Project data
Core platform
Tag Scraping
Extract and structure project tags from engineering drawings — then detect the tagging standard so scope of work can be assigned before anything hits the database.
- KKS
- RDS-PP
- ISA 5.1
- TIA-606
- CSI MasterFormat
- ISO 81346
- DC Functional
Tag scraping is how a drawing set becomes a completions register. Load P&IDs, single-line diagrams, and general arrangements into Prep Workbench, configure how tags look on your project, and extract equipment tags, document numbers, revisions, and drawing titles with coordinates that sit back on the sheet. Text PDFs are read from the page; scanned or image sheets go through OCR. Tag detection holds 90%+ accuracy across both.
Detection is the other half of the job. If the project already follows KKS, RDS-PP, ISA 5.1, TIA-606, CSI, or ISO 81346, we identify that standard and use it to describe the tag and suggest system, subsystem, and equipment type — a first cut of scope of work. If you do not have a tagging procedure yet, predefined libraries pick the best match. If your site numbering is unique, you apply a custom configuration instead.
Standards, custom patterns, and scope of work
Configure custom tag patterns, or use the predefined libraries. Detection type can run Auto, lock to a single standard, or switch Off if you only want raw pattern matches. Once the standard is identified — KKS, RDS-PP, ISA 5.1, TIA-606, and the rest — we can describe the tag and suggest system, subsystem, and equipment type. That is the first assignment of scope of work, not a second spreadsheet exercise after scrape.
No tagging procedure? Best match, or your own rules
Auto-detect tag types across the industry conventions we ship. If the project has no tagging procedure yet, predefined libraries fit the best match from the drawings in front of you. If the site already has a house standard, lock Detection type to that library. If numbering is unique, add custom N/C patterns. You are not forced into one philosophy — Auto, a named standard, or a custom configuration.
Region fallback when a sheet will not play ball
Document numbers, revisions, and titles are detected automatically from the title block. Optional region-based configuration is there for the unlikely case auto-detection fails — mark a drawing-number, revision, or title area on a preview. Auto-detection still runs first on every sheet; the regions are a fallback, not a mandatory markup pass.
See the extract on the drawing
Review is visual. Extracted tags, document numbers, and revisions sit as overlays on the PDF, with toggle controls so you can show or hide each layer and the quality flags. Side-by-side revision compare is a separate drawing feature we will cover on its own page — the overlay you have here is for verifying this scrape, not for a formal revision walk.
Review, edit, then commit
Nothing lands in the project database until Confirm. The review stage is a full data-management pass: unified tag table, inline edits, status against what already exists, and cleanup before upload. Manual changes happen here — descriptions, hierarchy suggestions, and rejects — so the register you commit is the one you meant.
What it does
- ✓Load multiple PDF drawings in one Prep Workbench session and process them together
- ✓OCR on scanned and image PDFs, plus pattern matching on text drawings — including rotated title blocks
- ✓Configure custom N/C tag patterns, or start from quick presets and predefined industry libraries
- ✓Auto-detect tagging standards — KKS, RDS-PP, ISA 5.1, TIA-606, CSI, ISO 81346 — so scraped tags can be identified and scoped
- ✓If you have no tagging procedure, Auto picks the best-matching library; if you do, lock to that standard or apply a custom configuration
- ✓Extract tags, document numbers, revisions, and drawing titles with overlay coordinates on the sheet
- ✓Optional title-block region fallback when a non-standard sheet beats automatic detection
- ✓Visual overlay on the drawing with toggles for tags, document numbers, revisions, and quality issues
- ✓Full review stage: unified tag table, inline edits, and quality flags before database commit
- ✓Quality analysis for duplicate tags, partial matches, and engineering warnings
Highlights
- • 90%+ tag detection accuracy
- • OCR on scanned sheets; pattern matching on text PDFs
- • File Select → Configure → Process → Review → Confirm before anything is committed
- • Corner-point coordinates on every extracted tag for the drawing overlay
- • Works on P&IDs, SLDs, and general arrangement drawings
Demo
The scrape walkthrough — configure, extract, and review on the drawing.
See the rest of the platform, or talk to us about a demo.