Accelerating PAS-X Migration with Structured AI Automation

Structural differences increased migration risk

The migration from PAS-X 3.1.8 to 3.4 introduced significant structural gaps between system versions. Documentation of underlying XML structures was limited, increasing the risk of transformation errors and import failures.

All generated MBR and PVL files had to remain fully XML valid to ensure successful system import. Early attempts to train AI models directly on raw MBR XML files delivered limited results due to the architectural differences between versions.

A more controlled, use-case-driven approach was required to ensure reliable and compliant migration.

A controlled AI-enabled migration framework

NNIT refined the approach to focus on two defined migration scenarios:

• Automated conversion of material placeholders into SAP BOM references

• Dynamic generation of Global Custom Attribute lists with associated formulas

Instead of training AI directly on raw XML, NNIT used Alera to provide a low-code environment where MBR designers can describe migration logic in natural language. Alera generated and executed Python-based migration scripts while preserving XML validity and ensuring full transparency. All Alera services are hosted within NNIT’s Azure tenant to ensure compliance with strict EU data processing requirements. Validation confirmed successful and repeatable reproduction of the targeted migration scenarios in a controlled setup.

Execution approach and expertise

NNIT combined deep PAS-X migration expertise with practical AI implementation. Rather than applying AI generically, the team structured the problem into concrete, validated use cases and embedded automation within a controlled low-code framework aligned with real MBR design practices.

This ensured technical reliability, XML integrity, and measurable value creation driven by process and system knowledge.

Business and operational outcome

The structured AI framework delivered measurable efficiency improvements and established a scalable migration model.

• Approximately 50 hours saved per use case

• 20+ additional migration use cases identified and in progress

• Validated XML output ensuring successful system imports

• Controlled, repeatable migration scenarios

The initiative positioned the organization to modernize its MES landscape more efficiently while maintaining compliance and technical integrity.