IT/OT integration

Yes, but only if you treat KPI remapping as a controlled translation layer, not a same-day metric replacement. The practical approach is to preserve legacy calculations during transition, define explicit crosswalks to the new taxonomy, and run both in parallel until data quality, source alignment, and reporting impacts are understood. In brownfield plants, disruptions usually come from hidden formula differences, timing mismatches, and weak system integration.

You do not need perfect MES data before starting AI in aerospace, but you do need enough trusted, traceable context to answer a specific use case. The minimum is usually event history, routing and work order context, part and serial or lot genealogy, nonconformance outcomes, and stable timestamps. Gaps in integration, master data, and validation can limit results quickly.

In aerospace, there is no single correct system for configuration control. Most mature organizations split responsibilities: engineering baselines in PLM or a dedicated configuration system, commercial and revision identifiers in ERP, and as-built execution and genealogy in MES. The right pattern depends on integration quality, validation burden, and how much change you can safely introduce to existing qualified systems.

ISA-95 (IEC 62264) is an international standard that defines models and terminology for integrating business systems (like ERP) with manufacturing operations systems (like MES, SCADA, and control). It does not guarantee interoperability by itself, but provides a common structure for data, functions, and interfaces that must still be engineered, validated, and maintained in each plant.