23-Ind-A5 Quality Planning, Control, and Assurance · May 2016
Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)
National Exams, May 2016. Closed-book examination. Any five of the six questions constitute a complete paper; all six are answered in full below. Relevant statistical tables (cumulative standard normal distribution, MIL-STD-105E sample-size code letters and master sampling table) are reproduced/applied from the paper's own attached appendices.
Reference texts: Montgomery, Introduction to Statistical Quality Control (8th ed.) — Ch. 1–2 (quality philosophy and management), Ch. 5–6 (variables control charts), Ch. 7 (attributes charts and average run length), Ch. 9 (EWMA/CUSUM and the SPC/EPC interface), Ch. 8 & 13 (designed experiments, Taguchi methods, reliability and life testing), Ch. 15 (acceptance sampling by attributes, MIL-STD-105E and Dodge–Romig plans).
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.
A quality plan is the documented set of specific practices, resources, and sequence of activities relevant to a particular product, project, or contract that translates an organization's general quality-management-system requirements into concrete, verifiable actions for that specific piece of work. Its purpose is threefold: it converts customer requirements and specifications into measurable acceptance criteria and inspection/test points before work begins (so quality is designed and planned in, not inspected in afterward); it assigns clear responsibility, resources, and timing for every quality-related activity (who inspects what, when, and against which standard); and it gives both the organization and the customer objective evidence, ahead of time, of how conformance will be demonstrated — reducing disputes, rework, and the cost of poor quality discovered late.
Quality planning itself operates at three organizational levels that mirror general management planning. Strategic quality planning is long-range (multi-year) and set by top management: it defines the organization's overall quality mission, vision and policy, the competitive role quality will play (e.g. quality as a differentiator versus cost leadership), and the broad resource commitments (technology, training infrastructure, supplier-development programs) needed to get there. Tactical quality planning translates that strategy into medium-term (annual to multi-year) plans at the business-unit or functional level: specific quality objectives (e.g. reduce field-return rate 30% in two years), programs (a Six Sigma deployment, an ISO 9001 certification project), budgets, and organizational structure to support them. Operational quality planning is short-range and product/process-specific: it is the quality plan described above — control plans, inspection and test plans, process-control procedures, and work instructions for a specific product or contract, executed day to day on the shop floor. The three levels must nest: an operational control plan for a single part should be traceable back through the business unit's tactical objectives to the organization's strategic quality policy, or the plan is disconnected from purpose.
Total Quality Management (TQM) is a management philosophy, not a technique: quality is defined by the customer and pursued as the responsibility of every person in the organization, continuously, rather than delegated to an inspection department at the end of the line. Its key elements are: (1) customer focus — quality is defined relative to customer requirements (fitness for use), internal and external; (2) leadership and organization-wide commitment — top management sets policy and visibly participates, quality is not delegated downward alone; (3) continuous improvement (kaizen) — quality is pursued as a never-ending process using structured cycles such as Plan-Do-Check-Act, not a one-time project with an end date; (4) employee involvement and empowerment — the people doing the work are closest to the process problems and are trained and authorized to identify and correct them; (5) process focus and prevention — quality is built into the process design (poka-yoke, statistical process control) rather than inspected in after the fact; (6) fact-based decision making — decisions are driven by data and statistical methods, not opinion; (7) supplier partnerships — suppliers are treated as an extension of the process and developed jointly rather than selected purely on price.
The difference between traditional Total Quality Control (TQC), as originally articulated by Feigenbaum, and modern TQM is one of scope and ownership rather than a wholesale change of subject. TQC was already "total" in the sense of coordinating quality-related activities across departments (design, purchasing, production, marketing), but the coordinating function was still typically vested in a dedicated quality department that set standards and audited compliance — quality remained a specialist function that other departments cooperated with. Modern TQM removes that intermediary: quality becomes an explicit, personal responsibility of every employee and especially of line management and top leadership, embedded in daily decision-making and culture rather than administered by a separate department; it also broadens the definition of "customer" to include internal customers (the next process step) and places much stronger emphasis on continuous improvement, employee empowerment, and organizational culture change as the improvement mechanism, rather than primarily audit-and-standard enforcement.
A quality circle is a small group (typically 6–12) of employees, usually from the same work area, who voluntarily meet on a regular schedule to identify, analyze, and propose solutions to quality and productivity problems within their own area, using basic problem-solving tools (cause-and-effect diagrams, Pareto charts, brainstorming) with training and facilitation support from management. Quality circles are a concrete implementation of the employee-involvement element of TQM: they give front-line workers a structured channel to apply their process knowledge to improvement, and management retains the role of implementing feasible recommendations and providing recognition, rather than dictating the solution.
The traditional (goalpost) definition of quality treats any unit that falls within the specification limits ($LSL$ to $USL$) as equally "good," and any unit outside those limits as equally "bad," regardless of how close it sits to the limit — the implied cost-of-quality function is a step function: zero loss anywhere inside spec, a fixed loss (scrap or rework cost) the instant a unit crosses the limit. Taguchi's definition rejects this discontinuity: quality is the loss a product imposes on society (the customer, and eventually the wider economy) from the time it is shipped, and that loss grows continuously as the characteristic departs from its ideal target value $\tau$ — a unit measuring just inside spec is treated as meaningfully worse than a unit sitting exactly on target, not as equivalent to it. Taguchi models this with the quadratic loss function $$L(y)=k(y-\tau)^2,$$ where $y$ is the measured characteristic, $\tau$ the target, and $k$ a cost constant fixed by relating a known loss at the specification limit ($k=A_0/\Delta_0^2$, where $A_0$ is the loss incurred at a deviation $\Delta_0$ from target, e.g. the cost of scrap or warranty repair at the edge of spec). The expected loss over a population of product is then $E[L]=k\big(\sigma^2+(\bar y-\tau)^2\big)$, which explicitly rewards BOTH reducing variance around the target AND centring the process mean exactly on target — a process that is perfectly centred but has some spread, and a process that is off-centre but has zero spread, are penalized on an equal, additive footing, whereas the traditional goalpost view would treat the off-centre-but-tight process as "perfect" so long as it stayed inside spec.
Taguchi's approach to quality improvement follows from this philosophy: since loss accrues continuously with deviation from target, the improvement objective is to make the product/process performance INSENSITIVE (robust) to sources of variation — manufacturing variation, environmental conditions, component aging, and customer usage variation ("noise factors") — rather than to try to eliminate every noise source individually, which is usually far more expensive. This is achieved through his three sequential design stages (system design, parameter design, tolerance design; detailed further in Question 5(a)), with the emphasis deliberately placed on the middle stage, parameter design: choosing the nominal levels of the CONTROLLABLE design parameters, using designed experiments (orthogonal arrays, signal-to-noise ratios), so that the product's performance varies as little as possible in the face of noise — achieving robustness through clever, low-cost design choices before ever resorting to the expensive last resort of tightening manufacturing tolerances.
Quality certification (third-party audit and registration of an organization's quality management system against a recognized standard, e.g. ISO 9001) serves as an independent, externally credible signal of quality-system capability: it reduces the need for every customer to individually audit every supplier, lowers transaction costs and buyer risk in the supply chain, and gives the certified organization a documented, externally verified management system that supports continuous improvement and market access (many customers and public-sector contracts require it as a condition of doing business). Quality prizes (e.g. the Malcolm Baldrige National Quality Award, the Deming Prize, the EFQM Excellence Award) serve a related but distinct purpose: rather than a pass/fail conformance audit, they benchmark an organization holistically against a detailed excellence framework (leadership, strategic planning, customer focus, workforce, process management, and RESULTS), publicly recognize sustained excellence, and — importantly — publish the assessment criteria themselves as a widely-used self-assessment and improvement framework, so the prize's educational/diffusion effect on industry as a whole is often considered as valuable as the award itself.
The ISO 9000 family is structured as a small set of complementary standards rather than one document: ISO 9000 gives the fundamental concepts, principles, and vocabulary (defining terms such as "quality," "conformity," "process approach" consistently across the family); ISO 9001 is the only standard in the family against which an organization can be CERTIFIED — it specifies the actual requirements for a quality management system (customer focus, leadership, planning, support/resources, operation, performance evaluation, and improvement, organized around the Plan-Do-Check-Act cycle and risk-based thinking) that a third-party auditor assesses; ISO 9004 provides guidance (not a certifiable requirement) for organizations that want to go beyond ISO 9001's baseline toward sustained organizational success; and a wider set of ISO 19011-type documents provide guidance on auditing management systems generally. The family is deliberately generic (applicable to any size or type of organization, any product or service) and process-based: it requires the organization to identify its key processes, their interactions, and the metrics used to manage and improve them, rather than prescribing a specific quality technique.