Access Official Technical PDF

Share Summary:LinkedInEmail

Digital industry is not a single technology project. For manufacturers, it involves connecting data from equipment and production systems with models and analytical tools that can support decisions across factory operations. A 2018 strategic research agenda for electronic components and systems groups this work into four priorities: digital twins and simulation, artificial intelligence (AI) and machine learning, condition monitoring that can support earlier decisions, and digital platforms that integrate sensors and systems.

Together, these priorities offer a useful way for automation engineers, system integrators, and plant leaders to think about digital transformation. They describe not just tools to adopt, but capabilities that must work across industrial assets, factory levels, and product or system life cycles.

Digital industry starts with a connected view of assets

Industrial data often comes from different parts of an operation: sensors, machines, control systems, and other factory systems. The strategic agenda highlights the need for digital platforms and application development frameworks that integrate sensors and systems. In practical terms, the goal is to make relevant information usable across the digital applications that depend on it, rather than treating each data source as an isolated project.

Integration is a foundation for the other priorities. A simulation model needs information about the asset or process it represents. Analytical methods need data to identify patterns or unusual behavior. Condition monitoring depends on information that can help people assess equipment state. If systems cannot exchange or meaningfully relate their data, those applications are harder to develop and use coherently.

Industrial digital twins connect models with factory assets

The agenda identifies digital twins and simulation models as a priority for evaluating industrial assets at all factory levels and across system or product life cycles. A digital twin can be understood here as a digital representation used alongside an industrial asset or system. Simulation models provide a way to examine how that represented asset or system behaves under defined conditions.

The value of this approach depends on the purpose and scope of the model. A representation may concern an individual asset, a larger system, or a broader factory context. The agenda’s emphasis on multiple levels and life cycles points to a challenge for engineering teams: deciding what must be represented, what information is needed, and how models relate to the operational systems they are intended to help evaluate.

For plant and engineering leaders, this makes model definition and integration important planning questions. A digital twin initiative should be connected to a specific evaluation need, not treated as an end in itself. Teams also need to consider how the model fits into existing system architecture and how it can remain relevant as the asset or system changes over its life cycle.

AI in manufacturing needs a defined operational purpose

Another priority is applying AI and machine learning to detect anomalies or similarities and to optimise parameters. These are distinct analytical tasks. Anomaly detection looks for behavior that differs from an expected or recognized pattern. Similarity detection identifies cases that resemble one another. Parameter optimisation concerns choosing or adjusting values in pursuit of a defined objective.

These methods are most useful when their role is clear. Engineers need to understand what data an application uses, what patterns it is intended to identify, and how its output will inform operational decisions. The agenda establishes these capabilities as research priorities; it does not imply that every process is suited to the same analytical approach or that a model can replace engineering judgment.

Connecting AI applications to industrial systems also brings the platform question back into view. Data access, system integration, and the context needed to interpret results are central to making analytical tools relevant to plant operations.

Move condition monitoring toward earlier decisions

The agenda describes a progression from condition monitoring toward pre-damage warning and online decision-making support. This points to an important distinction: monitoring provides information about equipment condition, while earlier warning and decision support aim to make that information useful before damage occurs or as decisions are being made.

That progression has practical implications for maintenance and operations teams. Information must be timely and understandable enough to support assessment, and the people responsible for equipment need a clear role in responding to it. The agenda identifies the direction of development, rather than prescribing a particular warning method or maintenance workflow.

When considering such applications, decision-makers can ask:

  • What equipment or system condition is the application intended to monitor?
  • What information is needed to identify a relevant change?
  • How will the output support an online operational decision?
  • Which roles are responsible for reviewing and acting on that information?

Integration is the thread linking the priorities

Digital twins, AI, condition monitoring, and industrial platforms are interdependent. A platform can help connect sensors and systems; connected information can support models and analytics; and those applications can provide inputs for evaluation or decision-making. The agenda’s structure therefore encourages a systems perspective rather than a collection of disconnected digital pilots.

For system integrators and technical decision-makers, a sensible planning approach is to define the operational objective first, identify the assets and data involved, and then determine which models or analytical methods are relevant. The architecture should account for how applications connect with the sensors and systems they rely on. This keeps technology choices tied to a defined industrial use case.

Conclusion: Treat digital industry as an integration challenge

The strategic research agenda frames digital industry around four connected areas: digital twins and simulation, AI and machine learning, more forward-looking condition monitoring, and platforms that integrate sensors and systems. For manufacturers, the practical implication is to plan these capabilities together. Start with the decision or evaluation need, then consider the data, models, analytical methods, and system connections required to support it. This approach keeps digital initiatives grounded in factory operations and makes integration a central engineering concern.

Access Official Technical PDF

About The Author