Question

Think of a structural system with a sensor of humidity that fault models are developed and implemented.

Please describe your approach to analyzing the outcomes of the fault models in the event of a model giving false-positive results.

Answer 1

I will be using 7 Statistical Quality Control tools to analyse the outcomes of the fault models in the event of a model giving false-positive results . These are described below :

Statistical Quality Control Tool	Description
Stratification	Here any data is converted into a meaningful Dara for analysis . So for example we have a data during last week 80 fault models were produced . We can convert it into by saying how many were produced on each day of the week . Which may give us following information : Monday Tuesday Wednesday Thursday Friday Saturday Total 50 18 6 3 2 1 80 So as we can see most were produced on Monday and after that there is a decreasing trend . So it may have something to do with Sunday Maintenance
Histogram	Here we convert the defects into major and minor and see what is their frequency of occurrence on each day of the week . Day 1 Day 2 Day 3 Major Minor Major Minor Major Minor 48 2 15 5 2 4
Check Sheet	Here we completely identify each and every defect mentioned above . As we can see above there were 50 faulty models were produced on account of 50 defects out of which 48 were major and 2 were minor . The defects can also be repetitive in nature .
Cause and Effect diagram	Here route cause are identified for each and every defect that is what is causing what .
Pareto Chart	This is basically an analysis where we try to find out what are those major 20% of the problems which are causing 80% of the defects leading to faulty models being produced.
Scatter Diagram	Here we plot Dependent Variables on Y axis and Independent variables on X axis and try to draw relationship between them by joining dots .
Control Chart	This helps us to understand whether an industrial is under control and capable of meeting customer or product related specification limits . Here we have upper tolerance and lower tolerance limit.