Product Lifecycle Management in the age of AI: Engineering trust for regulated industries

Artificial intelligence (AI) is rapidly becoming part of products that operate in some of the world's most demanding environments. From medical devices in healthcare and life sciences and automotives and appliance manufacturing to aeronautical and telecom services and critical infrastructure in aerospace, defense and energy, AI is enabling products to become increasingly intelligent, adaptive and autonomous. While this presents enormous opportunities for innovation, it also raises an important question for organizations operating in regulated industries:

How do we ensure AI-enabled products remain safe, explainable and trustworthy throughout their lifecycle?

For decades, Product Lifecycle Management (PLM) has provided the engineering backbone for managing complex products — from requirements and design through manufacturing, service and continuous engineering change. As AI becomes an integral part of modern products, the challenge is no longer simply about introducing another software capability. It is about understanding how engineering governance must evolve when product behavior itself becomes increasingly intelligent.

Engineering trust in regulated industries

PLM is often associated with managing engineering data, Bills of Materials or product configurations. While these capabilities remain fundamental, the real value of PLM is much larger. It provides the governance that allows engineering organizations to build products that can be trusted throughout their operational life.

Every requirement is connected to an engineering decision and every software release follows a controlled change process. Verification activities produce evidence, validation confirms expected behavior and engineering changes remain traceable long after a product enters service. In regulated industries, where products frequently remain operational for decades, this discipline is essential. Today, manufacturers must be able to explain not only how a product was originally designed, but also every significant change that has influenced its behavior over time.

This ability to maintain continuity across requirements, engineering, manufacturing, operations and service through a digital thread is one of PLM's greatest strengths. Ultimately, PLM has never been simply about managing product information — it has been about engineering confidence.

The future of AI in regulated industries will not be defined by how intelligent the models become, but by how confidently we can engineer, validate and govern them throughout their lifecycle.

Expanding the PLM scope

AI changes the engineering conversation because it introduces a new category of engineering artefacts.

Traditionally, product behavior has been largely determined by mechanical design, electronics and software. AI-enabled products, however, are increasingly influenced by models, training data, inference pipelines, validation datasets and continuous operational feedback. As these evolve, the behavior of the product itself may evolve, even when no conventional software changes have been introduced.

This presents a different engineering challenge. If AI influences how a product is built and behaves, then AI itself becomes part of the product. Consequently, AI models and the engineering evidence supporting them can no longer be viewed as isolated technology assets. They become part of the overall product lifecycle and should be governed with the same engineering discipline applied to every other critical component.

AI does not diminish the role of PLM in the future — it reinforces it.

With AI, the product lifecycle naturally expands to include AI artefacts alongside mechanical, electrical and software components, allowing organizations to maintain complete traceability from product requirements through deployment and continuous improvement.

Engineering Documents

Product Data

Product Configuration

Product Lifecycle

AI-Enabled Product Behaviour

Building trustworthy AI with a robust PLM foundation

Fortunately, organizations do not need entirely new engineering disciplines to support responsible AI adoption. Much of the required foundation already exists within modern PLM environments.

While PLM provides lifecycle governance across engineering artefacts, systems engineering connects stakeholder requirements with system behavior. The Digital Thread maintains traceability across disciplines, while Digital Twins create virtual environments where increasingly complex operating scenarios can be validated before products are deployed. Together, these capabilities enable organizations to move beyond validating individual software functions towards validating complete product behavior.

Imagine an engineering environment where AI models are directly connected to product requirements, hazard analyses, verification evidence and Digital Twin simulations. Every engineering change can be assessed for its impact, every validation activity produces traceable evidence, and operational feedback continuously strengthens future product releases. Human expertise remains central throughout the lifecycle, providing oversight and accountability whenever product behavior evolves.

Viewed independently, PLM, Digital Threads, Digital Twins and systems engineering each deliver significant value. Combined through an integrated lifecycle approach, they provide the engineering foundation required to deploy AI responsibly within regulated industries.

AI in Product Lifecycle Management: Enhancements. Advances. Evolution.

Perhaps the most significant implication of AI is not the technology itself, but the way it changes the role of Product Lifecycle Management

For many years, PLM has evolved from managing engineering documents to governing product data, product configurations and, ultimately, the complete product lifecycle. AI now presents the next logical step in that evolution.

As AI becomes embedded within products, what differentiates one product version from another may no longer be limited to its physical or software configuration. Increasingly, it will also be the product's behavior — how it interprets information, responds to changing operating conditions and supports human decision-making such as in humanoid robots.

This suggests that the future of PLM may extend beyond managing what a product consists of to governing how that product behaves throughout its operational life. AI models, validation evidence, simulation results and operational feedback may increasingly become first-class engineering artefacts connected through the Digital Thread and managed under the same lifecycle governance that has served regulated industries for decades.

This evolution is not about replacing existing engineering disciplines. It is about extending them to address a new generation of intelligent products while preserving the engineering trust that regulated industries depend upon.

These themes were explored in greater depth during my recent presentation at the “WeAreDevelopers World Congress”, where I discussed how Product Lifecycle Management, Digital Twins and engineering governance can support the validation of AI-enabled products operating in regulated industries. You can learn more about engineering architecture and practical implementation patterns in this presentation.

The future is trustworthy AI

Ultimately, AI will not redefine the purpose of Product Lifecycle Management — it will reinforce it. Engineering has always been about building confidence that products will behave as intended throughout their lifecycle. In the age of AI, that confidence will increasingly depend not only on managing product configurations, but also on understanding, validating and governing product behavior. By combining AI innovation with the proven disciplines of PLM, organizations will be well positioned to build the next generation of trustworthy, intelligent products.