In manufacturing, a process Failure Mode and Effects Analysis (pFMEA), is an indispensable tool for identifying potential production problems, their causes, and effects. Most often this analysis is augmented with a Criticality assessment (FMECA), where the Severity (S) of the effects, and the probability of Occurrence (O) and Detection (D) of the failure mode are rated. A scale of 1 to 5, or 1 to 10, is assigned to each S, O, and D, and then multiplied to obtain a risk priority number (RPN). The higher the RPN, the higher the assigned criticality.
Because pFMEAs are such an important risk-mitigation tool, it is critical to avoid the following six (6) mistakes during their development:
⚠️ Mistake #1: Not assembling the right team.
Make sure your FMEA team is cross-functional and includes a person intimately familiar with the process – e.g., an operator, if possible.
Just as importantly, make sure it includes someone not familiar with the process at all because…
“In the beginner’s mind there are many possibilities, but in the expert’s, there are few.” - Shunryu Suzuki.
Note: this is the same reason why having someone not connected to a new product development project (i.e., an independent reviewer) is used to great effect during Design Reviews (see my blog post on Design Verification & Validation for more information)
⚠️ Mistake #2: Showing up to the first meeting with a blank FMEA.
Always come prepared with a draft FMEA, which the team will review, add, change, or remove from. This will avoid the paralysis that comes with “blank page syndrome”, saving precious time.
⚠️ Mistake #3: Not breaking down the process into smaller steps.
Use the existing work instruction (WI) to break down the manufacturing process into smaller steps the team can easily tackle. For new processes, draft a WI before starting the PFMEA.
⚠️ Mistake #4: Waiting too long to draft the FMEA.
The PFMEA should be started as soon as possible in process development for highest impact. Too late, and it risks becoming a paperwork exercise (because later on, you may no longer be able to implement the most effective controls).
⚠️ Mistake #5: Not using quantitative data.
While qualitative analysis is appropriate for a first-of-its-kind process, using quantitative data [scrap, complaints, non-conformances (NCs), etc.] to support Occurrence levels is always preferred. Data from a predicate (similar) process can be used as a first order approximation when possible.
⚠️ Mistake #6. Listing ambiguous causes.
‘Operator Error’ and other ambiguous causes do not have clear controls. Instead, clarify that the ‘incorrect setpoint’ was used (for example). Including the correct setpoint in the work instruction and operator training on the document, can be used as controls.
In addition to avoiding the mistakes above, here are two tips to get the most out of your pFMEAs:
✅ Tip #1: Meeting in person is much more effective than through Zoom or Microsoft Teams.
You will get tons more engagement and a better FMEA this way!
✅ Tip #2: PFMEAs are great tools for assessing risks arising from process changes.
When a process change is proposed, the existing PFMEA can be consulted to determine the affected step, potential impact (Effects), and risk level.
Although developing a pFMEA can be time-consuming, the benefits gained through risk-reduction to the patient, customer, manufacturing organization, and operators, are well worth the effort.
If you found this post helpful, click the ️❤️ icon at the bottom of the page and consider subscribing to the blog.