Quality assurance and process control
Quality improves when the definition is operational, the measurement is trustworthy and variation is separated from noise. This path moves from customer requirement to control chart to standard work, without treating every wobble as a crisis.
6 steps · 5 cited sources · free during beta
How do we define quality, stabilise the process and hold the gain?
The path, step by step
-
Define quality from the customer and the cost of failure
Name the characteristic that matters, write an operational definition of a defect, and estimate what poor quality costs in rework, delay and lost custom.
Why it worksQuality is conformance to a requirement, not intensity of effort. Without an operational definition, two people can disagree about whether the same unit passed.
EvidenceDeming (1986) · Montgomery (2019)
CheckpointOne agreed definition of a defect, one target characteristic, and a cost-of-poor-quality estimate.
Common failureLaunching a quality initiative with a slogan and no measurable requirement.
-
Make the measurement trustworthy
Check calibration, repeatability and reproducibility. Test whether two trained people measuring the same item reach the same conclusion, and fix the method before collecting improvement data.
Why it worksStatistical process control assumes the measurement system can see the process variation. If gauge error is large relative to the process, the chart is mostly measuring the gauge.
EvidenceMontgomery (2019) · Deming (1986)
CheckpointMeasurement error is small relative to the variation the process needs to detect.
Common failureBuilding a control chart on a measure nobody has validated.
-
Separate common cause from special cause
Plot the measure in time order, establish limits from a stable baseline, and annotate every intervention. React to signals, not to single points inside the limits.
Why it worksControl and run charts distinguish routine variation from a signal that something changed. Reacting to common-cause noise adds variation and makes the process harder to learn from.
EvidenceMontgomery (2019) · Perla et al. (2011)
CheckpointA time-ordered chart with limits, annotations and a documented response rule.
Common failureTreating every uptick as a special cause and adjusting the process continuously.
-
Find the root cause with a designed test
Use the chart and process knowledge to form a causal hypothesis, then change one factor deliberately and measure the effect. Move from symptom to mechanism before standardising.
Why it worksDesigned experiments and disciplined problem solving identify which factor actually moves the outcome. Brainstorming without a test produces plausible causes that survive because nothing disproves them.
EvidenceMontgomery (2019) · Spear & Bowen (1999)
CheckpointA changed factor reproduces the effect, or the hypothesis is rejected and replaced.
Common failureFixing the most visible symptom and declaring the root cause found.
-
Standardise and mistake-proof the new method
Update the standard work, make the correct action the easy one, train against the standard and audit adherence. Design out the error where possible instead of asking for more care.
Why it worksToyota-style process design makes deviations visible at the point of work, and lean evidence shows that quality practices work as a bundle with standard work and employee involvement.
EvidenceSpear & Bowen (1999) · Shah & Ward (2003) · Deming (1986)
CheckpointThe new standard is in use, adherence is audited, and first-pass yield has moved.
Common failurePublishing a procedure and assuming it changed the work.
-
Hold the gain and escalate systemic issues
Keep the chart running at the new level, review signals on a fixed cadence, and escalate problems that require design, capital or supplier change rather than local effort.
Why it worksImprovement decays when the measurement stops or when local teams are asked to solve problems owned by the system. Deming's point is that most variation is built into the process, not the person.
EvidenceDeming (1986) · Montgomery (2019)
CheckpointThe process remains stable at the improved level for at least three review cycles.
Common failureClosing the project when the chart looks good for one week.
The evidence base
Each step above cites hand-selected research. These are the studies behind the path.
- Deming (1986) — Out of the Crisis
- Montgomery (2019) — Introduction to Statistical Quality Control
- Perla et al. (2011) — The run chart: A simple analytical tool for learning from variation in healthcare processes
- Spear & Bowen (1999) — Decoding the DNA of the Toyota Production System
- Shah & Ward (2003) — Lean manufacturing: context, practice bundles, and performance
More from the library
Further reading from Portolan's research library, matched by subject. These papers are not proof that any step works; check each checkpoint against your own operation.
- Psarommatis et al. (2021) — Zero-defect manufacturing the approach for higher manufacturing sustainability in the era of industry 4.0: a position paper
- Zu et al. (2008) — The evolving theory of quality management: The role of Six Sigma
- Lee & Li (2018) — Supplier Quality Management: Investment, Inspection, and Incentives
- Saraph et al. (1989) — An Instrument for Measuring the Critical Factors of Quality Management
- Ahire & Dreyfus (2000) — The impact of design management and process management on quality: an empirical investigation
- Flynn et al. (1994) — A framework for quality management research and an associated measurement instrument
- Flynn et al. (1995) — The Impact of Quality Management Practices on Performance and Competitive Advantage
- Abrahams et al. (2015) — An Integrated Text Analytic Framework for Product Defect Discovery
- Fine (1986) — Quality Improvement and Learning in Productive Systems
- Balachandran & Radhakrishnan (2005) — Quality Implications of Warranties in a Supply Chain
Other playbooks
- Demand forecasting — How should we forecast demand so that plans, staffing and stock follow reality?
- Capacity, flow and scheduling — Where is work waiting, what is the real constraint, and how should we release and staff work?
- Inventory and replenishment — How much stock should we hold, where, and what should trigger a replenishment?
- Supply chain and supplier resilience — Where is the network exposed, and which capability or buffer actually reduces the exposure?
- Service operations and customer experience — Where does the customer experience break, and what should we change in the service system?
- Maintenance and asset reliability — Which assets matter, how do they fail, and which maintenance tasks are actually justified?
- Continuous improvement and change adoption — How do we make improvement routine, and how do we get a change to stick?
- Measurement and operating cadence — Which measures should we run the operation on, and how should we review them?
- Incident response and on-call — How should we handle incidents and on-call so the service recovers quickly without burning out the team?