What Asset Administration Shells Deliver for Maintenance

In the era of Industry 4.0, data-driven processes are the foundation of modern maintenance strategies. Two key terms are increasingly coming into focus: predictive maintenance and predictive servicing. While often used interchangeably, they differ significantly – both in their objectives and technical approaches.
A decisive factor for the success of both strategies is the Asset Administration Shell (AAS) – the digital twin of a physical asset. This article explains the differences between these concepts and shows how the AAS serves as a central data platform enabling both approaches.

What is predictive maintenance?
Predictive maintenance focuses on carrying out maintenance actions not at fixed intervals, but based on real-time condition data. It relies on sensors, real-time data, and algorithms that detect anomalies or wear early on.

What is predictive servicing?
Predictive servicing takes a broader approach: it includes not only maintenance but also inspection, repair, and optimization – all based on up-to-date condition information.

What they share: Both approaches use sensor data, AI-powered analysis, and digital tools to reduce failures, increase availability, and cut costs.

The Asset Administration Shell (AAS) is the digital representation of a physical asset – structured, standardized, and machine-readable. It forms the basis for interoperability across systems and enables consistent data processing across manufacturer and system boundaries.

How the AAS Supports Predictive Maintenance

• Access to condition data: Parameters like temperature, vibration, or runtime are structured via submodels such as “Operational Data” – semantically defined, machine-readable, and interoperable.
• Simple data integration: Analytics tools or AI models can access relevant data – such as technical specifications or usage statistics – through defined interfaces.
• Documentation and spare parts: The AAS can standardize and provide access to maintenance histories, manuals, and spare parts lists.
• Standardization for heterogeneous machine fleets: With the AAS, even diverse machines can be integrated into a unified maintenance strategy.

AAS in the Context of Predictive Servicing

• Lifecycle data at a glance: The AAS tracks the full lifecycle of an asset – from commissioning to failure analysis and repair cycles.
• Decision support: Submodels provide data for assessing failure probabilities, wear progression, and maintenance costs.
• Process automation: Integration with MES, ERP, or CMMS systems enables automated workflows – for example, planning and triggering servicing actions.
• Industry 4.0 communication: Open standards like OPC UA ensure seamless communication between systems, equipment, and users.

A medium-sized company specializing in CNC machining centers implements a predictive servicing strategy. The goal: to operate both in-house production and delivered machines more intelligently.
Implementation steps:

• Sensors & data acquisition: Vibration sensors on the main spindles send data via OPC UA to an IIoT platform.
• Digital twins for each machine: Each asset receives its own AAS – including submodels for operational data, maintenance history, and spare parts logistics.
• Automated analysis: The platform uses AI to analyze AAS data – if anomalies are detected, a servicing task is triggered automatically.
• Seamless service process: Service technicians access all relevant data via mobile devices – including manuals and parts lists.

The benefits:

• Reduced unplanned downtime
• Shorter repair times (MTTR)
• Optimized spare parts logistics
• More efficient maintenance planning

Whether for precise maintenance actions or strategically planned servicing activities – without a structured, consolidated data foundation, these approaches remain inefficient. The AAS provides the foundation for true interoperability and automation.