Processing new types of quality data

Abstract

Quality engineers are increasingly faced with the need to deal with new types of data, which are significantly different from ordinary numerical data by virtue of their nature and the operations that can be performed with them. Basic concepts related to processing of such data, ie, data similarity, measurement system analysis, variation analysis, and data fusion, need to be thoroughly rethought. Reviewing recent publications in the field, we suggest a common approach to processing all data types on the basis of the idea of defining the distance metric for the appropriate data space. The article discusses six types of quality data (nominal, ordinal, preference chains, strings, tree structured, and product/process distribution) and four data processing aspects (calculating data similarity, error description, data fusion, and intradispersion and interdispersion studies). Necessary information and recommendations are given for each combination of data type and problem. They are also summarized in a table that refers the reader to various sections of the article. Any other data type for which the distance metric is definable can be included into the framework of the proposed unified approach.

Online detection of cyber‐incidents in additive manufacturing systems via analyzing multimedia signals

“…Therefore, we devote our efforts to identifying those deviations via monitoring the actual manufacturing process. A study published by Marmor and Bashkansky 19 has proposed a common approach to process new data sets and discussed six types of quality data. The data in our study fall into the category of process feature distribution.…”

“…A change in the statistical distribution of the process feature is expected after a change in the process has occurred. To detect the change, they 19 also suggest using distance metrics to capture the statistical divergence between two processes. For simplicity, we choose a single‐material Fused Deposition Modeling (FDM) printer for our study, which is the most common and economic printer in the market.…”

A Bayesian Approach for Lifetime Modeling and Prediction with Multi-Type Group-Shared Missing Covariates

“…In addition to the improved precision, the posterior mode of the estimated GSLH approaches closer to the true value when M becomes larger. This is because the group-specific data contributes to the estimation of GSLH, as shown in Equation (11). The estimation performance among all groups are further illustrated in Figure 7.…”

“…Due to the above various causes of missing information, the missing covariates often demonstrate multiple types of attributes, including qualitative, quantitative, or a combination of both [10]. For instance, the covariates may be qualitative by taking nominal values, such as various descriptors related to materials and diverse configurations pertaining to design [8]; or ordinal levels, such as various levels of material quality and diverse usage conditions [11]. Moreover, the covariates may also be quantitative factors that take numerical values on a continuous scale, such as manufacturing process conditions (e.g., pressure, humidity, flow rates) during the manufacturing phase [12] or environment conditions (e.g., loading, temperature) at the operation stage [13].…”

SPC scheme to monitor surgery duration

“…In a more recent paper by same authors, 4 they address process/product feature – distribution distance , stating that “the statistical distance between two theoretical or empirical distributions, or between an empirical and theoretical one, … is widely applied in quality engineering”, and then they enumerate some measures of distributional distance. Each of these may potentially serve to detect significant divergence of a sample distribution from a base distribution (the in‐control distribution), as done by the authors with the Anderson–Darling metric in the aforecited earlier paper.…”