USE OF ANALYTIC HIERARCHY PROCESS (AHP) TO SUPPORT THE DECISION-MAKING ABOUT DESTINATION OF A BATCH OF DEFECTIVE PRODUCTS WITH ALTERNATIVES OF REWORK AND DISCARD João Cláudio Ferreira Soares, Anabela Pereira Tereso, Sérgio Dinis Teixeira Sousa Centre ALGORITMI, University of Minho, Campus de Azurém, 4804-533 Guimarães, Portugal id6293@alunos.uminho.pt; anabelat@dps.uminho.pt; sds@dps.uminho.pt
Oral presentation TOPICS: Introduction.................................3 1. Case study...................................4 2. Results and discussion....................19 3. Conclusions.................................20 4. References..................................21 5.
3/21 1. Introduction • Application of AHP - support the decision-making - destination of a batch of defective products. Alternatives of destination: rework / discard. • • Mathematical development of the model: Excel. From a flow of analysis of quality problems - AHP method adapted and applied - • using evaluation questions to establish the criteria for comparison. Evidence problem analysis -> answers and determination of criteria weights -> • influences of the answers on cost/quality of the product -> rework or disposal.
4/21 2. Case study • Study developed - Brazilian plant of a Japanese auto parts industry (SHOWA) - supplies world-renowned Japanese motorcycle manufacturers (Honda and Yamaha). • Defective product - steering column of one of the models - presented the weld bead displaced from the correct position. Six decision criteria were used in the form of objective questions with "Yes" or "No" • answers. The answers to the questions of the criteria - obtained from the evidence collected • and verified in the technical analysis of the problem. Each criterion undergoes a change of importance (weight) according to the answer • (yes or no) of the respective question. • Information - collected through the engineering manager. Criteria - weighted consensus specialists in the areas of manufacturing, quality and • engineering.
5/21 2.1 Problem Definition Figure 1: Complete steering column assembly Figure 3: Type of defects in the welding process of the steering column in March 2016 Figure 2: Weld bead of the displaced steering column (left), diagram of the welding process with alignment by the fork holes (center and right)
6/21 2.2 Definition of Decision Criteria Table 1: Criteria with two possible conditions and respective tendencies Responses of flow Analysis Attributes/Criteria Yes No Tendency to Tendency to Problem Solved? Rework Discard History of occurrence in Tendency to Tendency to the final customer? Discard Rework Is currently occurring in Tendency to Tendency to the final customer? Discard Rework Tendency to Tendency to Rework plan approved? Rework Discard Company has all the Tendency to Tendency to capabilities to rework? Rework Discard Figure 5: Hierarchical Problem Structuring Rework economically Tendency to Tendency to viable? Rework Discard
7/21 2.3 Weight and relationship of Criteria with alternatives Table 2: Saaty Fundamental Scale – AHP Table 3: Weight of the Criteria in the possibilities of answers "Yes" and "No" Intensity scale of imporance - AHP Intensity scale of Responses of flow Analysis Weight AHP Definition Explanation Attributes/Criteria importance Yes No Yes No 1 Equal importance Two elements contribute equally to the objective. Weak importance of one Experience and judgment moderately favor one element Problem Solved? Tendency to Rework Tendency to Discard 2 9 3 over another over another. History of occurrence in the end Experience and judgment strongly favor one element over Tendency to Discard Tendency to Rework 9 2 5 Strong importance customer? another. Is occurring currently in the end One element is favored very strongly over another; its Tendency to Discard Tendency to Rework 9 3 7 Very strong importance customer? dominance is demonstrated in practice. Rework plan approved? Tendency to Rework Tendency to Discard 4 9 Evidence favors one activity over another, with the highest 9 Absolute importance Company has all the capabilities to degree of certainty. Tendency to Rework Tendency to Discard 5 9 rework? Median of both 2, 4, 6 e 8 When compromise is needed. Rework economically viable? Tendency to Rework Tendency to Discard 9 9 neighboring judgments
8/21 2.4 Hierarchy, criteria analysis and weight assignment for alternatives Table 4: Result of the evaluation of required quality level and cost of rework options Welding production piece (1) Rework to remove the (2) Rework fill with Activity with cord displacement cord for new welding welding Time 28 seconds 83 seconds 16 seconds Condition of Cost 22,82 BRL 43,31 BRL 3,92 BRL Visual Inspection Not satisfy quality Satisfy quality Satisfy quality Rupture test Satisfy maximum load Satisfy maximum load Satisfy maximum load Test macrography Not satisfy penetration Satisfy penetration Not satisfy penetration Lack of penetration High cost, bigger than to possible premature produce a new piece Necessary to rework or fatigue Appraisal Report dispose of the part Not satisfy cost Not satisfy quality
9/21 2.4 Hierarchy, criteria analysis and weight assignment for alternatives Table 5: Problem hierarchy and assignment of analysis flow responses PROBLEM: WELD BEAD OF THE STEERING COLUMN MOVED Responses of flow Weight AHP Alternatives Analysis Goal Dimension Attributes/Criteria (1) Yes No Yes No Discard Rework Problem Solved? 1 0 2 9 2 1 History of occurrence in the 1 0 9 2 1 9 end customer? Quality Is occurring currently in the end Reduce the cost of quality, 0 1 9 3 3 1 customer? mainly with internal and external flaws (depending on Rework plan approved? 0 1 4 9 1 9 the external impact of rework in the field). Company has all the capabilities 1 0 5 9 5 1 to rework? Cost Rework economically viable? 0 1 9 9 1 9
10/21 2.5 Construction of the preference matrices of the alternatives for each criterion. Table 6: Matrices of preference of the alternatives for each criterion Preference for Preference for Criterion 1 Result of the analysis Criterion 2 Result of the analysis YES NO History of YES NO Problem Solved? occurrence in the 1 0 1 0 end customer? C1 Rework Discard C2 Rework Discard Rework 1 2 Rework 1 1/9 Discard 1/2 1 Discard 9 1 Preference for Preference for Criterion 3 Result of the analysis Criterion 4 Result of the analysis Is occurring YES NO YES NO Rework plan currently in the 0 1 approved? 0 1 end customer? C3 Rework Discard C4 Rework Discard Rework 1 3 Rework 1 1/9 Discard 1/3 1 Discard 9 1 Preference for Preference for Criterion 5 Result of the analysis Criterion 6 Result of the analysis Company has all YES NO Rework YES NO the capabilities to economically 1 0 0 1 rework? viable? C5 Rework Discard C6 Rework Discard Rework 1 5 Rework 1 1/9 Discard 1/5 1 Discard 9 1
11/21 2.6 Normalization of each criterion Table 7: Normalization of criteria
12/21 2.6 Normalization of each criterion Table 7: Normalization of criteria
13/21 2.7 Average of the alternatives for each criterion Table 8: Matrices of the averages of the alternatives for each criterion Calculation of the average of the Criterion 1 Calculation of the average of the Criterion 2 Problem Solved? History of occurrence in the end customer? Rework Discard Average Rework Discard Average C1- Criterion 1 C2- Criterion 2 0,667 0,667 0,667 0,100 0,100 0,100 Rework Rework 0,333 0,333 0,333 0,900 0,900 0,900 Discard Discard Calculation of the average of the Criterion 3 Calculation of the average of the Criterion 4 Rework plan approved? Is occurring currently in the end customer? Rework Discard Average Rework Discard Average C3- Criterion 3 C4- Criterion 4 0,750 0,750 0,750 0,100 0,100 0,100 Rework Rework 0,250 0,250 0,250 0,900 0,900 0,900 Discard Discard Calculation of the average of the Criterion 5 Calculation of the average of the Criterion 6 Company has all the capabilities to rework? Rework economically viable? Rework Discard Average Rework Discard Average C5- Criterion 5 C6- Criterion 6 0,833 0,833 0,833 0,100 0,100 0,100 Rework Rework 0,167 0,167 0,167 0,900 0,900 0,900 Discard Discard
14/21 2.8 Definition of preferences for each criterion Table 9: Averages of the alternatives for each criterion which is the array of preferences CRITERIA ALTERNATIVES C1 C2 C3 C4 C5 C6 Rework 0,667 0,100 0,750 0,100 0,833 0,100 Discard 0,333 0,900 0,250 0,900 0,167 0,900
15/21 2.9 Comparison between criteria Table 10: Matrix of comparison between the criteria C1 C2 C3 C4 C5 C6 History of Company has Is occurring Rework CRITERIA Problem occurrence in Rework plan all the currently in the economically Solved? the end approved? capabilities to end customer? viable? customer? rework? C1 Problem Solved? 1 4 2 2 2 1/5 History of C2 occurrence in the 1/4 1 1/4 1/4 1/4 1/5 end customer? Is occurring C3 currently in the end 1/2 4 1 1 1 1/5 customer? Rework plan C4 1/2 4 1 1 1 1/5 approved? Company has all the C5 capabilities to 1/2 4 1 1 1 1/5 rework? Rework C6 economically 5 5 5 5 5 1 viable? SUM 7,75 22,00 10,25 10,25 10,25 2,00
Recommend
More recommend
