Grab a Red Bull, or a double latte, because today, as promised, we're going to talk turkey about FMEA (Failure Mode Effects Analysis) - which should never be confused with FEMA (Federal Emergency Management Agency). Although, as you know, this agency has no shortage of quality issues of their own that need to be addressed (it might help too if that agency was funded appropriately - IMHO).
I attended a seminar recently and listened to a speaker from the Quality Support Group, Inc who spoke about FMEA. His lead-in regarding how the tool is used to prevent loss of life caught my attention pronto - as he was not talking about a missed delivery date, but rather in terms of human casualties.
The focus of this tool is determining Risk. And, to level set the definition bar here, let's revisit what Risk is. Risk is 'a potential future event that could result in adverse and unplanned consequences'. Risk is NOT 'a problem, issue or crisis'. Risk IS 'a measurement of the potential inability to achieve objectives*'.
In many quality and project management circles, the idea of Risk Management is bandied about with a whole lot of fervor (yes Virginia, there are geeks in every profession), but much of that hoopla might be over such areas as the Risk to insurance premiums when some insurance company book worm makes an error on an actuary table. However, FMEA is used when real Risk is involved, you know the kind where if the product doesn't work right, you're pushing up daisies and your hard earned assets are being redistributed to next of kin who, hopefully, have better luck with that product in the future.The FMEA tool is really quite remarkable. It has no less than 14 prime categories and multiple sub-categories for determining and tracking Risk entities. But in keeping with my 2010 credo of 'just-get-to-the-point' brevity for Quality Sphere articles, I will touch upon a few key areas of the tool, then I promise to attempt to make sense of it all by the end of the dirge.
Ok .... time to get into the weeds for just for a moment. Key areas:
#1: Potential Failure Mode (types of failures):
When manufacturing anything from medical devices to baby car seats, a great deal of resource energy and time is expended on determining every type of failure for every component or material used and every step in its construction for that product. This evaluation is critical, cannot be done in a vacuum, and it depends on experts who know what the heck they are talking about for every part of the end-to-end development of the product.
#2: Potential Effects of Failure (how bad can it be?):
Again, a critical step that must be done with experts involved. Along with this analysis is listing reasons for why it could fail (proactive root cause review (RCA)). Only through this step can you determine the probability of detection controls that may successfully deter the failure from occurring.
Next, the quantitative portion of the tool takes over to determine severity, probability and occurrence for each potential failure. Based on the collection of data gathered, a Risk Priority Number (RPN) can be assigned and then plotted in a probability/severity matrix chart called an ALARP (As Low As Reasonably Possible). Catchy name ... Huh.
Why did I take you for a quick walk on the wild side of esoteric quality measurements - it's not like most of us deal in the development of products that, if improperly built, can punch your ticket to The Farm? The point is that wouldn't it be interesting if on any project or initiative, we spent some time in the planning stage with SMEs that represent the end-to-end process brainstorming for potential risks - then, build a mitigation plan around the input provided?
Classic project management says that a Risk Management Plan should always be just one of several plans for any project. In real life, project needs are moving too fast, tend to be not very large in scope or size, and frankly, no one wants to talk Risk - as they are seen as 'speed bumps' on the way to an end result. Risk is usually given an obligatory nod - then we move on thinking everything will be .... 'fine'. Risk only seems to count when there are large amounts of cash or causalities on the line (Toyota pushed the envelop big time and are now trying to find their way back to the promised land).
Please let me know your experiences with attempting to develop Risk Management Plans.
Next article: Quality is Everywhere
* Reference: Risk Mgt Guide - DoD Acquisition,4th Edition
