TL;DR

This article discusses how Werk24 transforms technical drawings into customer-specific output formats, creating canonical part descriptions that integrate seamlessly with ERP systems and enable automated costing workflows.

From Drawing to Your Data Model: Customer-Specific Output Schemas for Canonical Part Descriptions

Most teams don’t struggle with finding technical drawings. They struggle with what comes next: turning a drawing into the exact fields their systems need.

ERP systems, quoting tools, supplier portals, and internal part catalogs all expect structured attributes (“inner diameter”, “wall thickness”, “material”, “length”, …). Technical drawings, however, encode those attributes in a thousand different ways—depending on the designer, the supplier, the country, the department, and sometimes the individual engineer.

Werk24 closes that gap by converting a drawing into customer-specific output formats: not only extracting dimensions, but producing a canonical description of the characteristic values that define a part family.

Why “dimensions extracted” is not the same as “part described”

Extracting dimensions is a necessary first step—but it rarely solves the real business problem.

A drawing might contain:

• an outer diameter (OD) only,
• an inner diameter (ID) only,
• OD + wall thickness,
• ID + wall thickness,
• a diameter called out on a sectional view,
• or even a note-based definition that doesn’t follow your company’s “standard” wording.

All of these can describe the same physical part.

What your downstream systems typically need is something like:

• Inner diameter
• Outer diameter
• Length

…regardless of how the drawing chose to dimension it.

That’s the difference between “we read the drawing” and “we understood the part.”

Part families need domain knowledge

To turn a drawing into a canonical part description, you need to know what characterizes the part family.

A tube (or a simple cylinder) is a good illustrative example: you can fully describe it with just a few characteristic values, even if the drawing provides different dimensional combinations.

But this is also where reality hits:

• Companies assume their naming conventions and definitions are universal.
• They are not.
• Even within one company, two departments can use different conventions for the same attribute.

So the key question becomes:

“What does your organization consider the defining attributes of this part family?”

Werk24 is designed to encode exactly that.

How Werk24 produces customer-specific output formats

Werk24 combines two layers:

1. Reliable extraction from technical drawings (text, symbols, dimensions, tolerances, notes).
2. Customer-provided domain knowledge that tells the system how to interpret what it finds.

Instead of stopping at “these are the dimensions on the page,” Werk24 uses your domain knowledge to output:

• a normalized, canonical attribute set (your preferred fields, names, and definitions),
• with conversions and derived values where needed (e.g., wall thickness → ID/OD mapping, unit normalization, tolerance normalization),
• in the format your systems require.

For the tube example, that means you can feed Werk24 a drawing with any of these inputs:

• OD + wall thickness
• ID + wall thickness
• OD only (plus a note)
• ID only (plus a note)

…and receive a consistent output like:

• inner_diameter
• outer_diameter
• length

Even if the drawing never explicitly labels them that way.

Where this approach reaches its limits

The goal is not to reconstruct every feature coordinate and fully parameterize highly complex parts.

When a component requires dozens of interdependent measurements (think: many holes with positions, patterns, offsets, and reference frames), “canonical description” becomes a different problem category.

Werk24’s sweet spot for customer output formats is:

• simple to mid-complexity components,
• where the business value comes from quickly identifying and normalizing the defining characteristics,
• rather than exhaustively encoding every single geometric detail.

Business outcomes: ERP automation and top-down costing

1) Automate ERP population (e.g., SAP)

Many ERP environments already have fields that are currently filled manually from the drawing:

• dimensions and key attributes
• material and treatments
• tolerance class / surface requirements (where applicable)

Once Werk24 outputs those canonical fields in your preferred schema, you can:

• automatically populate ERP master data,
• validate incoming RFQs against existing part definitions,
• reduce manual typing and interpretation errors,
• standardize how parts are described across teams.

Because the output is already in your “system language,” integration becomes straightforward—whether you prefer JSON, CSV, XML, or a system-specific payload.

2) Enable top-down price calculation

Canonical part descriptors are also the foundation for top-down costing:

• identify similar parts in your history,
• compare canonical attributes (not the raw drawing dimensioning style),
• estimate price based on similarity and past quotes.

This is especially effective when implemented together with experienced partners who already have costing logic and historical data—turning drawings into consistent, comparable features is the missing link that makes similarity-based pricing practical.

The core idea: drawings in, customer outputs out

Different customers (and different departments) define “the right output” differently.

Werk24’s approach is to make that difference explicit:

• You define the characteristic values that matter for a part family.
• Werk24 interprets the drawing through that lens.
• You receive consistent, customer-specific output formats that plug into ERP, quoting, and analytics.

Technical drawing in. Canonical part description out.

If you want to explore what this looks like for your part families, we can start with a small set of sample drawings and define the attribute schema together.