Quality and Inspection

Floor Issue Density: A Switchgear Assembly KPI Worth Stealing

Counting issues on a switchgear build tells you almost nothing, because panels differ enormously in content. Normalize by section and the same data becomes a genuine quality signal.

Switchgear assembly is high-mix, engineer-to-order work where two jobs bearing the same description can differ several-fold in content. That makes raw defect counts nearly useless as a quality measure: a job with more issues may simply be a larger job. Floor issue density, issues normalized by a unit of content, fixes this and turns an unusable count into something you can trend and compare.

Choose the denominator carefully

Take a job logging 34 floor issues across 12 panels comprising 47 vertical sections. Per panel that is 2.8 issues; per section it is 0.72. The second is the more comparable figure, because a panel is a commercial unit of wildly varying size while a section is closer to a unit of assembly content. Better still where the data exists: issues per thousand terminations, which normalizes for the wiring density that actually generates most defects.

• Issues per panel: 2.8
• Issues per section: 0.72
• Raw count, which compares to nothing: 34

Segment by discovery point, or the metric misleads

An issue found during assembly is cheap. The same issue found at factory acceptance testing is considerably more expensive, and found at site commissioning it is expensive enough to threaten the project schedule. A single density figure that mixes all three hides the trend that matters. Report density by discovery point, and improvement becomes visible as issues migrating earlier even when the total count has not moved, which is exactly what a maturing quality process looks like.

Total issue count going flat while issues move from site back to the assembly bench is a large improvement that a single metric would have reported as nothing happening.

The causes cluster tightly

In most switchgear shops the recurring sources are

• Drawing errors or late engineering changes reaching the floor after assembly started, which is a process problem rather than a workmanship one.
• Incorrect or missing components, driven by kitting accuracy and by bills of material that were not updated with the design.
• Termination workmanship, particularly torque and labelling, which is the classic case for standard work and torque verification.
• Test failures traceable to wiring errors, which is where issues per thousand terminations earns its place as the denominator.

Why this metric transfers

Nothing about issue density is specific to switchgear. Any engineer-to-order assembly business with variable content and a defect log can use it, and most such businesses are currently reporting raw counts that move with order size rather than with quality. The work is choosing a denominator that tracks assembly content, which usually already exists in the routing or the bill of material.

Use the switchgear assembly hours calculator to relate content to labor and inspection load. Model the build

Published 2026-08-08.