The Manual-CAD Bottleneck in Harness Routing: An Industry Case Study Perspective

This entry is not about NX specifically — the case study does not name NX Routing or any CAD tool by product name — but it's directly relevant context for this knowledge base because it documents, from an industry (automotive powertrain) perspective, exactly the class of problem that NX Routing (and its ECAD counterpart Capital) exist to solve, and where organizations are pushing beyond traditional CAD-based routing workflows entirely. Included here so readers researching NX Routing understand the broader industry conversation about the limits of manual/CAD-native harness routing.

The problem, as described in industry practice

The case study states plainly that "harness routing is still done manually in CAD environments" across much of the industry, and quantifies the scale of the effort involved: in automotive alone, "hundreds of engineers can work on harness design," with "dozens of designers dedicated specifically to routing" on a single program. Modern vehicles can contain "several kilometers of wiring," and the manual, iterative nature of defining and verifying paths inside CAD is described as one of the primary bottlenecks — designers have limited practical ability to explore multiple routing alternatives because each alternative costs real CAD time to lay out and check.

Why this matters for NX Routing users specifically

This is a useful outside-in sanity check on what NX Routing's own toolset is trying to mitigate: features like Quick Path automatic collision-free path generation, batch design-rule checking (bend radius, minimum length, connection compatibility), and automated formboard flattening (see routing-bend-radius-checks.md and routing-electrical-harness-formboard-tips.md in this knowledge base) are direct responses to the same manual-effort problem this case study describes at an industry level. NX Routing automates individual steps of the traditional CAD workflow; the case study describes a further step — generative/AI-driven routing that treats path generation itself as an optimization problem rather than a manual (even if tool-assisted) CAD task.

Results claimed for the AI-driven alternative

According to the case study, an AI/generative-design approach to harness routing:

• Produces constraint-compliant candidate routing paths automatically, "within minutes," versus traditional manual CAD iteration timeframes.
• Enables engineers to computationally evaluate multiple routing alternatives rather than being limited to the few variants that are practical to hand-lay-out.
• Still respects the same real-world constraints routing engineers care about: geometric clearances, bending radius limits, and manufacturability rules — i.e., it is not skipping the engineering checks NX Routing itself performs, but automating the exploration that precedes them.
• Is framed as addressing "one of the last major engineering tasks still performed largely manually" in product development.

Takeaway for this knowledge base

Treat this as directional industry context, not an NX-specific finding: it does not claim NX Routing is inadequate, and does not compare directly against NX. It is useful primarily to calibrate expectations — traditional CAD-based routing (whether in NX or any competing MCAD system) is widely acknowledged in industry as a manual-effort-heavy process even with strong tool support, and generative/AI routing approaches are an active area of development at the periphery of (not a replacement within) tools like NX Routing.

Source: AI-driven wiring harness routing on a powertrain environment, Dessia