Original ARAN VISION analysis and adaptation of the Siemens white paper “Evaluation of production data”, document DIFA-B90394-00-7600, published in 2026.
Data becomes valuable when it informs operational decisions
Recording large numbers of tags does not by itself improve productivity. Operators, production supervisors and maintenance teams need live data, process history, alarms and machine state in context to identify bottlenecks, reduce downtime and control quality. Start by defining the problem and required decision, then determine the necessary data, accuracy, frequency and retention period.
A shared database turns isolated HMIs into a coherent view
Many lines have separate HMIs and archives for each machine or area, leading to inconsistent timing, differing formats and difficult event tracing. SCADA can consolidate and centrally archive data and alarms from multiple subsystems, provided time synchronization, tag quality, naming, data ownership and retention policies are standardized from the start. An attractive dashboard built on invalid data merely hides errors.
Alarm analysis must go beyond counting messages
Analyze occurrence frequency, active duration, acknowledgement time and event sequence. A hit list identifies frequent alarms, but a frequent alarm is not necessarily the root cause. Reliable analysis aligns process state, interlocks, operator changes and pre-stop events on a shared timeline. The goal is to reduce alarm floods and define clear actions for important messages.
Site-wide analysis without replacing every system
The white paper describes PM-ANALYZE for SIMATIC WinCC Unified and WinCC V8, which can collect, map and filter messages and trends from different sources and produce documented reports. A broader operational view does not necessarily require replacing all existing equipment. Before implementation, assess each source interface, event ordering, archive capacity, access permissions and reporting load.
Calculate OEE, MTBF and MTTR using shared definitions
Performance indicators are comparable only when their calculation boundaries are clear. OEE requires definitions of production schedule, downtime, reference speed and acceptable product quality. MTBF without an agreed failure definition, or MTTR without accurate repair start and end times, can mislead. Each KPI should support drill-down from plant and line to machine level, linking changes to tags, machine states, faults and user actions.
AI requires structured data and controlled access
WinCC Unified and WinCC V8 can provide structured access to live values, history, alarms and project context. The document discusses open interfaces and MCP-based integration patterns as foundations for operator assistants, including natural-language event search, report preparation and troubleshooting guidance. Implement these with role-based access, complete request logging, write-operation limits and human approval. Model responses must not replace interlocks or safety decisions.
The electronic shift log is the plant's operational memory
Verbal shift handovers often lose important details. An electronic shift log records events, priorities, status, follow-up owners and revision history, and can link events to equipment or areas in the plant hierarchy. Search and access through SCADA or a browser accelerate recurring-event troubleshooting, provided permissions reflect roles and responsibilities.
Combining WinCC, PM add-ons and industrial computing
This architecture can use WinCC V8 or WinCC Unified for visualization and operation, PM add-ons for analysis and process management, and SIMATIC IPCs for industrial execution. The document also mentions OPC UA, REST APIs and MQTT. Select components based on tag counts, logging rates, retention, redundancy, users, network requirements and compatible versions; product names alone do not define an architecture.
A practical implementation path: start with one bounded problem
Choose a critical line or asset and set a measurable objective, such as reducing downtime diagnosis time. Record tags, alarms, data quality, time sources and signal owners, then build a shared timeline, a few KPIs and a pilot shift report. Expand after verifying data accuracy and actual user adoption. Acceptance criteria should cover calculation correctness, display latency, recovery after disconnection, access controls and change traceability.
Scope and limits of technical information
This article supports conceptual design and preliminary assessment. Verify exact version, license, add-on and hardware capabilities against the applicable documentation and project contract. As Siemens notes in the original document, general descriptions may change as products evolve, and binding features must be explicitly included in the final contract.