Rule-Based PMI Creation
Rule-Based PMI Creation
Most PMI on a part is routine. Let rules prepare the routine — and keep your engineers for the judgement.
Rule-based PMI creation prepares the repetitive share of product manufacturing information in Siemens NX from your norms and templates: datum schemes per your family strategy, standard callouts for threads, fits and surfaces, note sets and model views per your MBD template — written as semantic PMI through NXOpen. The engineer then does what only an engineer can: adds the judgement-heavy tolerances, adjusts what the rule got generically right but specifically wrong, and releases. No magic, no percentages promised — time saved on the routine share, measured on your parts.
It is the “create” engine of our AI-Assisted Geometry Creation discipline — the rules-and-templates counterpart to the conversational MBD & PMI Authoring Assistant, and the creating counterpart to our MBD/PMI Compliance Automation, which checks what gets written.
Who this is for
- Your MBD rollout keeps stalling on the same fact: annotating parts takes hours, and there are many parts.
- Within a part family, most annotations follow the same pattern — and are still placed by hand every time.
- Datum strategies and standard callouts exist in your norms, but applying them consistently depends on who annotates.
- You want the annotation work reduced honestly — not a tool that claims to replace the engineer’s judgement.
// One flange housing — the rule set prepares, the engineer completes
- Datum scheme per family strategy (A: mounting face, B: pilot bore) · semantic
- Standard callouts: threads, fits and surfaces per WN-311 series rules · semantic
- Note set + model views per MBD template TN-12 · semantic
- Function-critical tolerances — the judgement the rule cannot make
- Case-specific adjustments where the generic rule fits generically
- Review & release — signed, logged, same responsibility as hand-made PMI
The honest split is the product. Rules prepare the routine share — how large that share is depends on your family and your norms, which is why the PoC measures it on your parts instead of promising a percentage. The judgement share stays with the engineer, on purpose.
What the rule sets prepare
Family datum strategies applied
Your preferred datum strategy per part class, placed as semantic datum features — consistent across the family instead of depending on who annotates.
Threads, fits, surfaces per norm
The callouts your norms already define — thread series, fit pairings, surface classes — written as machine-readable PMI, not decorative text.
Note sets & model views
Standard note blocks and view organization per your MBD template — the boilerplate share of every annotation job, prepared in seconds.
Prepared across a family — where it carries
Homogeneous families can be prepared in series runs with a review queue. What carries and what doesn’t, the PoC measures — we don’t promise rates we haven’t seen on your parts.
How it works
Your norms and MBD templates are encoded once, with your senior designers — datum strategies, callout rules per feature class, note sets. The preparation then runs through NXOpen’s semantic PMI APIs on your part templates; AI assists where interpretation genuinely helps (feature classification, matching norm rules to actual geometry), and everything is written as reviewable, editable PMI the engineer completes. On your infrastructure, on-premises or EU cloud.
Honest limits
Rules prepare; they don’t replace judgement. Function-critical tolerances — the ones that decide whether the part works — are engineering decisions, and the workflow is built so an engineer always makes them. A tool that claims to fully annotate your parts automatically is overselling; we’d rather tell you that here than after a pilot.
The rule set is only as good as the norm behind it. Encoding your standards is real workshop effort and part of the PoC — where your norms are silent, the rules stay silent too and the item lands with the engineer. For one-off parts and on-request annotation at the designer’s desk, the conversational authoring assistant is the better fit; this engine pays off where families and templates repeat.
Fixed price — fixed scope
PMI Creation PoC
€16,900 fixed
- One part family: datum strategy, callout rules and MBD template encoded with your designers
- Preparation running on your parts, on your infrastructure — engineers complete and release
- Honest measurement: prepared share vs. engineer share, per annotation type
- Verification through our separate compliance toolset — no self-grading
- Recommendation for the next families — based on measured numbers, not hope
4 weeks · encoded rules are yours either way.
For comparison
All-manual annotation
hours per part
- Routine callouts placed by hand, part after part
- Consistency depending on annotator and deadline
- MBD plans postponed because the math never works
Frequently asked questions
What is rule-based PMI creation?
How much of the PMI can the rules really prepare?
How is this different from your PMI authoring assistant?
Can it invent tolerances?
Is the prepared PMI machine-readable?
Where does it run?
Related engines: AI Geometry Manipulation edits the models these rules annotate, and Parametric Part & Variant Generation creates the family members these rules annotate. Overview: AI-Assisted Geometry Creation.
Where rules run out, an agent can place PMI on request in the live NX session. What it covers today, and what stays with engineers, is on NX MBD with AI.
Take one part your team annotated last week. How much of that PMI was judgement — and how much was routine?
Send us that part and the norm chapter that applies — we’ll return it with the routine share prepared as semantic PMI, and an honest list of what stayed with the engineer.
Talk to an EngineerPoC as an NX Sprint Plus, €16,900 · one family · on-premises option