23-Ind-A5 Quality Planning, Control, and Assurance · December 2019
Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)
Reference texts: Montgomery, Introduction to Statistical Quality Control (8th ed.) — cost/philosophy of quality, control charts for variables and attributes, CUSUM, acceptance sampling (MIL-STD-105E, Dodge-Romig); Montgomery, Design and Analysis of Experiments (9th ed.) — fractional factorial designs, aliasing, robust (Taguchi) parameter design; ISO 9001:2015 (successor to ISO 9000:2000) — quality management system certification.
Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.
Specification limits (USL/LSL) are engineering requirements set by the designer or customer, based on how the part must function — they describe what the process is allowed to produce and have nothing to do with how the process actually behaves. Natural tolerance limits are a statement about the process itself: for a normally distributed quality characteristic, they are conventionally set at $\mu\pm3\sigma$, the range within which nearly all (99.73%) of the process output actually falls when the process is operating in a stable, in-control state. Control limits ($\mu\pm3\sigma_{\bar X}$ on an $\bar X$ chart, or the analogous $3\sigma$ limits on any other chart) are statistical limits on the sampling statistic plotted on a control chart — they describe the amount of variation expected in that statistic (e.g. a sample average) from sample to sample when only common-cause variation is present, and are used to detect the onset of assignable causes.
The three limits answer three different questions and there is no necessary numerical relationship among them. Specification limits represent a requirement; natural tolerance limits represent what the process can do; control limits represent how much a sample statistic should vary if nothing has changed. A process can be in statistical control (all points inside its control limits) while still producing nonconforming product, if its natural tolerance limits are wider than the specification limits — control and capability are distinct questions.
ISO 9000 registration (now ISO 9001) exists to give a customer independent, third-party assurance that a supplier's quality management system is capable of consistently delivering conforming product, without the customer having to audit every supplier itself. It converts quality assurance from a bilateral, ad-hoc relationship into a documented system verified against a common international standard, which lowers transaction costs across a supply chain and is frequently a contractual precondition for doing business (particularly in automotive, aerospace, and government procurement).
Structurally, the standard specifies requirements for a documented quality management system built on management responsibility, resource management, product/service realization (the operational processes that create the product), and measurement/analysis/improvement, all closed by a continual-improvement loop. Registration is granted and maintained through periodic third-party audits against these clauses.
The ISO 9000:2000 revision was the most significant change to the standard's history. It replaced the earlier three-tier, procedure-heavy standard (ISO 9001/9002/9003, essentially a checklist of 20 required procedures) with a single standard (ISO 9001) built around the process approach — managing the organization as a set of interrelated processes with defined inputs, outputs, and owners, rather than as a stack of isolated procedures. It explicitly adopted the Plan-Do-Check-Act cycle as its organizing structure, added an explicit requirement to measure customer satisfaction as an output of the system, and required continual improvement as an ongoing system objective rather than a one-time compliance exercise. It was also written to be compatible in structure with ISO 14001 (environmental management), easing integrated audits.
Quality prizes — the Deming Prize, the Malcolm Baldrige National Quality Award, and similar national awards — exist to publicly recognize and benchmark organizational excellence in quality management, going beyond minimum-compliance certification. Their objective is to drive continuous improvement by rewarding demonstrated business results (not just documented procedures), to disseminate best practice across industry through the self-assessment criteria and site visits the award process requires, and to raise the visibility and prestige of quality management as a source of competitive advantage at the executive level, which registration to a standard alone does not achieve.
A company whose focus is narrowly on short-term profit maximization typically achieves it, in the near term, by cutting the costs most visible on this period's income statement — prevention and appraisal spending (training, process control, inspection) — because failure costs from that decision are deferred and often externalized onto the customer. Deming's central argument is that this trade is a false economy: cutting prevention/appraisal spend increases variation and the escaping-defect rate, which (per the "1-10-100" cost-of-quality escalation) raises internal and, more damagingly, external failure costs faster than the near-term savings, while also degrading the product's actual and perceived quality.
The longer-term consequence is a self-reinforcing decline: lower quality erodes customer satisfaction and repeat business, which erodes market share; a shrinking, price-sensitive residual customer base and lower unit volumes raise unit costs and further pressure margins, which drives more cost-cutting and a further quality decline (Deming's "chain reaction" running in reverse). Competitiveness suffers because competitors who instead invest in quality improvement capture the customers being lost, and because the firm's reputation for quality — once damaged — is expensive and slow to rebuild. The correct long-term reading is the opposite direction of causality: improving quality lowers total cost (fewer defects, less rework, less scrap) and grows market share, which is what sustains profit; profit is the outcome of a quality focus, not a target to be pursued by cutting quality investment directly.
The main difference is scope and ownership. Total Quality Control (TQC), as originally defined by Feigenbaum, is fundamentally a manufacturing/engineering discipline: quality is coordinated across the design-production-inspection cycle and is the responsibility of a dedicated quality function/department that plans, controls, and audits it, with other departments contributing inputs to that function. Total Quality Management (TQM) broadens this into a company-wide management philosophy: quality is the responsibility of every employee and every function (not a delegated department), is driven from the top by senior leadership, and is treated as a strategic, customer-focused, continuously-improving management system rather than a technical control activity bolted onto manufacturing.
The key elements of TQM are: customer focus (quality is defined by the customer, both external and internal); top-management leadership and commitment, since a culture change of this scope cannot be delegated; total employee involvement and empowerment, including cross-functional teams; a process-centered view of the organization, with process ownership and process improvement (not just outcome inspection); an integrated system linking business functions horizontally rather than managing them as silos; a strategic and systematic approach, with quality built into strategic planning; continual improvement (kaizen) as a permanent objective rather than a project with an end date; fact-based decision making, using data and statistical methods (SPC, DOE) rather than opinion; and open, two-way communication throughout the organization to sustain the culture.
