacturing Systems has taken on the challenge of compiling a comprehensive issue set for MDA implementation (CISM). “Comprehensive issue identification for manufacturing data analytics implementation” hopes to present the worlds first CISM for MDA that addresses the full context of implementation.
Technology is a barrier
Many of the issues that emerged during the literature study revolved around technology. Investing in the right ecosystem of tools that are both future-proof and user-friendly, without straining tech budgets, is a significant barrier. These implementations must be accurate and timely, as well as cover the entire data lifecycle.
But manufacturers often face a confusing mix of legacy systems, partial upgrades, and pressure to adopt cutting-edge solutions that may not integrate well with existing workflows. Investing in an ecosystem that’s both interoperable and scalable is a tall order, especially for mid-sized manufacturers without large IT budgets.
This means that while technology is often pitched as the enabler of data-driven manufacturing, it’s also one of its biggest stumbling blocks.
More so than tech, talent is in short supply
Even with tools, a lack of broad understanding of data analytics within context is also a barrier. Data goes unexamined, or siloed systems don’t provide the level of comprehensive support these organizations need. Current teams may be unfamiliar with data extraction methods or miss hidden insights from the data they do collect. On the other side, teams well-versed in data may not understand the complexity of manufacturing.
Even bigger, implementing MDA requires a shift within the entire organization, not just a technological upgrade. It transforms business processes and creates cultural shifts that may be uncomfortable or unwelcome at first. Without the right buy-in and communication between data experts and domain experts, these projects might be short-lived.
Ultimately, MDA isn’t a plug-and-play solution. It requires cross-functional fluency, trust between teams, and a willingness to learn new ways of working. Without that, even the most powerful analytics engine may well stall out before it ever delivers the value that stakeholders are hoping for.
See also: 2025 IoT Developments to Spur Industrial Innovations
How a better MDA framework can help implementation
The team read 35 papers on MDA implementation after a thorough literature review and selection. Challenges were extracted from these papers with careful focus on TOE contexts. The team categorized recorded issues based on similarity and sorted them into five different categories: data analysis planning, data preparation, data analysis, evaluation and interpretation, and implementation into manufacturing systems.
This comprehensive reading mattered because each of these papers focused on a specific subset of MDA implementation, leaving the broad contextual view lacking. With this study, the team was able to create a comprehensive set of challenges mapped to implementation stages, setting the stage for a better understanding of the process as a whole.
Organizations with foreknowledge of potential bottlenecks and roadblocks can set up more suitable implementation parameters. They can address issues proactively, which may lead to greater implementation overall. The entire point is to create better avenues for success for this process, which has such high potential.
Why is this paper different from previous studies?
Previous papers tackled only a narrow piece of MDA implementation. The team here was able to tease out 29 different parameters, with potential for problems. This reshapes the entire conversation surrounding MDA by using a full field view.
Manufacturers aren’t failing to implement MDA because they don’t want to. They’re failing because they can’t see what’s ahead. A clearer map like this one helps manufacturers see possible roadblocks and pivot when they come up, or just prevent them entirely. This includes smarter planning, more targeted training, and a cultural shift that isn’t left to chance. If manufacturing is to evolve into the adaptive, efficient, data-driven system so many envision, then understanding these friction points is foundational to the entire process. This paper gives manufacturers a better shot at making the MDA promise a reality.
