The Future of PLM Starts with Data, Not Tools
Talk to manufacturing leaders about the future of PLM, and the conversation quickly turns to AI. Ask what stands in the way of real progress, and a different answer often emerges: disconnected systems, inconsistent data, and processes that were never designed to support today’s requirements.
For Dr. Steffen Kunnen, Head of Product for PRO.FILE, that’s where the discussion should start. New technologies matter, but only if the underlying data can support them. In the face of mounting pressure to improve efficiency, manufacturers must balance AI initiatives and compliance requirements. A strong product data foundation is now a prerequisite for long-term success.
In this interview, Kunnen shares his perspective on the trends shaping PLM and manufacturing, the growing importance of product data quality, and why the future of PLM begins with data instead of tools.
Which technology trends do you believe will have the greatest impact on PLM and the manufacturing industry during the second half of 2026? Where do companies need to think long-term, and what practical steps should they take today to prepare?
AI remains the defining topic. The question isn’t whether manufacturers are “using AI in some way,” but whether they can identify use cases that deliver measurable value. In PLM and engineering environments, I currently see the greatest potential for generative AI in information retrieval, input completion, and the automation of clearly defined processes.
In the near term, companies need to focus on their data foundation and prioritize a limited number of workflows, while bringing users into the process early. Approval workflows, BOM transfers, and technical documentation are areas where this approach can be applied.
At the same time, AI is only as effective as the information it can access. If information isn’t structured and consistently available, results quickly become unreliable. That’s exactly where PLM becomes relevant.
The latest PRO.FILE Smart Manufacturing Report found that 29% of manufacturers consider a Product Data Backbone the most important investment for future readiness. Why is a strong data foundation more important than adopting individual technologies?
New technologies are only as good as the information behind them. The pace of change is enormous. What looks innovative today may be standard tomorrow. If the underlying data foundation isn’t there, the value of any new tool is short-lived. Instead of driving progress, implementation projects often create new problems.
You can see this in many PLM initiatives. They rarely fail because functionality is missing. More often, they struggle because data quality isn’t where it needs to be or because information can’t be integrated into existing systems. At the same time, the most important use cases rely on dependable master data. A central Product Data Backbone, like the one provided by PRO.FILE, is what makes reliable digital processes possible in the first place.
EU regulations such as CSRD and CBAM are dramatically increasing requirements for traceability and data continuity. How much are these regulations influencing PLM strategies among mid-sized manufacturers? What role does the digital thread play?
More than many manufacturers realize. However, the EU isn’t mandating specific software platforms. These regulations raise expectations for data consistency, traceability, and auditability. That doesn’t just affect companies with direct reporting obligations. In many cases, those requirements are pushed down the supply chain, for example, when OEMs require emissions data, material declarations, or traceability information from their suppliers.
To me, this goes hand in hand with the Digital Product Passport. It requires material, supplier, and lifecycle information to be available in a digital, standardized, and machine-readable format. Manufacturers that don’t have this information readily available often end up relying on manual processes and spreadsheets, which creates a significant maintenance burden.
That’s why the digital thread is becoming increasingly important. It provides the foundation needed for compliance and traceability, while ensuring that the right information is available when it’s needed.
The Corporate Sustainability Reporting Directive (CSRD) is an EU regulation that requires companies to disclose detailed sustainability and ESG information. It increases expectations around data transparency, traceability, and auditability across the organization.
The Carbon Border Adjustment Mechanism (CBAM) requires reporting of carbon emissions associated with certain imported goods. It places greater emphasis on emissions data, supply chain visibility, and standardized reporting.
Engineering Change Management is often considered an Achilles’ heel in manufacturing. Changes are still managed through email, spreadsheets, or informal conversations. Why are so few companies taking a more systematic approach?
Engineering Change Management isn’t really a tool issue. It affects how a company works and involves multiple departments. Once you start formalizing the process, responsibilities become visible, familiar shortcuts disappear, and existing organizational resistance comes to the surface. Many manufacturers know the current approach isn’t ideal, but they’re reluctant to confront those issues directly.
There’s also the risk of over-customization. When ECM processes are heavily tailored around special requirements, complexity increases, and user acceptance tends to drop. At the same time, many companies underestimate how quickly modern ECM solutions can be implemented today. As a result, email, spreadsheets, and informal coordination often remain the most convenient option—even if they’re also the most error-prone.
Many organizations still view data security and data connectivity as competing priorities rather than complementary ones. How can companies balance the need for strict IP protection and compliance with the demand for more connected data and systems?
Data sovereignty is a fundamental requirement. That’s understandable. Engineering data, product knowledge, and intellectual property sit at the core of a manufacturer’s business. At the same time, isolation isn’t a viable strategy. Organizations that lock down their systems too aggressively limit the flow of information that modern processes and technologies depend on.
The answer lies in clearly defined access models and an architecture that supports both data sovereignty and data continuity. That’s why hybrid approaches will continue to gain importance. They allow sensitive information to remain where companies need full control, while enabling collaboration through web-based or cloud-connected components where it makes sense.
As AI becomes part of the equation, the challenge becomes even more important. The more open and capable a system is, the greater the need for verification, traceability, and a zero-trust mindset. In the end, connectivity only works when security is built into the system design from the start.
Many mid-sized manufacturers start with a PDM system and eventually reach its limits. What does it take to successfully move from product data management to a scalable PLM environment, and what principles or milestones matter most along the way?
I think too many discussions start with the end solution. When a PDM system reaches its limits, the first question shouldn’t be, “What PLM system do we need now?” The better question is, “Where are we running into problems, and which issue do we need to solve first?”
My recommendation is to take a structured approach. Start by identifying the challenges, then look for gaps in the existing data and process model. From there, define the requirements, strengthen the underlying data foundation, and build the PLM strategy around it.
Scalable PLM doesn’t come from implementing more functionality. It comes from creating a foundation that can evolve over time. If the data model is sound and new capabilities are added based on specific use cases, the system becomes far more resilient. That’s where the Product Data Backbone approach comes in. It provides the structure needed to support meaningful growth and expansion.
Cloud software is often seen as the only viable path forward. Your research shows that 44% of manufacturers already use cloud-based PLM solutions. Is a cloud-first approach truly the future of industrial PLM, or will on-premise and hybrid models continue to have a place?
The term cloud covers a wide range of deployment models, from SaaS and private cloud environments to hosted infrastructure solutions. While cloud-first has gained significant traction in some areas, adoption in traditional PDM environments still lags behind.
Data sovereignty, performance requirements, and established CAD-centric processes remain strong arguments for on-premise and hybrid models. Wherever sensitive engineering data, deeply integrated systems, or highly specialized configurations play a critical role, on-premise deployments aren’t simply going away.
That’s why I see hybrid models less as a transition phase and more as a deliberate target architecture, and the future isn’t about choosing one approach over another. It’s about using the right architecture for the right purpose. The PDM core remains where control and performance matter most, while additional PLM capabilities can be delivered through web-based or cloud-enabled technologies. Companies that take this approach gain flexibility without giving up control.
The term “digital thread” is everywhere, but not always clearly defined. How do you define it in the context of your work, and how can manufacturers establish a digital thread without creating unnecessary complexity? Looking ahead, auditability will become increasingly important in areas such as CSRD and CBAM. How far is today’s reality from that vision?
In theory, the digital thread is the continuous connection of information created throughout the product lifecycle. What’s important is that information doesn’t exist in isolation. It should be connected and available in the right context whenever it’s needed.
In practice, that’s often where things break down. Incomplete processes, missing integrations, and disconnected systems remain common challenges. Once critical information starts moving through email, spreadsheets, or processes outside the system, the thread is broken. Many manufacturers still have a long way to go before they achieve true end-to-end connectivity.
The goal shouldn’t be to build everything at once. A more practical approach is to develop the digital thread step by step, starting with the most critical information flows. The digital thread is the destination. For most companies, the journey has only just begun.
Let’s stay on the topic of AI. Are manufacturers already using AI-powered assistants or automation in PLM and engineering environments? Where do you see tangible value today, and which applications are likely to become more important across the product lifecycle?
Many manufacturers are exploring AI, but only a fraction have a clear understanding of what they actually want to achieve with it. As a result, productive deployments are still relatively limited.
Where AI is being used today, the benefits tend to appear quickly. It can reduce repetitive tasks, accelerate access to information, and support content creation. In all of these areas, AI helps employees work more efficiently.
The same principle applies here as everywhere else: first build the data foundation, then scale. AI quickly reaches its limits when PDM, ERP, and CAD systems are not working together properly.
The PRO.FILE Smart Manufacturing 2026 Report found that manufacturers consider AI skills their most urgent capability gap. Where do you see the most practical value of AI in PLM and engineering today, and what needs to be in place for AI to scale across the product lifecycle?
Most AI applications today are fairly pragmatic. We’re seeing value in areas such as semantic search, automated classification, extracting information from documents, anomaly detection, and early-stage analysis. These are all use cases where AI helps reduce the effort involved in finding, processing, and working with information.
As always, scaling requires the right foundation. That means high-quality data, complete and digitally maintained datasets, reliable process integration, and data sovereignty. Without a structured product data model, AI lacks the context it needs and quickly reaches its limits.
To identify potential gaps early, I recommend starting with an AI readiness assessment.
Industry 4.0 is giving way to Industry 5.0. What does that mean for the role of engineers in the future development process? Will AI fundamentally change the engineering profession, or will it remain a tool that complements human expertise?
I think the answer depends on the industry, but AI will change engineering—it won’t replace it. The more important question is which activities within engineering are actually affected. What disappears? What gets supported? And what becomes more important?
Routine tasks and documentation can already be handled very effectively by AI. The picture looks different when it comes to areas that can’t easily be delegated, such as systems thinking, requirements gathering, collaboration, and evaluation. As a result, engineers will spend less time executing tasks and more time providing direction, making decisions, and putting results into context.
Regardless of how the technology evolves, one thing remains essential: outcomes still need to be verified by people.
How are manufacturers responding to the growing skills shortage, and what role can a modern PLM system play in preserving knowledge and helping new employees become productive faster?
The skills shortage highlights a simple reality: manufacturers can’t secure their future by hiring more people alone. When experienced employees leave, companies lose more than capacity. They lose knowledge. And without that context, even AI-based approaches become less effective.
A PLM system such as PRO.FILE can help when decisions, changes, and relationships are documented consistently within a Product Data Backbone. The benefits become visible very quickly: faster onboarding, better reuse of existing knowledge, and quicker access to relevant information. Ultimately, the goal is to make product knowledge accessible rather than keeping it locked inside individual teams or employees’ heads.
Markets and technologies are changing rapidly—and in some cases fundamentally. Looking back, what do you think will be remembered as the defining turning point for manufacturing during the 2020s?
I believe the 2020s will be remembered as the decade when data took on a fundamentally different role in engineering. Today, data is no longer just the output of a development process. It’s becoming the foundation for decisions, development, innovation, and entirely new products. AI is accelerating that shift, but it isn’t what caused it.
Manufacturers have had to learn that generating data isn’t enough. The real challenge is treating data as a productive asset and managing it accordingly.
The term PLM is constantly being debated. Do we need a new term, or have expectations simply changed when it comes to what PLM is supposed to deliver?
Rather than phrasing it as a terminology problem, the real question is what companies actually mean when they talk about PLM. Many still think of PLM as a tool. In reality, it’s a strategy for bringing together product information and processes across the organization. Expectations can also create challenges. Companies often assume that PLM will solve their problems. In practice, it tends to do the opposite at first. It exposes gaps, highlights weaknesses in existing processes, and creates a need for clearer structures. That’s not always comfortable, and it can affect adoption.
AI is pushing expectations even further. More and more, PLM is being viewed as a cure-all for challenges that are only partially related to the system itself. Whether the term changes or not, I think the industry is moving in the right direction: from thinking about tools to thinking strategically about data.
One final question. Do you have a guiding principle for separating meaningful innovation from short-term hype? How do you distinguish genuine progress from the latest trend?
My approach is fairly simple: understand the problem first, then talk about technology.
I’m skeptical of evaluating new tools in isolation before there’s a clear understanding of the challenge they’re supposed to solve. Real progress isn’t measured by how modern something sounds but by whether it improves a process. And sometimes the most valuable step is not the most spectacular one. A solution may seem unspectacular at first, yet prove to be strategically important over time.
Dr. Steffen Kunnen is Head of Product for PRO.FILE at Revalize and a key member of the PLM product leadership team. With a background in mechanical engineering and a doctorate focused on product development processes and data management, he leads the strategic direction and ongoing development of the PRO.FILE portfolio.
