Standardisation
AI and automation: Scaling the solution
Manual mapping is no longer necessary. AI and machine learning now enable automated, highaccuracy transformation from JIP33 procurement data to CFIHOS handover models— leveraging the semantic structure we just described: 1. Extract: Ingest vendor data sheets
( PDF, Excel) 2. Normalize: Convert units( psi → bar), standardise codes( WCB → A216-WCB) 3. Map: Match fields to CFIHOS RDL predicates using fuzzy logic and ontology lookup.
4. Attach: Embed links to API / ASME standards and test reports.
5. Score & Validate: AI assigns confidence scores— high matches auto-process; low matches flag for SME review
This approach reduces mapping effort from weeks to hours, with accuracy exceeding 90 % on standardised fields. Combined with semantic anchors, AI ensures that even custom attributes in the 20 % zone are mapped intelligently without breaking alignment.
Bridging the gap: How do we get there?
We now have the standards( JIP33) and the destination( JIP36 / CFIHOS), but implementation remains the challenge. Most manufacturers and EPCs do not have dedicated semantic data
engineering teams to build these AI pipelines from scratch. As Onno emphasizes:“ You don’ t build your own website; you hire an expert.”
Digital interoperability providers This need has given rise to a new category of partners: Digital Interoperability Providers. These are not traditional software vendors selling proprietary design tools. Instead, they function as data logistics hubs, offering:
• Hosted master indexes and predicate libraries where mappings between API, ISO, ASME, and CFIHOS are pre-established.
• AI-assisted mapping workflows that scale across projects.
• Compliance-ready documentation frameworks for global operations.
Readers looking to explore these solutions should search for providers specialising in“ CFIHOS Data Solutions,”“ Capital Project Information Handover,” or“ Technical Asset Information Management.” The goal is awareness— not promotion— so the industry understands these capabilities exist and can accelerate adoption.
Conclusion: From stalemate to execution
For decades, the valve industry has faced a stalemate: too many standards, too little alignment and too much manual effort.
Both Onno and Glen agree that volunteerdriven standards work and thin budgets make progress slow. Today, the pieces fit together:
• JIP33 standardizes procurement.
• CFIHOS( JIP36) secures handover.
• Semantics preserve meaning.
• AI automates scale.
• Digital interoperability providers make implementation practical.
The question is no longer if interoperability is possible— it’ s whether the industry will act collectively to adopt it. This is not a future vision. It’ s achievable today and already in motion. For an industry built on precision, it’ s time our data matched our engineering.
About the author Michael Maar is a student at the University of Houston studying Supply Chain & Logistics Technology. In his Supply Chain roles at MRC Global, he has supported capital projects and quality / automation initiatives, including developing valve data tools and streamlining procurement workflows.
34 Valve World February 2026 www. valve-world. net