Full citation

Dai,W., Maropoulos, P.G. &Tang, X.Q. (2010). Proceedings from Failure Knowledge Based Decision-Making in Product Quality.

Format: Conference proceeding

Type: Research — Non-experimental

Experience level of reader: Fundamental

Annotation: This conference paper suggests that product quality may be enhanced by building a knowledge network that represents and shares mechanical failure scenarios. The Failure Knowledge Network provides quantitative information that can be applied to quality related decision-making (and represents an enhanced alternative to the quality function development approach). The authors advocate a six-step process for applying failure knowledge to quality decision-making: (i) predicting and identifying risks and faults, (ii) analyzing the cause and mechanism of the past similar failures, (iii) presenting optional proposals, (iv) selecting the optimal scheme, (v) conducting the designated plan, and (vi) verifying the execution results. The paper includes the results from a short case study from a large aerospace parts and sub-assembly process.

Setting(s) to which the reported activities/findings are relevant: Large business, Small business (less than 500 employees)

Knowledge user(s) to whom the piece of literature may be relevant: Manufacturers

Knowledge user level addressed by the literature: Organization

This article uses the Commercial Devices and Services version of the NtK Model

Primary Findings

Methods:

  • Product quality-related decision-making using the Failure Knowledge Network (FKN) — The first step of the decision-making process is the identification of related failures and characteristics. The second step is determination of the important characteristics of the clusters. Next, there is a comparison between the characteristics of each target. Finally, the interdependent priorities of the characteristics are determined by analyzing dependencies among the targets and characteristics.
    Failure knowledge based decision-making in product quality.
    Occurrence of finding within the model: Tip 2.4, KTA Step 3.A, Step 4.12, Step 4.2, Step 1.3, Step 9.2, Step 8.4
  • Failure Knowledge Network (FKN) — captures and inter-relates mechanical product quality knowledge from five areas: (i) the connection between failures and product functions, (ii) the relationship between failures and product components, (iii) the correlation between failures and organizations, (iv) the association between failures and product processes, and (v) the conjunction among different failures. FKN information is represented in a four-dimensional matrix that includes components, functions, processes and organization. Each element in the matrix is a failure scenario and represents the related failures within the corresponding dimensions. Conventional factors of failures are embodied in the FKN representation. They include event, detection, effect, severity, solution weight, cause, monitor, reappearance, operation, efficiency and precaution. The indexes of each factor are provided by subject matter experts and are set in accordance with the correlation between corresponding characteristics and failures
    Failure knowledge based decision-making in product quality.
    Occurrence of finding within the model: Tip 2.4, KTA Step 3.A, Step 5.4, Step 5.1, Step 4.12, Step 4.2, Step 1.3, Step 9.2, Step 8.4, Step 6.4, Step 6.2

Secondary Findings

Barriers:

  • One drawback to using Failure Modes and Effects Analysis (FMEA) is that it has deficiencies in the expression of the relationship between different failure components. As a result it can not be used as a technique for knowledge formulation. One way to represent and share failure information is to construct a knowledge network of failure scenarios. (Dai [2009])
    Occurrence of finding within the model: KTA Step 3.A, Step 9.2, Step 8.4
  • One of the reasons that product quality failures reoccur is that the knowledge of past failures is not well represented or readily-available to respective parties. One way to represent and share past failures is to construct a knowledge network of failure scenarios. (Hatamura [2003])
    Occurrence of finding within the model: Tip 2.4, KTA Step 3.A, Step 1.3, Step 9.2, Step 8.4