Design for Manufacturing (DFM) is one of the most important feedback loops in product development. It is where engineering intent meets the realities of tooling, process capability, cost, quality, assembly, and production.

But the way DFM is performed is still often centered on the individual CAD model. A designer creates or modifies geometry, a manufacturing or quality check is run, issues are reviewed, the model is updated, and the cycle repeats. That workflow can work well at part level, but it becomes harder to scale when the real question is not about one model. It is about an assembly, a project, a product family, or thousands of components across the product lifecycle.

At Nora IPLM, we have been exploring an internal initiative that could change that relationship. For now, we call it Project X.

The idea is simple to describe, even if the engineering behind it is not: make valuable engineering information that is usually locked inside CAD files accessible as connected product intelligence inside PLM, then make that information queryable and actionable with AI.

DFM Today: Valuable Checks, but Often CAD-Centric

Modern CAD environments already provide powerful mechanisms for checking design quality and enforcing engineering standards. Examples include Q-Checker in the CATIA ecosystem, NX Check-Mate, and Creo ModelCHECK. These tools can evaluate models against rules, company standards, modeling practices, or geometry-quality criteria, and they help engineers catch problems earlier in the design process.

That is valuable. Project X is not based on the idea that CAD checking is missing. The opportunity is different: what if the information produced by geometry and design analysis could participate directly in the broader product lifecycle?

Today, many analysis workflows still depend on CAD-specific applications, CAD-specific processing, or model-by-model access. Even when batch checking is possible, the engineering context often remains separated from the rest of the lifecycle information that PLM already understands: BOMs, revisions, requirements, suppliers, changes, manufacturing structures, documents, and release status.

The Missing Link Between CAD Geometry and Product Context

PLM is already good at answering product-context questions such as:

• Which revision is released?
• Where is this component used?
• Which BOMs are affected by this change?
• Which supplier is connected to this item?
• Which documents and requirements belong to this product?
• Which manufacturing structure implements this engineering definition?

CAD, meanwhile, contains another layer of highly valuable information: geometry, design features, dimensions, wall thicknesses, holes, fillets, draft angles, undercuts, tolerances, and other characteristics that directly affect manufacturability.

Project X explores what happens when those two worlds stop behaving like separate information systems.

What Product-Wide DFM Questions Could Look Like

Imagine being inside the PLM environment and asking questions such as:

• “Check every molded part in this project for insufficient draft angles.”
• “Find all fillets smaller than 2 mm.”
• “Give me a table of every hole in this product, including diameter, depth, and type.”
• “Which parts contain features that may create manufacturing problems?”
• “Which released components contain the same geometry issue?”
• “Which product variants would be affected if this manufacturing rule changed?”
• “Show the manufacturing-risk findings for the parts included in this ECO.”

The shift here is important. The user would not be thinking in terms of opening files one by one. The user would be asking a product question, and the system would use the connected lifecycle and geometry context to determine what must be examined.

From Part-Level Checking to Product-Level Manufacturing Intelligence

Traditional DFM analysis usually starts with a known model. Product-wide DFM starts with a question.

This could make DFM less of a one-time gate and more of a continuous source of product intelligence. Geometry findings could be evaluated alongside lifecycle state, supplier context, product configuration, change history, BOM position, and manufacturing process information.

Why AI Changes the Question

Rule-based checking is very good at answering: “Does this geometry violate the rule?”

AI creates the possibility of moving one level higher, toward questions such as:

• “Which of these issues should we worry about first?”
• “Why is this feature a manufacturing risk?”
• “Which products are most exposed to this problem?”
• “What change would reduce the risk without disrupting the rest of the product definition?”
• “Which findings are likely to matter for this manufacturing process, supplier, or product configuration?”

That distinction matters because engineering teams rarely struggle from a lack of findings. They struggle with prioritization, context, and the amount of manual work required to understand what a finding means for the product.

Nora Prima AI is already positioned around using connected lifecycle data to analyze impact, explain risk, and support decisions across BOMs, changes, documents, suppliers, and workflows. Project X explores what could happen if geometry and manufacturing characteristics became another layer of that connected context.

Also Read: AI Change Impact Simulation: See the Impact Before the Change Moves Forward

A Practical Example: Molded Components Across an Entire Product

Consider a product with 800 managed components, including 120 molded plastic parts spread across several assemblies and variants.

A manufacturing engineer wants to know whether any released or pending components contain draft angles below the organization’s preferred threshold. In a file-centric process, the team may need to identify the relevant parts, retrieve the models, run checks, collect the results, and then reconcile those results back to BOM and revision context.

A product-centered approach could begin differently:

1. Identify the components in scope from PLM product structures and configuration context.
2. Access the geometry intelligence associated with those components.
3. Evaluate the applicable draft-angle rule across the selected population.
4. Return a structured result tied to item number, revision, BOM location, lifecycle state, and related change context.
5. Use AI to explain which findings appear most important and why.
6. Create follow-up actions, change work, or review tasks against the affected product records.

The value is not only faster checking. The result becomes connected to the product lifecycle, which means it can participate in engineering review, manufacturing readiness, change management, supplier collaboration, and release decisions.

What This Could Mean for Different Teams

Could AI Eventually Move From Diagnosis to Action?

The most provocative possibility is not simply asking AI to identify a problem. It is asking whether AI could eventually help propose, or even execute, a controlled dimensional change that resolves it.

That is a much harder problem. A dimensional change is not just geometry. It can affect fit, function, tolerance stack-up, interfaces, requirements, simulation assumptions, drawings, tooling, suppliers, cost, configurations, and released product structures.

For that reason, any move from “find the issue” to “change the design” would require strong engineering guardrails, clear ownership, traceable approvals, and validation. The interesting part of a PLM-connected approach is that those dependencies already live in the same lifecycle environment.

Also Read: How CAD and PLM Integration Transforms Product Teams

DFM as Part of the Digital Thread

The broader opportunity is not only DFM. It is the digital thread.

A digital thread becomes more useful when it contains more than files and metadata. If engineering geometry can be represented as accessible product intelligence, then manufacturing questions can be connected directly to:

• the BOM and product structure
• the exact item and revision
• the engineering change that introduced the geometry
• the requirement the feature is intended to satisfy
• the supplier or manufacturing context
• the product variant and effectivity
• the release and lifecycle state

That is where the line between CAD management and lifecycle intelligence starts to change. CAD remains the authoring environment for engineering design, but the knowledge embedded in the model can become useful far beyond the person who created it.

Project X Is Not About Replacing CAD

The goal is not to turn PLM into another CAD system. Designers still need professional CAD environments for modeling, detailed engineering, simulation, and design authoring.

The opportunity is to make CAD-derived intelligence more available to the rest of the lifecycle so teams can search, analyze, compare, prioritize, and act without requiring every question to begin by manually opening a CAD model.

That could help shift CAD from being only a source file managed by PLM to being a richer source of product knowledge connected to PLM.

Where We Are Going With Project X

We are still exploring what Project X can become. DFM is one of the areas that makes the opportunity easiest to see because geometry, manufacturing knowledge, product context, and AI all have a clear reason to meet.

The question we are interested in is no longer only:

“Can this CAD model pass a manufacturing check?”

It is becoming:

“What can we understand about the manufacturability of the entire product, right now?”

That is a much more interesting problem, and it is one of the doors Project X is opening for us.

More soon.

Frequently Asked Questions

What is Design for Manufacturing (DFM)?

Design for Manufacturing is the practice of evaluating and improving a product design so it can be manufactured reliably, efficiently, and at the required quality and cost. DFM considerations can include geometry, material, process capability, assembly, tooling, tolerances, and supplier constraints.

How is DFM usually performed in CAD?

CAD and engineering-quality tools can evaluate models against geometry, modeling, or company-specific rules. Examples include CATIA Q-Checker, NX Check-Mate, and Creo ModelCHECK. The exact checks and workflows depend on the CAD environment and how each organization configures its standards.

What would PLM add to DFM analysis?

PLM adds lifecycle context. A geometry finding can be connected to the item, revision, BOM, product configuration, requirement, supplier, engineering change, manufacturing structure, and release state that give the finding business and engineering meaning.

How could AI be used in DFM?

AI could help summarize findings, explain why a feature may create manufacturing risk, prioritize issues across many components, and connect geometry findings with broader product context. Any automated design-changing capability would require engineering controls and validation.

Does Project X replace CAD software?

No. The concept is about making CAD-derived engineering intelligence accessible within the wider product lifecycle. CAD remains the primary environment for detailed design authoring.

Is Project X available today in Nora IPLM?

Project X is described here as an internal exploration. The examples in this article illustrate the direction being investigated and should not be interpreted as a commitment that every capability is currently available or will be released in the exact form described.