23-Ind-A5 Quality Planning, Control, and Assurance · Undated paper
Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)
National Exams, May 2019. 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, factors for constructing variables control charts, MIL-STD-105E Table I) 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, cost of quality, Six Sigma/TQM, ISO 9000/TS16949), Ch. 4–6 (magnificent seven SPC tools, process capability, X̄-R and attributes control charts), Ch. 9 (average run length), Ch. 13–14 (designed experiments/factorial designs, acceptance sampling and MIL-STD-105E).
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.
The traditional (conformance-based) philosophy defines quality as meeting engineering specifications: a unit is good if every characteristic falls inside its tolerance band and bad otherwise. This "goal-post" view treats a dimension just inside a limit identically to one at nominal, and the implied cost function is a step — $0 loss inside the spec, a fixed scrap/rework cost the instant a limit is crossed.
Juran's philosophy defines quality as "fitness for use," judged from the customer's point of view across several dimensions (design quality, conformance quality, availability, reliability, field service). Juran frames quality management as a trilogy — quality planning (identify customers and their needs, design a product/process to meet them), quality control (operate the process to stay on target, using feedback and corrective action), and quality improvement (a project-by-project breakthrough process to raise performance beyond the historical level) — and famously argues that a large share of quality problems (roughly 80%) are management-controllable "system" causes rather than worker-controllable "special" causes.
Taguchi's philosophy reframes quality as "the loss imparted to society from the time a product is shipped," a societal-cost view including cost to the customer (poor performance, repairs) and to the producer (warranty, reputation). Its central device is the quadratic loss function $L(y)=k(y-T)^2$ (the leading nonzero term of a Taylor expansion of any smooth loss about the target $T$), which replaces the goal-post step function with a continuous cost that grows with any deviation from target, even while still inside the tolerance band.
Implications for product and process design. The traditional view only drives designers to keep the process inside tolerance, with no incentive to reduce variance further once inside spec ("just barely conforming" production is treated as free). Juran's trilogy explicitly separates the up-front planning activity (choosing targets and designing a capable process) from ongoing control and later improvement, so the design phase is responsible for building in capability, not just meeting a spec on paper. Taguchi's loss function goes furthest: it drives on-target, minimum-variance design — parameter design chooses factor settings that minimize $\sigma$ around $T$ (Q5(a) below), and tolerance design allocates tighter, costlier tolerances only where the economic loss coefficient $k$ is largest, directly connecting statistical variance reduction to a dollar-valued design objective instead of a pass/fail gate.
Six Sigma is a prescriptive, project-based, statistically rigorous methodology that targets a specific numerical performance goal — a process capability corresponding to $\pm6\sigma$ between the mean and the nearest specification limit, i.e. $\approx3.4$ defects per million opportunities once a $1.5\sigma$ long-term mean shift is allowed for. It is executed through the structured DMAIC roadmap: Define the project scope and customer requirements (CTQs); Measure the current process performance and validate the measurement system; Analyze the data to identify root causes of variation/defects; Improve the process by selecting and implementing a solution (often via designed experiments); Control the improved process with control charts, mistake-proofing, and a documented response plan so gains do not erode. Projects are run by trained Green/Black Belts, are financially justified, and are executed on a discrete project-by-project basis.
TQC (Total Quality Control), the Feigenbaum-era predecessor, is an organization-wide system for integrating quality development, maintenance, and improvement effort across every department (not inspection alone), emphasizing an internal control system and cost-of-quality accounting. TQM (Total Quality Management) broadens TQC into a management philosophy: continuous improvement (kaizen), strong customer focus, employee empowerment and teamwork, process ownership, and top-management leadership as the driver of a quality culture — it is cultural/managerial rather than a specific technical toolkit.
Difference: TQC/TQM are broad, continuous, culture-and-system-oriented programs with no fixed numerical target or standard project roadmap; Six Sigma is a narrower, quantitative, project-driven execution methodology with an explicit numerical goal (3.4 DPMO) and a fixed five-phase roadmap (DMAIC), and is typically run within a TQM-style culture to deliver specific, measurable improvements rather than replacing it.
Quality awards (e.g. the Malcolm Baldrige National Quality Award, the Deming Prize) are competitive, holistic assessments of an organization's overall quality management maturity — leadership, strategic planning, customer/market focus, workforce focus, process management, and results — scored against a detailed criteria framework, usually with a limited number of winners per cycle. Their objective is to recognize and publicize role-model performance and to give organizations a rich self-assessment/benchmarking framework, not to certify any single supplier relationship.
Quality certification (e.g. ISO 9001) is a pass/fail, standard-based audit of whether a documented quality management system exists and is followed, performed by an accredited third-party registrar, with no ranking or scoring — an organization either meets the standard's clauses or it does not. Its objective is narrower and more transactional: to give a customer objective, portable, third-party assurance about a specific supplier's quality system, substituting for costly individual second-party (customer) audits.
Recent trends in supplier-producer relations have shifted from adversarial, price-driven, multiple-competing-source relationships with heavy incoming inspection toward partnership-based, single/preferred-source relationships: long-term contracts, supplier development and joint problem-solving, sharing of process-capability and SPC data instead of lot-by-lot inspection, and certification/registration (ISO 9001 or a sector-specific scheme) used to qualify suppliers so that incoming inspection can be reduced or eliminated ("ship-to-stock").
ISO/TS 16949 is the automotive-sector-specific quality management standard, built on the ISO 9001 process-based framework but adding automotive-industry-specific requirements (developed jointly by IATF members and major automakers): mandatory use of core APQP/PPAP tools (Advanced Product Quality Planning, Production Part Approval Process), FMEA, MSA (measurement systems analysis), SPC, continuous improvement and defect-prevention requirements, and a strong emphasis on reducing variation and waste in the supply chain. It replaced the earlier fragmented set of national automotive schemes (QS-9000 in North America, VDA6.1 in Germany, EAQF in France, AVSQ in Italy) with a single, globally recognized certification so a supplier needs only one audit/certificate to qualify with automakers worldwide — a direct example of the "certification substitutes for individual customer audits" trend above.
The traditional (sequential, "over-the-wall") approach hands a design from marketing to design engineering to process/manufacturing engineering to quality to production, one function at a time, each group discovering the next group's constraints only after its own work is "finished." Concurrent engineering instead runs design, manufacturing, quality, and supplier/service inputs in parallel on cross-functional teams from the start of the project.
Advantages: (1) manufacturability and quality problems surface during design, when changes are cheap, rather than during production ramp-up or in the field, where they are far more expensive to fix; (2) shorter time-to-market, since downstream activities (tooling design, process planning, supplier qualification) start before the design is fully frozen; (3) fewer engineering-change orders after release, because tooling/assembly/quality constraints are designed in rather than discovered afterward; (4) better alignment with Design for Manufacturability/Assembly (DFM/DFA) and Taguchi robust-design principles, since process capability and variation are considered while the design is still flexible; (5) improved cross-functional communication and shared ownership.
Role of QFD (Quality Function Deployment). QFD is the structured tool that makes concurrent engineering's parallel, cross-functional work coherent: starting from the "House of Quality," it translates the voice of the customer (WHATs) into engineering characteristics (HOWs), ranks them by customer importance, benchmarks against competitors, and then cascades that same WHATs→HOWs translation down through successive houses — from product characteristics to part characteristics, to process operations, to production/control requirements. Because every function (marketing, design, manufacturing, quality) works from the same cascaded set of prioritized requirements simultaneously, QFD gives concurrent engineering teams a common, quantified reference point instead of relying on informal hand-offs, directly enabling the parallel work the sequential approach could not support.
