Statistical Methods and Quality Engineering in Advanced Manufacturing

value design EWMA Gradual movement in process location Flexible weighting of recent and historical information Performance depends on smoothing and limit parameters Multivariate chart Correlated quality characteristics Preserves joint process relationships Signals may be difficult to diagnose Distribution-free chart Non-normal or poorly specified processes Reduces dependence on parametric assumptions Often needs substantial reliable reference data Advanced manufacturing increasingly produces skewed, high-dimensional and mixed-format data for which normality assumptions are doubtful. Distribution-free multivariate charts can maintain in-control performance without specifying a complete parametric distribution, although reliable reference samples and computational discipline remain essential. Goodness-of-fit-based procedures extend monitoring beyond mean shifts to broader distributional changes, including alterations in variance, dependence or shape (Zhang, Chen & Zou, 2016). Such platforms should not convert every limit breach into automatic adjustment. Signals require contextual validation because sensor failure, communication latency, maintenance states and planned recipe changes can imitate instability. A layered response should distinguish warnings, operator confirmation, controlled shutdown and engineering review, while preserving time- stamped evidence for root-cause analysis and model refinement. Implementation also requires an operational response architecture. Programmable logic controllers, supervisory interfaces and integrated engineering platforms can acquire signals, execute control logic and communicate alarms rapidly. The Nigerian Siemens TIA Portal case demonstrates how unified automation environments support real-time visualisation, coordinated control and structured process intervention, although it does not itself establish control-chart theory (Sunday &Omoegun, 2022). In advanced manufacturing, SPC is therefore most effective when chart design, measurement integrity, diagnostic expertise and automated escalation are integrated into a documented reaction plan that assigns responsibility for containment, investigation, correction and verification. 2.3 Process Capability, Performance, and Tolerance Analysis Process capability analysis evaluates whether a stable manufacturing process can consistently produce output within engineering tolerances. Its purpose is not merely to report historical conformance, but to compare the process distribution with the allowable specification interval and identify whether improvement should target dispersion, centring, or both. The distinction between capability and performance is essential: capability indices estimate short-term potential under controlled conditions, whereas performance indices reflect longer-term variation, including shifts, drifts, and operational disturbances. Capability claims are defensible only after process stability, measurement adequacy, representative sampling, and specification validity have been established (Kotz & Johnson, 2002). The classical index Cp expresses the ratio of specification width to natural process spread and therefore measures potential capability without considering process location. Cpk incorporates the distance between the process mean and nearest specification limit, exposing losses caused by off-centring. Indices such as Cpm and Cpmk additionally penalise departure from a designated target, making them valuable where nominal performance, economic loss, or functional consistency matters more than simple conformance. These measures are complementary rather than interchangeable because identical values may imply different yields, expected losses, and customer risks under different centring conditions (Wu, Pearn & Kotz, 2009). Capability indices also require statistical inference. A point estimate computed from a limited sample may overstate the true process condition, especially when variability is underestimated or observations contain contamination. Confidence intervals and lower confidence bounds provide evidence for supplier qualification, process release, and improvement verification. Robust bootstrap procedures are useful when outliers distort conventional estimates of the mean and standard deviation, although resampling cannot compensate for an unstable process or a poorly designed measurement system. The relationship between capability and yield must nevertheless be interpreted carefully. An index should reflect the probability that future output will satisfy specifications, yet conventional indices may not map uniquely to actual yield when assumptions concerning normality, independence, or centring are violated. Yield-oriented formulations emphasise the proportion conforming and can improve interpretation, but they still depend on reliable models of process tails, where few observations are available (Chao & Lin, 2006). Advanced manufacturing frequently involves several correlated quality characteristics, such as dimensional accuracy, surface integrity, strength, thermal response, and assembly alignment. Analysing each characteristic independently can exaggerate overall capability because a unit must satisfy all requirements simultaneously. Multivariate indices address this problem by comparing the joint process region with a multidimensional tolerance region; however, their value depends on covariance estimation, scale standardisation, distributional form, and diagnostic transparency (de-Felipe & Benedito, 2017). Figure 1: Illustrative Distribution of Factors Contributing to Process Capability Loss Note: This is a conceptual illustration developed for the review. The percentages are not empirical estimates. Figure 1 offers an illustrative, non-empirical distribution of major contributors to capability loss. It emphasises that excessive process spread and mean off-centring require different interventions, while measurement uncertainty, tolerance mismatch, non-normality, and sampling limitations can distort apparent capability. Tolerance analysis should connect statistical evidence with design intent. Overly narrow tolerances may create unnecessary cost and rejection, whereas excessively broad tolerances can compromise fit, safety, reliability, and customer value. Statistical tolerance design must account for component accumulation, assembly relationships, functional risk, and feasible production variation. African manufacturing evidence reinforces the importance of contextual interpretation. In a Nigerian polypropylene-bag company, capability analysis revealed production spreads exceeding specification intervals across two plants, demonstrating that centring alone could not compensate for excessive variability and that improvement required process-level action (Odeyinka et al., 2021). For uncertain or limited evidence, Bayesian methods offer a structured means of combining prior knowledge with observed data and expressing capability decisions probabilistically, although conclusions remain sensitive to prior assumptions and model choice (van der Merwe, Sjölander & van Zyl, 2019). Process capability and tolerance analysis must operate as linked activities, joining stable data, valid specifications, uncertainty quantification, and proportionate improvement decisions. 3. Design of Experiments and Response Surface Methods for Process Optimisation Design of experiments provides a structured means of learning how controllable manufacturing factors influence quality, productivity and process robustness. Unlike one-factor-at-a-time trials, factorial experimentation changes several inputs systematically, permitting efficient estimation of main effects, interactions and curvature. Its value lies in converting process improvement from informal adjustment into reproducible evidence. The classical sequential philosophy begins with a screening region, estimates the local direction of improvement and progressively moves experimentation towards better operating conditions, thereby reducing the cost and ambiguity associated with exhaustive trial-and-error searches (Box & Wilson, 1951). In advanced manufacturing, the experimental strategy must correspond to the maturity of process knowledge. Full factorial designs offer comprehensive interaction information but become expensive as factor numbers increase; fractional factorial and Plackett–Burman designs are therefore useful for screening numerous variables with relatively few runs. Once influential factors have been identified, central composite, Box–Behnken or other second-order designs can estimate curvature and construct response surfaces around a promising region. Randomisation protects against time-related bias, replication estimates pure error, and blocking separates nuisance variation from factor effects. Model adequacy must be assessed through residual analysis, lack-of-fit testing, leverage diagnostics and confirmation experiments rather than through coefficient significance alone under realistic industrial production constraints and resource limitations (Khuri & Mukhopadhyay, 2010). Process optimisation rarely concerns only the response mean. Manufacturers may seek a target dimension while simultaneously reducing variance caused by noise factors such as material batches, ambient conditions or equipment wear. Dual-response modelling treats the fitted mean and variance as related optimisation objectives, supporting operating conditions that are both accurate and robust. This approach strengthens quality engineering because a setting that maximises average performance may remain commercially unacceptable if variability produces defects, rework or unstable downstream assembly. Robust optimisation should therefore distinguish controllable design factors from difficult-to-control environmental factors and explicitly evaluate their interactions (Vining & Myers, 1990). Multiple-response problems are equally common. Machining parameters may influence surface roughness, tool life, cutting force and material-removal rate in conflicting directions, while additive or thermal processes may require simultaneous control of density, strength, dimensional accuracy and energy use. Desirability functions transform individual responses into dimensionless objectives and combine them into an overall criterion, allowing engineers to reflect upper limits, lower limits, targets and relative priorities. The resulting optimum must nevertheless be interpreted cautiously because weightings and desirability shapes embody managerial judgement, while a high composite score can conceal poor performance in a critical response (Derringer & Suich, 1980). Recent African applications demonstrate the practical capacity of RSM to reduce experimental burden in materials development. In South African spark-plasma sintering research, a user- defined response-surface design modelled the effects of sintering temperature and heating rate on microhardness and relative density in a high-entropy alloy. The work illustrates how fitted quadratic relationships, analysis of variance and experimental validation can identify feasible settings while avoiding unrestricted combinations that would be costly or technically unsafe to test (Olorundaisi et al., 2023). RSM is similarly valuable in machining difficult materials, where physical mechanisms interact nonlinearly. A South African investigation of Ti6Al4V milling used designed trials, predictive models and numerical optimisation to compare carbide, SiAlON and cubic-boron-nitride inserts. By relating cutting speed, feed per tooth and depth of cut to tool-life responses, the study demonstrated that optimisation must integrate statistical fit with metallurgical interpretation, microscopy and confirmation machining rather than treating software-generated optima as self- validating results. Within equipment restoration, DOE can be combined with root-cause analysis to move beyond replacing failed components towards determining why degradation occurred and which recovery settings produce durable performance. Factorial designs, Taguchi approaches and response surfaces can test restoration variables, while fault trees, fishbone analysis and failure-mode evaluation identify plausible causal pathways. Nigerian scholarship positions this integration as a means of improving restoration accuracy, limiting recurrent faults and extending asset life, although the quality of conclusions remains dependent on reliable failure data, competent experimental execution and post-restoration validation (Akanbi, Ganiu & Sunday, 2023). 4. Measurement Systems, Sampling Strategies, and Quality Data Integrity Measurement systems, sampling strategies, and data integrity form a single evidential chain in advanced manufacturing: if any link is weak, subsequent statistical analysis may be technically sophisticated yet operationally misleading. Measurement-system analysis distinguishes genuine part-to-part variation from error introduced by instruments, fixtures, operators, methods, and environmental conditions. Repeatability captures variation when the same appraiser measures the same item under unchanged conditions, whereas reproducibility reflects differences across appraisers or operating settings. Gauge repeatability and reproducibility studies therefore require representative parts, randomised measurement order, sufficient replication, and models that estimate relevant variance components rather than merely reporting a percentage of tolerance (Burdick, Borror & Montgomery, 2003). Contemporary practice extends beyond elementary crossed studies. Automated optical systems, coordinate-measuring machines, destructive tests, and multivariate sensors may involve nested structures, within-part heterogeneity, non-normal errors, or correlated outputs. A systematic review of Gauge R&R research shows that design choice, estimation method, acceptance criterion, and industrial context materially affect judgments about whether a measuring system is adequate. Consequently, a gauge should be assessed for its intended decision, measurement range, resolution, stability, bias, linearity, and traceability, not approved through a universal threshold detached from engineering risk (Soares et al., 2022). Traceability links shop-floor measurements to recognised standards through an unbroken calibration chain with stated uncertainty. This infrastructure is especially important where suppliers, regulators, and customers must compare results generated by different laboratories or plants. African metrology development has historically faced shortages of calibration facilities, trained personnel, reference standards, and sustained institutional support, which can weaken confidence in test results and constrain industrial competitiveness. Regional networks and competence-building programmes are therefore essential for embedding calibration, uncertainty evaluation, accreditation, and measurement culture within manufacturing practice (Charki et al., 2009). Sampling determines which portion of production becomes visible to decision-makers. Random, stratified, systematic, and rational-subgroup designs serve different purposes; their suitability depends on lot structure, process dynamics, defect rarity, inspection cost, and the consequences of error. Acceptance sampling does not estimate every feature of a process or improve an unstable operation. It provides a lot-sentencing rule that balances producer’s risk, consumer’s risk, sample size, and an acceptance criterion. Capability-based plans are particularly useful when defect rates are extremely low and conventional samples are unlikely to contain nonconforming units, but their reliability still depends on stable processes and credible distributional assumptions (Pearn & Wu, 2007). Advanced manufacturing further requires quality data to remain complete, accurate, timely, consistently labelled, and traceable across interconnected stages. Isolated machine databases, incompatible identifiers, duplicated records, missing timestamps, and changing inspection definitions can sever the relationship between process conditions and product outcomes. Digital quality-management platforms address this fragmentation by integrating machine, inspection, and product data across the production chain, thereby supporting defect-propagation analysis, virtual quality gates, and targeted inspection. Their analytical value nevertheless depends on common data models, contextual metadata, disciplined master-data governance, and reliable product genealogy. Figure 2: Illustrative Distribution of Quality-Data Risk Across the Manufacturing Information Chain Note: This conceptual chart was developed for the review. The percentages are illustrative and do not represent empirical estimates. Figure 2 presents a conceptual—not empirical—distribution of quality-decision risk across the manufacturing information chain. Its emphasis on calibration and Gauge R&R reflects the principle that downstream models cannot recover information already distorted at measurement. The other categories demonstrate that sampling, capture, labelling, integration, traceability, and access control are mutually reinforcing rather than independent administrative tasks. Quality-data integrity also has behavioural and security dimensions. Although conducted in university registries rather than factories, Nigerian evidence on information-security awareness is defensibly transferable to manufacturing record systems because both environments depend on personnel handling confidential, operationally significant data. Weak training, inconsistent managerial commitment, and routine non-compliance can compromise records despite the presence of technical controls (Onche, Ogbonna & Adegbite, 2020). Similarly, research on password and authentication behaviour among Nigerian administrative professionals is not a measurement-science source, but it illuminates human vulnerabilities relevant to manufacturing execution systems, laboratory information systems, and electronic quality records. Strong authentication, role-based access, audit trails, backup controls, and recurrent user education are necessary to prevent unauthorised alteration, deletion, or disclosure of evidence used for release and improvement decisions within highly automated, regulated, and geographically distributed production networks worldwide (Onche& Ogbonna, 2022). 5. Six Sigma, Lean Quality Engineering, and Continuous Improvement Lean Six Sigma integrates Lean’s emphasis on flow and waste elimination with Six Sigma’s statistical discipline for reducing variation and defects. Within advanced manufacturing, the combined approach is valuable because production losses rarely arise from a single source: excessive motion, waiting, changeover delays, unstable equipment, measurement error and inconsistent materials frequently interact. Its principal strength is therefore not the accumulation of isolated tools, but the organisation of improvement around customer-defined quality requirements, verified process evidence and sustained operational control. Comparative scholarship across industrial sectors indicates that Lean Six Sigma is most effective when project selection, analytical rigour, leadership sponsorship and workforce participation are treated as mutually dependent elements (Singh & Rathi, 2019). The DMAIC sequence provides the principal architecture for improving an existing process. Define clarifies the defect, customer requirement, project boundary and financial or operational consequence. Measure establishes the baseline through reliable operational definitions and validated data. Analyse tests plausible causal mechanisms rather than accepting symptoms as root causes. Improve develops, pilots and verifies countermeasures, while Control embeds the revised method through standard work, monitoring plans, ownership and escalation rules. Lean techniques increase process visibility and speed, whereas statistical methods establish whether observed improvement is larger than routine variation. This discipline also prevents local gains from displacing losses elsewhere systemically. Table 3: Integrated Lean Six Sigma Framework for Manufacturing Improvement DMAIC phase Principal purpose Illustrative tools Required quality- engineering output Define Specify value and the improvement problem Voice of customer, SIPOC and project charter Measurable critical-to- quality requirement Measure Establish trustworthy baseline performance Process map, Gauge R&R and capability analysis Valid data and quantified current state Analyse Identify verified causes of loss Pareto analysis, regression and cause-and-effect analysis Evidence-supported root causes Improve Remove waste and reduce variation DOE, mistake-proofing, 5S and SMED Validated future-state process Control Prevent deterioration and institutionalise learning SPC, visual management and control plan Stable ownership and response system Digital manufacturing is extending this framework beyond periodic improvement events. In high-speed packaging, sensor networks, advanced SPC, machine learning and digital twins can strengthen DMAIC by accelerating defect detection, testing interventions virtually and supporting predictive maintenance. However, technology does not remove the need for causal validation, competent interpretation or standardised reaction plans; poorly governed analytics may simply automate incorrect assumptions (Akanbi, Ganiu & Sunday, 2023). African manufacturing evidence demonstrates the practical relevance of integrating flow and variation reduction. In a South African paper-production case, value-stream mapping, 5S, Pareto analysis and process-efficiency measures exposed excessive downtime, labour movement and non-value-added time, enabling targeted redesign of the production system. The case also shows that sustainable improvement requires performance indicators to be connected to customer dissatisfaction and operational constraints rather than pursued as abstract efficiency targets (Adeodu, Kanakana-Katumba & Rendani, 2021). The three healthcare references supplied for this section do not constitute direct manufacturing evidence; they are defensibly useful only as cross-sector illustrations of continuous-improvement governance. Nurse-led quality improvement in clinical decision units demonstrates the importance of frontline ownership, structured observation and coordinated handovers when processes are time-sensitive and failure consequences are substantial (Ayinde, Ojukwu &Ohunyon, 2023). Infection-control leadership similarly illustrates how standard protocols, resource coordination, communication and safety culture sustain compliance during operational disruption (Ohunyon, Ojukwu & Ayinde, 2023). Sepsis-management interventions show how explicit time targets, rapid escalation, digital alerts and multidisciplinary workflows can reduce delay in a critical process, principles analogous to takt-time management and abnormality response in manufacturing. These transferable lessons should inform implementation design without being presented as evidence of factory-level performance. 6. Reliability Engineering, Failure Analysis, and Risk-Based Quality Management Reliability engineering provides the analytical foundation for ensuring that manufacturing assets, processes and products perform required functions for specified durations under defined conditions. Its scope extends beyond component failure rates to system complexity, failures, human intervention, uncertainty and lifecycle decisions. Advanced manufacturing intensifies these concerns because tightly coupled equipment can allow a local defect to disrupt a value stream. Reliability models must combine physical understanding, probabilistic evidence and operational context rather than treating historical failure data as representative of future behaviour (Zio, 2009). Failure analysis converts adverse events into structured knowledge. Failure mode and effects analysis identifies modes, causes, consequences and controls before prioritising corrective action. Risk-priority numbers may generate identical rankings for risk profiles, assume equal weighting and obscure uncertainty. Defensible applications supplement ordinal scoring with criticality, fuzzy logic, multi-criteria methods or probabilistic assessment, while retaining expert review and traceable assumptions (Liu, Liu & Liu, 2013). Table 4: Comparative Reliability and Risk-Based Quality Management Methods Method Primary question Typical evidence Quality contribution FMEA How can the process fail? Modes, causes and controls Preventive prioritisation Fault-tree analysis How can a top event occur? Logical combinations of events System-level causal analysis RCM Which maintenance preserves functions? Functions, consequences and task effectiveness Criticality-based maintenance Condition-based maintenance When is deterioration actionable? Sensors, trends and diagnostics Timely intervention Risk-based maintenance Where does failure create greatest exposure? Probability and consequence Risk-prioritised allocation Reliability-centred maintenance shifts attention from preserving components to preserving functions at acceptable risk. It evaluates functional failures, consequences and tasks before assigning preventive, predictive, redesign or run-to-failure policies. A ferrochrome case from Southern Africa shows that neglected preventive maintenance can reduce availability, destabilise operating temperature and compromise production quality, demonstrating the connection between asset reliability and process conformance (Fore & Msipha, 2010). Condition-based maintenance strengthens this approach by using vibration, temperature, lubricant, acoustic or electrical signals to identify degradation. Diagnostics determine the present fault state, while prognostics estimate future condition or remaining useful life. Effective programmes require validated sensors, feature extraction, dependable thresholds and decision rules linking indications to maintenance action; otherwise, false alarms or missed detections generate avoidable cost and risk (Jardine, Lin & Banjevic, 2006). Risk-based maintenance integrates failure likelihood with consequence severity to prioritise inspection and maintenance. It prevents scarce resources from being distributed uniformly across assets whose safety, environmental, production and financial consequences differ. Quantitative schedules remain sensitive to data quality, model assumptions and changing conditions, requiring periodic review rather than static compliance (Khan &Haddara, 2003). Nigerian research on the transition from reactive to predictive maintenance emphasises monitoring, analytics and intelligent diagnostics as mechanisms for reducing unplanned downtime and enabling intervention. Its practical relevance lies in repositioning maintenance as a data-governed reliability function, although adoption depends on digital competence, integrated records and organisational readiness (Sunday et al., 2020). Reliability and quality management become mutually reinforcing when asset-criticality analysis is integrated with Lean Six Sigma. This architecture links failure prevention with variation reduction, root-cause removal, lifecycle cost control and sustained process capability, while requiring coherent operational measures and cross-functional governance (Olutimehin& Ganiu, 2022). 7. Artificial Intelligence, Machine Learning, and Real-Time Quality Analytics Artificial intelligence and machine learning are transforming quality engineering from retrospective defect classification into predictive, adaptive and real-time decision support. Conventional inspection identifies nonconforming output after value has already been added, whereas data-driven models can infer emerging quality loss from sensor streams, process histories, images and production data. Their principal contribution lies in discovering nonlinear interactions that are difficult to specify through conventional parametric models. However, successful deployment requires a defined quality objective, representative data, appropriate validation and an operational response that converts predictions into controlled action rather than generating isolated analytics (Fahle, Prinz &Kuhlenkötter, 2020). Machine vision is among the most mature applications. Industrial cameras, structured illumination and image-processing pipelines enable non-contact examination of surfaces, dimensions, assemblies and labels at production speed. Deep convolutional networks can classify defects, localise anomalies and segment damaged regions, but their reliability depends on stable illumination, suitable optics, image resolution and realistic training samples. Accuracy obtained on curated datasets may deteriorate when production introduces reflections, contamination, product variants or unseen defect morphologies, making robustness and requalification essential (Ren et al., 2022). Predictive quality extends analysis beyond visible defects by estimating product conformity from upstream process conditions. Edge-cloud architectures can acquire shop-floor data, execute models near equipment, and communicate results to manufacturing systems with lower latency. This approach can reduce unnecessary inspection where predictive confidence is adequate, yet it must preserve traceability, fail-safe escalation and selective physical verification. Models should support rather than silently replace statutory, safety-critical or customer-mandated inspections (Schmitt et al., 2020). In multistage manufacturing, quality is shaped by variation propagated across sequential operations. Multi-task deep-learning architectures can share information among stages and outputs, allowing related quality characteristics to be predicted jointly rather than independently. Such models offer richer representations of cross-stage dependency, but they also require synchronised records, consistent product genealogy and treatment of missing or noisy labels. Without these foundations, high predictive accuracy may reflect data leakage or transient correlations rather than stable causal relationships (Wang et al., 2023). Explainability is therefore central to credible industrial AI. Engineers must understand which variables, image regions or process stages influenced a prediction before authorising adjustment, containment or release. Explainable artificial intelligence can expose influential features and support root-cause investigation, although post-hoc explanations are not guarantees of causal correctness. Validation should evaluate calibration, class-specific error, uncertainty, drift and operational consequences alongside aggregate accuracy, particularly where false acceptance and false rejection carry unequal costs (Hoffmann & Reich, 2023). Real-time analytics also depend on sensing architecture. Sensor-based HVAC monitoring illustrates how temperature, pressure, airflow, and equipment-condition measurements can support earlier fault recognition and more targeted intervention. The Nigerian study aligns with intelligent quality monitoring as an applied diagnostic example, although its publisher archive presents an unresolved inconsistency between the supplied 2022 date and a 2020 issue label. Its relevance lies in linking continuous measurement with automated fault awareness, not in establishing general machine-learning theory (Sunday &Omoegun, 2022). African manufacturing adoption further demonstrates that technical capability alone is insufficient. South African evidence associates AI adoption with improved quality control and productivity, while also emphasising workforce transformation and organisational readiness. Manufacturers require data governance, cybersecurity, model ownership, interdisciplinary competence and procedures for human override. Real-time quality analytics become strategically valuable when integrated with statistical process control, metrology and continuous improvement, ensuring that algorithmic signals are tested against engineering knowledge and translated into auditable decisions. 8. Implementation Challenges, Emerging Directions, and Research Priorities Advanced manufacturing implementation is constrained less by the availability of statistical and digital tools than by the organisational capacity to integrate them into stable, governed and economically justified systems. Legacy machinery, fragmented databases, incompatible communication protocols and uneven metrological practices prevent firms from establishing a trustworthy data architecture. These limitations are consequential for small and medium-sized manufacturers, which must balance automation expenditure against uncertain returns, volatile demand and shortages of specialist personnel. South African evidence indicates that Industry 4.0 can influence manufacturing performance, but its benefits depend on advanced education, information and communication technology investment and industrial policy rather than technology acquisition alone. Workforce readiness constitutes a second implementation barrier. Statistical quality engineering increasingly requires hybrid competence in process physics, programming, data science, cybersecurity and operational improvement. Employees may resist new systems when automation is introduced without role clarity, participation or development pathways. Although derived from university administration rather than manufacturing, Nigerian research on contemporary office management provides transferable evidence that technology availability does not guarantee productive use; sustained training and practical competence are required to convert modern equipment into improved organisational output (Onche, 2021). Governance and professional responsibility are equally important because advanced quality systems distribute decision authority across engineers, operators, analysts and algorithmic platforms. Poor documentation, reporting or uncritical acceptance of automated recommendations can compromise traceability and regulatory confidence. Evidence concerning professional ethics in Nigerian higher education is not manufacturing-specific, yet it illustrates how moral standards, workplace expectations and institutional treatment shape responsible conduct. Manufacturing research should therefore examine ethical accountability for model validation, data alteration, automated release decisions and the equitable redesign of work (Onche& Ogbonna, 2021). Cybersecurity has become inseparable from quality assurance as connected gauges, programmable controllers, cloud platforms and digital twins enlarge the attack surface. Bring- your-own-device practices can introduce unmanaged applications, weak access controls and unverified data pathways. The Nigerian university evidence is transferable only at the level of organisational information risk, but it reinforces the need for device governance, network segmentation, authentication, encryption and auditable access to production and quality records (Onche, Adegbite & Ogbonna, 2023). Technical controls must be supported by recurrent behavioural intervention. Awareness programmes should progress beyond compliance exercises towards role-specific simulations, feedback and measurable reinforcement. The integrative Nigerian review of insider-threat behaviour supports research into how knowledge, self-efficacy and organisational culture influence secure conduct, although manufacturing studies must validate these relationships within operational technology environments where downtime and unsafe process manipulation carry consequences. Sustainability and infrastructure resilience represent further priorities. Nigerian small-scale manufacturing illustrates how unreliable electricity, capital constraints and dependence on fuel- based generation can obstruct stable production and quality control. Distributed solar systems may improve continuity and environmental performance, but implementation requires lifecycle costing, storage planning, maintenance capability and alignment with load profiles rather than symbolic installation (Sunday &Omoegun, 2018). Hybrid-energy research similarly shows that load distribution must coordinate generation, storage and demand to prevent instability, equipment stress and interruption of quality-critical operations (Sunday &Omoegun, 2019). Emerging research should prioritise interoperable quality architectures, uncertainty-aware digital twins, explainable and drift-resistant artificial intelligence, privacy-preserving federated analytics and human-centred automation. Attention is also required for affordable adoption pathways, transferable validation protocols, circular-manufacturing metrics and standards that allow smaller firms to participate securely in digital supply networks. Comparative longitudinal studies across African, Asian, European and American plants would clarify which technical capabilities remain portable and which require institutional adaptation, while implementation research should report failed deployments, hidden maintenance burdens and workforce consequences alongside demonstrations. Conclusion This review has fulfilled its aim of examining how statistical reasoning and quality engineering support reliable, efficient and continuously improving manufacturing systems. It traced the progression from inspection-based quality control to statistical process control, experimental optimisation, Lean Six Sigma, reliability engineering and digitally enabled analytics, demonstrating that contemporary quality performance depends on the coordinated use of technical, organisational and human capabilities. The analysis established that variation, uncertainty and process instability remain fundamental sources of manufacturing risk, while effective control requires valid measurement systems, appropriate sampling, capable processes and evidence-based intervention. The review further showed that design of experiments and response surface methods provide systematic means of identifying influential variables, modelling interactions and optimising multiple responses under practical constraints. Measurement-system integrity, calibration, traceability and secure data governance were also found to be indispensable because unreliable data can invalidate otherwise sophisticated statistical analyses. Lean Six Sigma and continuous-improvement frameworks strengthen these methods by linking defect reduction, waste elimination and organisational learning, whereas reliability-centred and risk-based approaches align maintenance resources with functional criticality and failure consequences. Artificial intelligence, machine learning and real-time analytics extend quality engineering through predictive inspection, anomaly detection and adaptive decision support. However, their value depends on representative data, explainability, cybersecurity, model governance and continued engineering oversight. Implementation remains constrained by legacy infrastructure, skills shortages, fragmented information systems, financial limitations and weak institutional readiness, particularly in emerging manufacturing economies. Accordingly, manufacturers should integrate statistical methods with robust metrology, interoperable digital infrastructure, workforce development and clearly defined accountability. Future research should prioritise explainable hybrid models, uncertainty-aware digital twins, affordable adoption pathways for smaller firms and longitudinal validation across diverse industrial settings. Overall, sustainable quality advancement requires a balanced architecture in which statistical discipline, technological innovation and human judgement reinforce one another throughout the manufacturing lifecycle. Such integration will strengthen resilience, competitiveness, sustainability and customer confidence across industries. References Adeodu, A., Kanakana-Katumba, M.G. and Rendani, M. (2021) ‘Implementation of Lean Six Sigma for production process optimization in a paper production company’, Journal of Industrial Engineering and Management, 14(3), pp. 661–680. https://doi.org/10.3926/jiem.3479. Akanbi, O., Ganiu, O.S. and Sunday, E.A. (2022) ‘A workforce-centric line-balancing model for throughput maximization in multi-shift operations’, Shodhshauryam: International Scientific Refereed Research Journal, 5(3), pp. 439–464. Available at: https://doi.org/10.32628/SHISRRJ. Akanbi, O., Ganiu, O.S. and Sunday, E.A. (2022) ‘Comprehensive review of statistical process control for precision weight management in global food manufacturing’, Gyanshauryam: International Scientific Refereed Research Journal, 5(3). Page range requires confirmation. Available at: https://gisrrj.com/paper/GISRRJ225387.pdf. Akanbi, O., Ganiu, O.S. and Sunday, E.A. (2023) ‘A review of industrial equipment restoration strategies: Integrating design of experiments and root-cause analysis’, International Journal of Scientific Research in Computer Science, Engineering and Information Technology, 9(10), pp. 444–462. https://doi.org/10.32628/CSEIT2361074. Akanbi, O., Ganiu, O.S. and Sunday, E.A. (2023) ‘Technological advances in Lean-Six Sigma methodologies for reducing process failures in high-speed packaging lines’, International Journal of Scientific Research in Computer Science, Engineering and Information Technology, 9(10), pp. 463–492. https://doi.org/10.32628/CSEIT2361075. Alwan, L.C. and Roberts, H.V. (1988) ‘Time-series modeling for statistical process control’, Journal of Business & Economic Statistics, 6(1), pp. 87–95. https://doi.org/10.1080/07350015.1988.10509640. Ayinde, M., Ojukwu, P.U. and Ohunyon, G. (2023) ‘Enhancing patient safety in clinical decision units through nurse-led quality improvement’, International Journal of Multidisciplinary Research and Growth Evaluation, 4(6), pp. 1478–1491. https://doi.org/10.54660/.IJMRGE.2023.4.6.1478-1491. Bersimis, S., Psarakis, S. and Panaretos, J. (2007) ‘Multivariate statistical process control charts: An overview’, Quality and Reliability Engineering International, 23(5), pp. 517–543. https://doi.org/10.1002/qre.829. Borror, C.M., Montgomery, D.C. and Runger, G.C. (1999) ‘Robustness of the EWMA control chart to non-normality’, Journal of Quality Technology, 31(3), pp. 309–316. https://doi.org/10.1080/00224065.1999.11979929. Box, G.E.P. and Wilson, K.B. (1951) ‘On the experimental attainment of optimum conditions’, Journal of the Royal Statistical Society: Series B (Methodological), 13(1), pp. 1–38. https://doi.org/10.1111/j.2517-6161.1951.tb00067.x. Burdick, R.K., Borror, C.M. and Montgomery, D.C. (2003) ‘A review of methods for measurement systems capability analysis’, Journal of Quality Technology, 35(4), pp. 342–354. https://doi.org/10.1080/00224065.2003.11980232. Chakraborti, S., Human, S.W. and Graham, M.A. (2009) ‘Phase I statistical process control charts: An overview and some results’, Quality Engineering, 21(1), pp. 52–62. https://doi.org/10.1080/08982110802445561. Chao, M.T. and Lin, D.K.J. (2006) ‘Another look at the process capability index’, Quality and Reliability Engineering International, 22(2), pp. 153–163. https://doi.org/10.1002/qre.692. Charki, A., Ferrier, C., Bonnier, G., Berrada, M., Hassani, E.A. and Osseni, L. (2009) ‘Metrology in Africa: Role of the CAFMET’, Accreditation and Quality Assurance, 14(4), pp. 227–230. https://doi.org/10.1007/s00769-009-0496-9. Costa, A.F.B. and Castagliola, P. (2011) ‘Effect of measurement error and autocorrelation on the X̄ chart’, Journal of Applied Statistics, 38(4), pp. 661–673. https://doi.org/10.1080/02664760903563627. Crosier, R.B. (1988) ‘Multivariate generalizations of cumulative sum quality-control schemes’, Technometrics, 30(3), pp. 291–303. https://doi.org/10.1080/00401706.1988.10488402. de-Felipe, D. and Benedito, E. (2017) ‘A review of univariate and multivariate process capability indices’, The International Journal of Advanced Manufacturing Technology, 92, pp. 1687–1705. https://doi.org/10.1007/s00170-017-0273-6. Derringer, G. and Suich, R. (1980) ‘Simultaneous optimization of several response variables’, Journal of Quality Technology, 12(4), pp. 214–219. https://doi.org/10.1080/00224065.1980.11980968. Escobar, C.A., McGovern, M.E. and Morales-Menendez, R. (2021) ‘Quality 4.0: A review of big data challenges in manufacturing’, Journal of Intelligent Manufacturing, 32(8), pp. 2319– 2334. https://doi.org/10.1007/s10845-021-01765-4. Fahle, S., Prinz, C. and Kuhlenkötter, B. (2020) ‘Systematic review on machine learning methods for manufacturing processes: Identifying artificial intelligence methods for field application’, Procedia CIRP, 93, pp. 413–418. https://doi.org/10.1016/j.procir.2020.04.109. Fore, S. and Msipha, A. (2010) ‘Preventive maintenance using reliability centred maintenance : A case study of a ferrochrome manufacturing company’, South African Journal of Industrial Engineering, 21(1), pp. 207–235. https://doi.org/10.7166/21-1-78. Hoffmann, R. and Reich, C. (2023) ‘A systematic literature review on artificial intelligence and explainable artificial intelligence for visual quality assurance in manufacturing’, Electronics, 12(22), article 4572. https://doi.org/10.3390/electronics12224572. Jardine, A.K.S., Lin, D. and Banjevic, D. (2006) ‘A review on machinery diagnostics and prognostics implementing condition-based maintenance’, Mechanical Systems and Signal Processing, 20(7), pp. 1483–1510. https://doi.org/10.1016/j.ymssp.2005.09.012. Jensen, W.A., Jones-Farmer, L.A., Champ, C.W. and Woodall, W.H. (2006) ‘Effects of parameter estimation on control chart properties: A literature review’, Journal of Quality Technology, 38(4), pp. 349–364. https://doi.org/10.1080/00224065.2006.11918623. Jones-Farmer, L.A., Woodall, W.H., Steiner, S.H. and Champ, C.W. (2014) ‘An overview of Phase I analysis for process improvement and monitoring’, Journal of Quality Technology, 46(3), pp. 265–280. https://doi.org/10.1080/00224065.2014.11917969. Katombe, M. and Munapo, E. (2016) ‘The quest for process operations variability reduction in manufacturing firms in South Africa’, South African Journal of Economic and Management Sciences, 19(3), pp. 448–466. https://doi.org/10.4102/sajems.v19i3.1327. Khan, F.I. and Haddara, M.M. (2003) ‘Risk-based maintenance : A quantitative approach for maintenance/inspection scheduling and planning’, Journal of Loss Prevention in the Process Industries, 16(6), pp. 561–573. https://doi.org/10.1016/j.jlp.2003.08.011. Khuri, A.I. and Mukhopadhyay, S. (2010) ‘Response surface methodology’, Wiley Interdisciplinary Reviews: Computational Statistics, 2(2), pp. 128–149. https://doi.org/10.1002/wics.73. Kotz, S. and Johnson, N.L. (2002) ‘Process capability indices—A review, 1992–2000’, Journal of Quality Technology, 34(1), pp. 2–19. https://doi.org/10.1080/00224065.2002.11980119. Kusiak, A. (2018) ‘Smart manufacturing’, International Journal of Production Research, 56(1–2), pp. 508–517. https://doi.org/10.1080/00207543.2017.1351644. Liu, H.-C., Liu, L. and Liu, N. (2013) ‘Risk evaluation approaches in failure mode and effects analysis: A literature review’, Expert Systems with Applications, 40(2), pp. 828–838. https://doi.org/10.1016/j.eswa.2012.08.010. Lowry, C.A., Woodall, W.H., Champ, C.W. and Rigdon, S.E. (1992) ‘A multivariate exponentially weighted moving average control chart’, Technometrics, 34(1), pp. 46–53. https://doi.org/10.1080/00401706.1992.10485232. Madanhire, I. and Mbohwa, C. (2016) ‘Application of statistical process control in manufacturing industry in a developing country’, Procedia CIRP, 40, pp. 580–583. https://doi.org/10.1016/j.procir.2016.01.137. Maisiri, W., van Dyk, L. and Coetzee, R. (2021) ‘Factors that inhibit sustainable adoption of Industry 4.0 in the South African manufacturing industry’, Sustainability, 13(3), article 1013. https://doi.org/10.3390/su13031013. Markatos, N.G. and Mousavi, A. (2023) ‘Manufacturing quality assessment in the Industry 4.0 era: A review’, Total Quality Management & Business Excellence, 34(13–14), pp. 1655– 1681. https://doi.org/10.1080/14783363.2023.2194524. Odeyinka, O.F., Ogunwolu, F.O., Popoola, O.P. and Oyedokun, T.O. (2021) ‘Process capability analysis in a polypropylene bag manufacturing company’, Journal of Applied Sciences and Environmental Management, 25(8), pp. 1477–1482. https://doi.org/10.4314/jasem.v25i8.30. Ohunyon, G., Ojukwu, P.U. and Ayinde, M. (2023) ‘Nurse leadership in infection control during healthcare crises’, International Journal of Multidisciplinary Research and Growth Evaluation, 4(6), pp. 1492–1506. https://doi.org/10.54660/.IJMRGE.2023.4.6.1492-1506. Olorundaisi, E., Babalola, B.J., Bayode, B.L., Teffo, L. and Olubambi, P.A. (2023) ‘Optimization of process parameters for the development of Ni–Al-Ti-Mn-Co-Fe–Cr high entropy alloy system via spark plasma sintering’, The International Journal of Advanced Manufacturing Technology, 126, pp. 3323–3337. https://doi.org/10.1007/s00170-023-11311-6. Olutimehin, A.O. and Ganiu, O.S. (2022) ‘Integrating Lean Six Sigma with reliability-centered maintenance: A unified framework for manufacturing excellence’, International Journal of Multidisciplinary Research and Growth Evaluation, 3(6), pp. 983–996. https://doi.org/10.54660/.IJMRGE.2022.3.6.983-996. Onche, V.O. (2021) ‘Contemporary office management at the University of Ibadan, Ibadan, Oyo State, Nigeria’, African Journal of Social Sciences and Humanities Research, 4(2), pp. 1– 9. No DOI verified. Available at: Publisher article page. Onche, V.O. and Ogbonna, C.S. (2021) ‘Professional ethics, moral standards, and the workplace experiences of female secretaries in higher education’, International Journal of Multidisciplinary Research and Growth Evaluation, 2(6), pp. 943–953. https://doi.org/10.54660/.IJMRGE.2021.2.6.943-953. Onche, V.O. and Ogbonna, C.S. (2022) ‘Password practices, authentication behaviour, and data- protection compliance among secretarial professionals in higher education’, International Journal of Multidisciplinary Research and Growth Evaluation, 3(6), pp. 1001–1015. https://doi.org/10.54660/.IJMRGE.2022.3.6.1001-1015. Onche, V.O., Adegbite, M.P. and Ogbonna, C.S. (2023) ‘BYOD adoption and information- security risk in university administrative offices: Evidence from Nigerian federal universities’, International Journal of Multidisciplinary Research and Growth Evaluation, 4(6), pp. 1549–1563. https://doi.org/10.54660/.IJMRGE.2023.4.6.1549-1563. Onche, V.O., Ogbonna, C.S. and Adegbite, M.P. (2020) ‘Information-security awareness and records confidentiality in Nigerian university registries: A behavioural assessment’, International Journal of Multidisciplinary Research and Growth Evaluation, 1(5), pp. 1032–1044. https://doi.org/10.54660/.IJMRGE.2020.1.5.1032-1044. Pearn, W.L. and Wu, C.W. (2007) ‘An effective decision making method for product acceptance’, Omega, 35(1), pp. 12–21. https://doi.org/10.1016/j.omega.2005.01.018. Pepper, M.P.J. and Spedding, T.A. (2010) ‘The evolution of Lean Six Sigma’, International Journal of Quality & Reliability Management, 27(2), pp. 138–155. https://doi.org/10.1108/02656711011014276. Psarommatis, F., Sousa, J., Mendonça, J.P. and Kiritsis, D. (2022) ‘Zero-defect manufacturing: The approach for higher manufacturing sustainability in the era of Industry 4.0’, International Journal of Production Research, 60(1), pp. 73–91. https://doi.org/10.1080/00207543.2021.1987551. Qin, S.J. (2003) ‘Statistical process monitoring: Basics and beyond’, Journal of Chemometrics, 17(8–9), pp. 480–502. https://doi.org/10.1002/cem.800. Rathilall, R. and Singh, S. (2018) ‘A Lean Six Sigma framework to enhance the competitiveness in selected automotive component manufacturing organisations’, South African Journal of Economic and Management Sciences, 21(1), article a1852. https://doi.org/10.4102/sajems.v21i1.1852. Ren, Z., Fang, F., Yan, N. and Wu, Y. (2022) ‘State of the art in defect detection based on machine vision’, International Journal of Precision Engineering and Manufacturing- Green Technology, 9, pp. 661–691. https://doi.org/10.1007/s40684-021-00343-6. Roberts, S.W. (1959) ‘Control chart tests based on geometric moving averages’, Technometrics, 1(3), pp. 239–250. https://doi.org/10.1080/00401706.1959.10489860. Schmitt, J., Bönig, J., Borggräfe, T., Beitinger, G. and Deuse, J. (2020) ‘Predictive model-based quality inspection using machine learning and edge cloud computing’, Advanced Engineering Informatics, 45, article 101101. https://doi.org/10.1016/j.aei.2020.101101. Shuaib, K.M. and He, Z. (2021) ‘Impact of organizational culture on quality management and innovation practices among manufacturing SMEs in Nigeria’, Quality Management Journal, 28(2), pp. 98–114. https://doi.org/10.1080/10686967.2021.1886023. Singh, M. and Rathi, R. (2019) ‘A structured review of Lean Six Sigma in various industrial sectors’, International Journal of Lean Six Sigma, 10(2), pp. 622–664. https://doi.org/10.1108/IJLSS-03-2018-0018. Soares, W.O.S., Peruchi, R.S., Silva, R.A.V. and Rotella Junior, P. (2022) ‘Gage R&R studies in measurement system analysis: A systematic literature review’, Quality Engineering, 34(3), pp. 382–403. https://doi.org/10.1080/08982112.2022.2069505. Stoumbos, Z.G., Reynolds, M.R. Jr, Ryan, T.P. and Woodall, W.H. (2000) ‘The state of statistical process control as we proceed into the 21st century’, Journal of the American Statistical Association, 95(451), pp. 992–998. https://doi.org/10.1080/01621459.2000.10474292. Sunday, E.A. and Omoegun, G.O. (2018) ‘Integrating solar power solutions in small-scale manufacturing industries in Nigeria’, International Journal of Scientific Research in Science, Engineering and Technology, 4(8), pp. 832–853. :https://doi.org/10.32628/IJSRSET23102174. Sunday, E.A. and Omoegun, G.O. (2019) ‘Optimizing electrical load distribution for hybrid solar installations in developing economies’, International Journal of Scientific Research in Mechanical and Materials Engineering, 3(6), pp. 27–47. :https://doi.org/10.32628/IJSRMME19349. Sunday, E.A. and Omoegun, G.O. (2022) ‘Automation of process control systems using Siemens TIA Portal: A case study approach’, International Journal of Scientific Research in Mechanical and Materials Engineering, 6(5), pp. 30–55. verified. Available at: https://ijsrmme.com/home/issue/view/v6i5. Sunday, E.A. and Omoegun, G.O. (2022) ‘Smart fault detection in HVAC systems using sensor- based monitoring’, International Journal of Scientific Research in Science, Engineering and Technology, 7(4), pp. 380–403. https://doi.org/10.32628/IJSRSET229473. Sunday, E.A., Omoegun, G.O., Essien, M.A. and Oluokun, O.A. (2019) ‘Thermodynamic efficiency and control strategies in residential air conditioning systems’, International Journal of Scientific Research in Civil Engineering, 3(3), pp. 52–76. https://doi.org/10.32628/IJSRCE19339. Sunday, E.A., Omoegun, G.O., Essien, M.A. and Oluokun, O.A. (2020) ‘Transitioning from reactive to predictive maintenance in mechanical systems’, International Journal of Scientific Research in Computer Science, Engineering and Information Technology, 6(6), pp. 425–447: https://doi.org/10.32628/CSEIT206643. Tao, F., Qi, Q., Liu, A. and Kusiak, A. (2018) ‘Data-driven smart manufacturing’, Journal of Manufacturing Systems, 48, pp. 157–169. https://doi.org/10.1016/j.jmsy.2018.01.006. Tegegne, D.A., Kitaw, D. and Berhan, E. (2022) ‘Advances in statistical quality control chart techniques and their limitations to cement industry’, Cogent Engineering, 9(1), article 2088463. https://doi.org/10.1080/23311916.2022.2088463. van der Merwe, A.J., Sjölander, M.R. and van Zyl, R. (2019) ‘Bayesian testing for process capability indices’, South African Statistical Journal, 53(2), pp. 87–113. https://doi.org/10.37920/sasj.2019.53.2.3. Vining, G.G. and Myers, R.H. (1990) ‘Combining Taguchi and response surface philosophies: A dual response approach’, Journal of Quality Technology, 22(1), pp. 38–45. https://doi.org/10.1080/00224065.1990.11979204. Wang, P., Qu, H., Zhang, Q., Xu, X. and Yang, S. (2023) ‘Production quality prediction of multistage manufacturing systems using multi-task joint deep learning’, Journal of Manufacturing Systems, 70, pp. 48–68. https://doi.org/10.1016/j.jmsy.2023.07.002. Woodall, W.H. and Montgomery, D.C. (2014) ‘Some current directions in the theory and application of statistical process monitoring’, Journal of Quality Technology, 46(1), pp. 78–94. https://doi.org/10.1080/00224065.2014.11917955. Wu, C.W., Pearn, W.L. and Kotz, S. (2009) ‘An overview of theory and practice on process capability indices for quality assurance’, International Journal of Production Economics, 117(2), pp. 338–359. https://doi.org/10.1016/j.ijpe.2008.11.008. Wuest, T., Weimer, D., Irgens, C. and Thoben, K.D. (2016) ‘Machine learning in manufacturing: Advantages, challenges, and applications’, Production & Manufacturing Research, 4(1), pp. 23–45. https://doi.org/10.1080/21693277.2016.1192517. Yang, H., Rao, P., Simpson, T., Lu, Y., Witherell, P., Nassar, A.R., Reutzel, E. and Kumara, S. (2021) ‘Six-Sigma quality management of additive manufacturing’, Proceedings of the IEEE, 109(4), pp. 347–376. https://doi.org/10.1109/JPROC.2020.3034519. Zhang, C., Chen, N. and Zou, C. (2016) ‘Robust multivariate control chart based on goodness- of-fit test’, Journal of Quality Technology, 48(2), pp. 139–161. https://doi.org/10.1080/00224065.2016.11918156. Zio, E. (2009) ‘Reliability engineering: Old problems and new challenges’, Reliability Engineering & System Safety, 94(2), pp. 125–141. https://doi.org/10.1016/j.ress.2008.06.002.