Skip to main content
JobCannon
All skills
✓

Quality Inspection and SPC

Inspect parts and monitor trends — control charts, Cpk, defect prevention

⬢ TIER 1Industry
💰
Medium
Salary impact
⏱
4 months
Time to learn
📊
Medium
Difficulty
👥
—
Careers
At a glance

Quality Inspection and Statistical Process Control (SPC) is the disciplined measurement and analysis of manufacturing processes to prevent bad parts from shipping. Inspectors measure parts using precision tools (micrometers, CMMs, gauges), record data, and plot it on control charts. SPC tools (control limits, Cpk capability indices, trend analysis) predict when a process is drifting out of spec before actual scrap occurs. Work spans incoming inspection (supplier parts), in-process inspection (during manufacturing), final inspection (before shipping), and source inspection (at supplier facility). Career path: Quality Inspector (entry, $32-45k) to Quality Technician ($45-60k) to Quality Engineer ($65-100k) over 5-10 years. Quality is non-negotiable in regulated industries (automotive, aerospace, medical, food); poor quality costs millions in recalls and liability. ASQ and NIMS certifications are industry standard.

What is Quality Inspection and SPC

Quality Inspection and SPC is the discipline of systematically measuring parts, analyzing data, and predicting when a process is drifting out of control before scrap happens. It combines hands-on inspection skill with statistical rigor to ensure every part meets spec. Quality Inspection is the measurement and evaluation of parts against blueprints and specifications. SPC (Statistical Process Control) is the use of statistics and control charts to monitor process stability and predict failures. An inspector measures parts using precision tools (micrometers, CMMs, calipers), records the data, plots it on a control chart, and analyzes trends. Control limits (derived from process statistics) are narrower than spec limits; if a measurement exceeds the control limit, the process is drifting and needs correction before the next part goes out of spec. Cpk (capability index) is a single metric that summarizes process fitness: can you reliably hold the tolerance? SPC is the early-warning system; without it, you only catch scrap after parts are made.

🔧 TOOLS & ECOSYSTEM
Measurement Instruments (Micrometers, Calipers, Gauges, CMM)Control Chart Software (Excel, JMP, Minitab, SPC Pro)Inspection Documentation (Check Sheets, Defect Logs)GD&T Interpretation ChartsSample Size CalculatorsCpk and Process Capability Analysis SoftwareFirst-Piece Inspection KitsAudit ChecklistsData Collection Devices (Tablets, Scanners)Statistical Analysis SoftwareQuality Management System (QMS) DatabaseTraceability and Lot Tracking

📋 Before you start

Precision Measurement And MetrologyBlueprint And Gd And T Reading

💰 Salary by region

RegionJuniorMidSenior
🇺🇸USA$32k$46k$62k
🇬🇧UK£20k£29k£39k
🇪🇺EU€24k€34k€46k
🇨🇦CANADAC$36kC$53kC$71k

🎓 Certifications

ASQ Certified Quality Inspector (CQI)→NIMS Quality Inspection and SPC Certification→ASQ Six Sigma Green Belt (for advanced roles)→MSSC Certified Quality Inspector→

⚖ Compare with

Precision Measurement And MetrologyLean Manufacturing And KaizenBlueprint And Gd And T Reading

❓ FAQ

▶What are control limits and how do they differ from specification limits?
Specification limits (upper and lower spec limit, USL and LSL) are the customer requirements: a part dimension must be between 2.000 and 2.010 inches. Control limits (upper and lower control limit, UCL and LCL) are statistical boundaries derived from your process performance: they're typically 3 standard deviations from the process mean. If you measure 25 parts and calculate the process average (2.004 inches) and standard deviation (0.002 inches), the control limits are 1.998 to 2.010 inches. Control limits are tighter than spec limits and predict when the process is drifting: if a measurement exceeds the control limit, the process is out of statistical control (and will soon produce out-of-spec parts) even if that single part is within spec. Acting on a control-chart signal (stopping production to investigate) prevents scrap.
▶What is Cpk and what does it tell me?
Cpk (Process Capability Index) is a single number that summarizes how well your process is performing relative to tolerance. Cpk = 1.0 means your observed variation exactly fills the tolerance (you're on the edge of scrap); Cpk = 1.33 means you have 25% margin (good); Cpk = 2.0 means you have huge margin (excellent). Formula: Cpk = (USL – Process Mean) / (3 × Std Dev). A customer demanding Cpk ≥ 1.33 is saying: 'I need confidence you can hold tolerance consistently.' If your Cpk is 0.8, you're making out-of-spec parts; if Cpk is 1.67, you're golden. Cpk is calculated from sample data (measure 25-50 parts), so it's not perfect, but it's an industry-standard shorthand for process fitness.
▶What is the difference between attribute inspection and variables inspection?
Attribute inspection: you check if a part passes or fails (go/no-go). Example: Does the hole have the right number of threads? Yes/No. Does the surface have any scratches? Yes/No. Fast, binary, but low information. Variables inspection: you measure the actual value (a dimension: 2.005 inches). Slower, but much richer data (you know how far off-spec the part is). SPC requires variables data (you need actual measurements to calculate Cpk and control limits). Attribute data (yes/no) is useful for simple go/no-go gauges and final acceptance, but it doesn't predict process drift.
▶How often should I measure samples and how many samples?
Sample frequency and size depend on process stability and risk: A stable, proven process might be measured once per shift (4 parts per sample, 25 samples total = 100 parts sampled out of 10,000 per day). A new process or a critical dimension might be measured every hour (4 parts per sample, 40+ samples per shift). A critical aerospace feature might be 100% measured (every part) to zero risk. The rule of thumb: the smaller the batch, the higher the sampling percentage (a batch of 50 parts might require 10-20 measured). Statistical formulas (AQL acceptance level, ANSI/ASQC Z1.4 tables) define sample sizes for different risk levels. Start conservative (higher sampling) and relax as process capability improves.
▶What is a control chart and how do I interpret it?
A control chart is a graph plotting sample measurements over time (X-axis = sample number, Y-axis = dimension). A horizontal line shows the process mean; two horizontal lines show the upper and lower control limits. If all points fall between the control limits with random scatter, the process is 'in control' (predictable, stable). If a point exceeds the limit or you see a trend (points climbing up, down, or oscillating), the process is 'out of control' (drifting, and action is needed). Rules: out-of-control: one point beyond 3-sigma (control limit), two of three points beyond 2-sigma, four of five points trending in one direction, eight points on the same side of the mean. Operators and supervisors watch control charts daily; when out-of-control signals occur, they stop production and investigate: tool wear, temperature drift, spindle issue, whatever.
▶What is first-article inspection (FAI) and when is it required?
First-article inspection (FAI) is the comprehensive inspection of the very first part from a new tool, die, or process, before production starts. It's mandatory in aerospace and automotive (OEM contracts require it), and best practice everywhere. FAI includes: measurement of every critical dimension (often 100%), verification of GD&T, surface inspection, dimensional documentation, sometimes metallurgical testing (hardness, grain structure). If any FAI dimension is out of spec, the part is rejected, the process is stopped, and the root cause is investigated before production resumes. A good FAI catches problems before 1,000 scrap parts are made. FAI documentation becomes part of the contract file (traceability: if a customer ever has a problem, you can prove the first part was good).
▶What is a defect and how is it categorized?
A defect is any non-conformance to specification: wrong dimension, scratch, missing feature, misalignment, discoloration, anything. Defects are often categorized by severity: Critical = part won't function or is unsafe (missing thread, wrong material, cracked weld); Major = part functions but poorly (wrong color, rough finish, slow response); Minor = cosmetic or negligible (light scratch, label slightly off). A critical defect causes 100% scrap; a major defect might be rework (re-machine, re-paint) or accept-as-is if the customer approves; a minor defect is usually accepted. The categorization is defined by the customer or your quality policy. Tracking defects by type and category reveals patterns: if 20% of parts have the same major defect, the process has a systematic issue (tool wear, setup error, temperature) that needs fixing.
🎯

Not sure this skill is for you?

Take a 10-min Career Match — we'll suggest the right tracks.

Find my best-fit skills →

Find your ideal career path

Skill-based matching across 2,536 careers. Free, ~2 minutes.

Take Career Match — free →
Browse more:All skillsCareer libraryPersonality testsMethodology