Digital industry is not a single technology project. It is the coordinated use of digital models, data analytics, connected equipment and manufacturing processes to support industrial assets across their lifecycles. For automation engineers, system integrators and plant leaders, the practical challenge is joining these capabilities into systems that can be evaluated, monitored and improved without losing sight of operational requirements.
A 2020 strategic research agenda for electronic components and systems identifies five priorities for digital industry: digital twins and simulation, artificial intelligence (AI) and machine learning, condition monitoring, integrated digital platforms, and sustainable manufacturing in a circular economy. Together, these themes describe a broad direction for industrial research and development—not a checklist of technologies that can be deployed independently.
Digital industry starts with lifecycle visibility
Industrial assets operate within larger systems, and their value and condition change over time. The agenda highlights digital twins and simulation models as tools for evaluating assets at different factory levels and across system or product lifecycles. A digital twin, in this context, is a digital representation used alongside the physical asset or system. Simulation models can help examine how that system behaves under defined conditions.
The emphasis on multiple factory levels matters. An asset-level view alone may not explain how a machine interacts with a production line or a wider factory system. Likewise, a model useful during design may need to remain relevant as equipment is operated and maintained. The research priority is therefore not simply to create a virtual model, but to develop ways to use models for evaluation across industrial contexts and over time.
Applying AI in manufacturing to support decisions
The agenda also identifies AI and machine learning as priorities for detecting anomalies or similarities and optimising parameters. These applications connect data analysis to specific industrial questions: Is observed behaviour unusual? Does it resemble a previously recognised pattern? Which operating parameters may warrant adjustment?
These are distinct tasks. Anomaly detection focuses on identifying deviations; similarity detection looks for patterns that resemble other observations; parameter optimisation concerns selecting or adjusting operating values. In each case, analytical results need to be considered within the relevant process and operational context. The agenda frames these methods as research challenges rather than suggesting that AI can make every decision autonomously.
For plant teams, that distinction supports a disciplined approach: define the problem first, then determine what data and models are needed to address it. Analytics can only be useful when its outputs connect to a clear evaluation or decision process.
Condition monitoring must extend beyond individual machines
Condition monitoring is another central priority, with the agenda pointing toward a broader progression: monitoring equipment, providing pre-damage warnings and supporting online decision-making. It also identifies standardised communication scenarios as important for collecting large volumes of data across extensive, including remote, sites.
This connects asset health to system architecture. Monitoring a machine produces limited value if its information cannot be communicated in a form that other systems can use. Consistent communication approaches can support data collection across equipment and locations, while decision support can help translate monitoring information into an operational response. The agenda does not prescribe a particular protocol or implementation; it identifies standardisation and scalable data collection as areas requiring attention.
Integrated platforms connect sensors, actuators and systems
Digital industry also requires platforms and application-development frameworks that integrate sensors, actuators and systems. Sensors provide information about equipment and processes; actuators carry out control actions. Connecting these components with broader systems creates a foundation for applications that use operational data and interact with industrial processes.
Integration is more than making devices communicate. Different systems may represent information in different ways, and applications need a suitable structure for accessing and using it. This makes platform development closely related to interoperability: systems must exchange information with enough consistency for it to retain practical meaning. For integrators, the agenda’s focus reinforces the importance of considering device integration, data handling and application requirements together rather than treating each as a separate layer.
Sustainable manufacturing belongs in the digital agenda
The fifth priority is sustainable manufacturing in a circular economy. Its inclusion signals that digital industry is not only about asset performance or factory connectivity. Manufacturing approaches also need to be considered in relation to resource use and the broader lifecycle of products and systems.
The agenda places this topic alongside digital twins, analytics, monitoring and platforms. That positioning encourages decision-makers to consider how digital capabilities relate to manufacturing processes and sustainability objectives. It does not set out a specific circular-economy method, but makes sustainable manufacturing a research and development priority in its own right.
Turning research priorities into practical work
The five themes are interdependent. Digital models need suitable information about physical assets. AI and machine learning depend on data and defined tasks. Condition monitoring relies on communication and decision-support approaches. Platforms must integrate devices and systems, while sustainable manufacturing needs to be considered as part of the wider industrial transformation.
For engineering and operations teams, the useful takeaway is to assess digital initiatives as connected system problems. When evaluating a project, consider:
- Lifecycle scope: Which assets, factory levels or product stages need to be evaluated?
- Decision purpose: Is the goal to detect anomalies, identify similarities, optimise parameters or support another defined task?
- Data and communications: How will information be collected and communicated across equipment and sites?
- Integration: How will sensors, actuators and existing systems fit into the application framework?
- Sustainability: How does the work relate to sustainable manufacturing and circular-economy priorities?
These questions do not replace detailed engineering or project planning. They help expose dependencies early and keep technology choices connected to the intended industrial use.
Conclusion
The digital industry priorities outlined in the strategic agenda point to a manufacturing environment where models, analytics, monitoring and integrated platforms work together, with sustainable manufacturing treated as part of the same transformation. For industrial decision-makers, the practical implication is to plan beyond isolated deployments: define the operational purpose, account for lifecycle and integration needs, and consider how data can support decisions across assets and systems.