Risk-Based Sampling

Risk-based sampling is an inspection and testing approach in which the size, frequency, and rigor of sampling are determined by the assessed level of risk associated with a product, process, supplier, or batch. Instead of applying a uniform sampling plan to all situations, organizations adjust how much they sample based on factors such as criticality, historical performance, process capability, and potential impact on safety, quality, or compliance.

Key characteristics

In industrial and regulated manufacturing environments, risk-based sampling commonly includes:

• Risk assessment as a driver: Sampling plans are informed by formal or semi-formal risk assessments (for example using severity, occurrence, and detectability) rather than only by fixed AQL tables.
• Variable sample sizes: Higher-risk items (such as safety-critical features, new processes, or high-defect suppliers) receive larger sample sizes or 100% inspection, while lower-risk items may be sampled at reduced rates.
• Dynamic adjustment: Sampling intensity can be increased or decreased over time based on nonconformance trends, process capability, audit findings, or changes in design or process.
• Integration with quality systems: The approach is often documented within quality plans, control plans, supplier quality requirements, or inspection plans in MES/QMS systems.
• Focus on critical characteristics: Particular emphasis is placed on critical-to-quality or safety-related characteristics, which may be sampled more heavily than non-critical features in the same lot.

Operational usage in manufacturing

Operationally, risk-based sampling shows up in:

• Incoming inspection: Adjusting sampling levels per supplier, part family, or risk classification, often linked to supplier scorecards and historical defect data.
• In-process checks: Setting higher sampling frequencies at process steps with higher failure risk, new tooling, or recent changes, and lower frequencies on stable, capable operations.
• Final inspection and release: Applying enhanced sampling for high-criticality assemblies, new product introductions, or parts with recent nonconformances.
• Electronic systems: Encoding different sampling rules in MES, LIMS, or SPC systems so that inspection plans are automatically adjusted based on part, process, or supplier risk attributes.

What risk-based sampling is not

• It is not the same as simple random sampling without regard to risk.
• It is not a guarantee of compliance to any specific standard, although some standards and guidance documents encourage risk-based thinking.
• It does not remove the need for documented rationale; sampling decisions are typically supported by written risk assessments and quality procedures.

Common confusion

• Risk-based sampling vs. AQL sampling: AQL (Acceptance Quality Limit) sampling uses predefined statistical plans based mainly on lot size and acceptable defect rates. Risk-based sampling may use AQL tables as inputs, but adjusts them according to risk factors such as criticality, process history, and supplier performance.
• Risk-based sampling vs. 100% inspection: For very high-risk situations, a risk-based approach may justify 100% inspection, but risk-based sampling also supports reduced sampling where risk is demonstrably lower.

Relation to risk and quality management

Risk-based sampling is typically part of a broader risk and quality management strategy. It supports prioritization of inspection resources toward areas with higher potential impact on product quality, regulatory compliance, or customer requirements. In regulated sectors, documented risk-based sampling rationales are often referenced during audits or inspections to show how inspection controls are aligned with identified risks.