General CAD is a drawing tool that knows nothing about what you make. Vertical CAD is a manufacturing tool that happens to draw. That difference is where the value sits, and it is why specialised industries keep commissioning their own software.
What general CAD cannot know
| Industry | Domain knowledge general CAD lacks |
|---|---|
| Sheet metal | Bend allowance, K-factor, minimum flange, tool access, flat pattern |
| Furniture | Joinery rules, panel thickness, edge banding, grain direction, hardware |
| Textile | Grading across sizes, seam allowance, fabric width, grain and lay rules |
| Glass / stone | Edge profiles, cut-out clearances, breakage risk, slab optimisation |
| Signage | Material limits, weeding for vinyl, tool paths for routing |
| Structural fabrication | Section libraries, connection standards, weld symbols, cut lists |
In every one of these, the knowledge exists in your experienced staff and in spreadsheets beside the workstation. Vertical software moves it into the tool, where it applies consistently and does not leave when someone does.
The features that actually justify a build
Nesting and material optimisation
For anyone cutting from sheet or panel stock, this is usually the single highest-return feature. Automatic nesting that packs parts to minimise offcut pays back in material cost on every job — and it accounts for real constraints general CAD ignores: grain direction, remnant reuse, common-line cutting, and clamp zones.
Parametric domain libraries
Not generic blocks, but parts that understand themselves. A cabinet that recalculates its own panels when width changes. A duct fitting that regenerates its development when the angle changes. This is where most of the drawing time disappears.
Manufacturability checking
Warning at design time that a flange is too short to bend, a hole is too close to an edge, or a panel exceeds sheet size. Catching those before the shop floor is worth considerably more than catching them after.
Automatic output
Cut lists, BOMs, labels, cost estimates, and machine files generated from the drawing rather than retyped. The re-typing step is where errors enter, and removing it is often what convinces a workshop to adopt.
The feature that decides adoption is rarely the drawing. It is what comes out the other end. A tool that produces a correct cut list and a machine-ready file wins over one that draws beautifully and leaves the operator to translate.
Why the scope is manageable
Vertical 2D CAD is a far smaller undertaking than general CAD, because you exclude almost everything:
- 2D only — no solid modelling, no B-Rep kernel, no booleans on trimmed surfaces.
- A closed set of entity types — lines, arcs, circles, polylines, text. Not arbitrary freeform surfaces.
- Domain-shaped workflow — guided steps rather than a general drawing environment.
- DXF for exchange — documented and adequate for most manufacturing handoff.
The reason vertical CAD is viable for a small team is that the industry itself narrows the problem. You are not building CAD; you are building one industry's workflow that happens to include drawing.
Where the real engineering is
Not in the drawing canvas, which is well-trodden. It is in:
- Encoding the domain rules correctly — bend allowances, joinery standards, grading rules. Getting these wrong produces confidently incorrect output, which is worse than no software.
- Nesting algorithms — genuinely hard optimisation, usually heuristic, and where a lot of the value concentrates.
- Unfolding and development — flat patterns for bent or curved parts, which must account for material behaviour.
- Machine output — correct files for your specific cutting, routing or bending equipment.
Domain rules must come from practitioners, not from a specification document. The bend allowance a shop actually uses, the joinery they actually cut, the seam allowance their machinists expect — these are often different from the textbook values, and the difference is what makes the software trusted or ignored.
Adoption is the real risk
Vertical CAD fails on adoption more often than on engineering. What determines it:
- Faster than the current method for the common job, not just the complex one.
- Respects existing drawing conventions — output that looks like what the shop already reads.
- Interoperates — imports what customers send, exports what machines need.
- Allows override. Experienced staff must be able to break the rules when the job requires it, or they will abandon the tool at the first exception.
- Works offline, because shop floors have unreliable connectivity.
Building CAD tooling for a specific manufacturing industry? Tell us the industry and the rules your team applies. See our CAD engineering service and the add-in vs standalone decision.