FMEA (Failure Mode and Effects Analysis) is one of those things most people do without actually realising it, albeit often in an informal and unstructured manner. What could possibly go wrong?To an HSE professional it’s risk assessment (and management); to a quality management professional, it’s a technique that should be far more commonly used in a structured manner.
FMEA is a systematic evaluation of what could go wrong with a product, process, service, etc., how that would happen, how critical it would be, how it can be avoided and what can be done if it actually happens. It’s sometimes defined as inductive reasoning single point of failure analysis (or forward thinking).
I’m not going to go into much detail of the tools as there are numerous books, journals, courses and online sites that will cover the range of FMEA variants around, providing forms, formulae and other tools to assist. To me, the most important considerations are:
- Conducting an FMEA on a scrap of paper and hoping to get it right is wishful thinking. A formal worksheet is almost a necessity for all but the most simple case.
- Set ground rules. Define the working environment, inputs, expectations, resource availability, users, etc.
- Ensure the necessary skill sets are available. FMEA is rarely a one-person job.
- Take formal, structured approach, one failure mode at a time – and follow it through.
- Be clear on assessing criticality and risk in order to prioritise subsequent actions.
- Don’t focus so hard on detail that the bigger picture is lost – occasionally take a step back and assess where you are.
- Consider change over time (where relevant). This is particularly important with products that degrade in use (or just over time) – corrosion, wear, battery life, user experience, fatigue (of the product and the user), etc.
- Don’t assume you’ll catch everything with FMEA. Consider complementary tools, such as Fault Tree Analysis, as well.
FMEA is often time-consuming. If not planned into a development project it may well extend the completion date, especially if (when) additional work is identified as necessary to reduce failure risk to an acceptable level.