Armeta's contextualization engine reads any PDF piping isometric — scanned paper, hand-drawn legacy drawings, rasterized CAD plots, flat image exports — and outputs structured spool data: pipe routing, fittings, welds, dimensions, materials, and quantities. The output feeds fabrication tracking, MTO, and procurement, and it's contextualized against your P&ID and line list data so the isometric becomes part of the unified engineering knowledge graph.
The isometric is the most granular piping document you have. Armeta extracts the engineering objects on it — not just text, but pipe geometry, components, welds, and the BOM — with each entity tied back to the source region on the drawing.
North arrow, elevation changes, and coordinates parsed from the isometric projection so the routing is preserved as structured data — not just an image.
Elbows, tees, reducers, flanges, and inline valves identified by symbol and tagged with size, type, and material spec.
Field welds, shop welds, and weld numbers captured with their type and position. Feeds straight into weld maps and NDT planning.
Cut lengths, center-to-face, and center-to-center dimensions extracted with their reference points so spool fabrication data is shop-ready.
Pipe spec, line number, service, and any spec breaks captured per segment, with the line number resolved against the P&ID and line list.
The bill of materials on the isometric is parsed into a structured table: every component with quantity, description, spec, size, and rating.
Drawing number, revision, sheet, project, and the line-number tie back to the P&ID — used as the indexing key for every other extraction.
The isometric does not stand alone. It is the spatial expression of a line that already exists on the P&ID and in the line list. Armeta resolves the same line across documents so connectivity, attributes, and BOM data live on one record — not three siloed copies.
The result: every fitting, every weld, and every dimension on the isometric is unambiguously attached to a line that lives end-to-end in the engineering knowledge graph.
The line number on the isometric title block is matched against the same line on the P&ID. The schematic-level connectivity from the P&ID and the spatial routing from the isometric resolve to one record in the engineering knowledge graph.
Spec, size, and design conditions on the isometric are reconciled against the line list. Where attributes disagree, both values are retained with their source documents and surfaced as a structured discrepancy — never silently averaged.
Once isometric content is structured and contextualized, the workflows that depend on it stop being manual roll-ups and become queryable.
Most piping isometrics in operating facilities are not smart PDFs. They are scanned paper, hand-drawn legacy sheets, rasterized CAD plots, and flat image exports — with no embedded metadata, no selectable text, and no vector layer.
Armeta reads them as-is. No conversion step, no pre-vectorization, no requirement to re-author the drawing in a modern CAD tool before it can be processed.
Every isometric extraction Armeta delivers is validated against the source drawing by our engineering team before it reaches you. Accuracy metrics on your own isometrics are produced as part of the initial document audit — our ten-document free assessment that establishes baseline performance before any paid engagement.
Every isometric extraction is reviewed by an Armeta engineer against the source drawing before delivery. Not raw model output.
Every component, weld, and dimension in the structured output links back to the exact region on the source isometric it was extracted from.
Random spool sampling against ground-truth BOMs on every engagement. Accuracy metrics produced on your own isometrics.
Discrepancies between the isometric BOM, the P&ID, and the line list are surfaced as structured reports rather than absorbed silently.
Three deployment modes, chosen to match the security and governance constraints your engineering records actually live under.
Fastest time to value
Armeta-hosted, isolated per customer. Your engineering documents and extracted data remain segregated from every other customer environment.
Your network, your perimeter
Armeta's contextualization engine runs inside your network. Your documents never leave your environment. Deployed with customers today under data residency and air-gap requirements.
Managed inside your tenant
Armeta-managed deployment inside your own cloud tenant. Cloud operational model, customer-controlled perimeter.
Before any procurement conversation, see Armeta run on your actual piping isometrics — spool by spool, with BOMs reconciled against your P&IDs and line lists. Not sample data. Not a generic demo. Your isometrics, contextualized, delivered as a structured report in one engagement.
