-
Accept / Reject Criteria
This refers to the measurement and decision of whether or not the quality and performance of a product or service is acceptable. These criteria can be technical specifications (such as level of size tolerances allowed in parts), contract terms and conditions, or performance of a process or service. Implied in the term is the idea that all processes, products, and actions should have an assigned level of acceptable performance that one can measure. Consistent with the understanding that criteria grow out of process capabilities, technicians should periodically measure for acceptability to find out how well a process or item is meeting specification.
-
Acceptable Quality Level (AQL)
AQL is limit of a satisfactory process average at a particular quality level when a continuing series of lots is considered.
-
Acceptance Number
In judging the acceptability of a lot or batch, this is the maximum number of nonconformities allowed within a sample, based on a count of the nonconformities. If the batch passes, that is, it’s amount of nonconformities falls at or below the acceptance number. If a batch does not pass, this usually indicates there is a problem with the process.
-
Acceptance Sampling Plan
This is a plan for setting up the acceptable sampling technique that includes criteria setting sample size, and for accepting and rejecting a component, product, or service based on samples meeting prescribed quality criteria. Plans may involve single, double, sequential, chain, multiple, or skip-lot sampling techniques. With variable samples, management may deploy single, double, and sequential sampling techniques.
-
Adequacy of Standards
This is a manufacturing term that refers to the ability of a standard to be used for calibrating the accuracy of a gage or instrument used to measure quality or performance. It is best if such standards are traceable to the National Bureau of Standards. The term also appears when dealing with the measurement of processes and procedures.
-
Agile Manufacturing
This is a set of flexible manufacturing processes that leads to the rapid design and manufacture of products that provides customers with what they want, when and where they want it. Using agile manufacturing rather than traditional manufacturing methods, firms can offer more customized products at competitive prices with no compromise in quality or damage to the environment.
-
Analysis of Means (ANOM)
This is a statistical technique used when running experiments to identify problems and/or capabilities of an industrial process to deliver an end product with the desirable characteristics.
-
Analysis of Variance (ANOVA)
This is a method for breaking down and analyzing the total variation in the outputs of any process, understanding the causes of this variation, and then assessing their significance. The goal is to come up with a process in which variation in outputs will be minimized.
-
AOQ
This is an acronym for average outgoing quality. This refers to the maximum percent of defective products that can go to a customer after inspection is performed by an acceptance-sampling plan. AOQ requires 100 percent inspection if defectives in a batch or lot of products exceed some minimum standard. This helps reassure customers that suppliers will deliver products that will not include more defectives than expected.
-
Audit (Quality)
An independent review conducted to compare some aspect of quality performance with a standard for that performance.
-
Average Run Length (ARL)
Assuming a process is in statistical control and the upper and lower control limits are known, the ARL is the expected number of subgroups inspected and measured before an operator detects whether a process has shifted to a new level of control with a new center line and control limits. There are statistical methods for calculating the ARL.
-
Batch
A definite quantity of some product or material produced under conditions that are considered uniform.
-
Bathtub Curve
This is a composite curve derived from data on early failure rates and constant failure rates combined with a normalized curve of wear-out failure rates. This curve is used in assessing products that have a certain degree of reliability over time. It assumes that failures will either come early on in product life or late in the product life. Between early and late, the rate of failure is lower and generally constant over the life of the product.
-
Bimodal Distribution
This is a distribution with two modes in the frequency of occurrence of whatever the population is being measured. It is usually displayed using a histogram. The bimodal distribution documents a relationship between two factors in a process, such as time of day and level of traffic.
-
Calibration
In quality management, this is a comparison of one measurement system or instrument not verified as accurate to another standard measurement system or instrument with verified accuracy. This is done to identify variations from a required specification for performance.
-
Capability Process Index (Cp)
This is a technique for making a ratio of the specification width (the range within which an output is considered to meet specification) in relation to the process’s natural distribution for assessing whether the process can regularly meet specifications. Process capability is used to determine which processes are not capable of meeting specifications, to identify those processes operating suboptimally, and to estimate the output proportion that will not conform to specifications.
-
Causal Analysis
This technique concentrates on identifying specific failures or defects, discovering root causes of those failures, and concluding with recommendations on how to eliminate those defects by implementing solutions that address the appropriate cause.
-
Continuous Sampling Plan
This is an approach to sampling that is most appropriate for the output of processes that deliver a continuous flow of a product. In this plan, a company begins by inspecting 100 percent of the units coming from the process. After a certain number of items that have been inspected with no defects, the plan goes to inspecting only a fraction of items. This continues until the company finds a nonconforming unit. At that point, the plan reverts back to 100 percent inspection following the same pattern.
-
Control Chart
This is a graph and a statistical process control method used to track the capability and performance of a process over time.
-
Control Chart Factor
This is a factor based on mathematical probability that facilitates the calculation of control limits on X-bar and range charts.
-
Cost of Poor Quality (COPQ)
These are the extra expenses caused by delivering poor quality goods or services to customers. These expenses have two sources: (1) internal failure costs (from defects before customers get the product) and (2) external failure costs (costs after a customer receives the poor product or service). Rework, repairs, lost future business, and warranty payments are all examples of costs associated with poor quality.
-
Cost of Quality (COQ)
This refers to all costs involved in the prevention of defects, assessments of process performance, and measurement of financial consequences. Cost of quality is the cost justification of quality efforts.
-
Critical Defect
This is a defect that, based on experience, is considered hazardous to employees or to have the potential to harm or injure end users of a product or process.
-
Defect
This is (a) any nonconformance from a customer’s requirement (b) any attribute of a product or service that fails to meet specifications (c) any state of unfitness for use. Defects are usually cataloged into four types by degree, being very serious, serious, major and minor.
-
Demerit Chart
This is a tool for tracking defects, or various types in process outputs. This type of chart is a way to measure the quality of outputs from a process. Demerits are usually broken down into the four categories of defects: critical, serious, major and minor. These are given weights, and each type is then tracked for a specified number of demerits by the total number of products. Using this as a baseline, the companies can then track additional lots or batches of outputs to measure improvement.