Quality failure analysis rests on the use of high level human cognition. Here are five aspects of human mental processes that form the foundation of the way effective thinkers use their mental skills. Let’s review them here, in reference to the basic steps of a failure investigation.

1. Clarify Context: Define the frame of reference of the event, events or service conditions that resulted in the failure. The opinions of the stakeholders requesting or requiring the investigation are important to understand. This does not mean that the failure analyst limits project suggestions to what they think will lead to the answer they want. We, as a project team, simply must ensure that we have a way to address all of the relevant concerns. If the stakeholders want suggestions for recurrence avoidance, we must understand the design and manufacturing processes, as well as the procurement processes for the raw material or outsourced treaments (heat treating or plating, for example). Suggesting an appropriate level of detail for the evaluation requires familiarity with the consequences of the failure.

2. Compare-Contrast forms the foundation of rational thinking, so it infuses every aspect of the investigation. Using standardized test methods, when possible, facilitates comparison of the subject part to others, as well as to published data and specifications. The inability to use or obtain standardized test methods complicates interpretation of the data. Perhaps the single most noteworthy item specific to failure analysis that is related to Compare-Contrast is selection of reference or control parts to analyze in some of the same ways that the subject part is analyzed. 

3. Connect-Create forms the foundation of creative thinking, “connecting the dots,” getting to “the AHA! moment” where the data stops being data and starts being useful information, perhaps sometimes even leading to breakthroughs in understanding or the emergence of new ideas.

4. Contemplate Complexity: Do we do this at the end of the project after we have selected a preliminary idea for a fix? Or do we do this before choosing a fix? Really, the experienced, competent and effective analyst has cultivated a habit of evaluating the data as it becomes available to try to make sure that all the relevant factors that influenced the production or service history of the component are revealed. That said, some failure-analysis procedures are performed in a more linear fashion. The small tasks are performed, and then someone else looks at the completed data packet for the first time. To me, such a procedure is not as desirable as a more flexible approach, where results from early analysis may allow the process to be tweaked to provide better information. But that might be the company policy in some situations.

The flexible process allows contemplation of the mutually interactive influences of factors revealed as the inspection proceeds. This allows us to work toward more clearly seeing the functioning of a dynamic system. Seeing the system rather than individual facts allows us to double check, triple check, quadruple check, etc., our work in an interesting way that sharpens the mind. Going down checklists, which always must oversimplify, dulls our intellect.

5. Choose with Confidence: We need to make peace with the fact that certainty is limited to our eventual death and the necessity to pay taxes, as the old saying goes. Confidence is always limited, and in that knowledge we can move forward rather than becoming paralyzed by indecision.