Failure Modes and Effects Analysis (FMEA) is currently the most used technique in risk analysis. Risk analysis is a very natural human activity. Imagine yourself as a hunter-gatherer charged with organizing the evening meal. Together with your team, you organize a hunting party and part of that plan will be some intuitive risk analysis. What are the dangers? You may consider predators, thirst, and injury during the chase. All being well the precautions you take result in a successful hunt, everybody returns home in one piece and the tribe eats their evening meal.

Risk analysis remains an essential part of our business activity. We routinely analyze situations and take actions to minimize business risks and safety concerns. FMEA is an effective risk analysis method for design and manufacturing.
It examines your design and manufacturing processes and identifies the opportunities for marginalities and defects which can result in customer dissatisfaction.

Before we go into the detail why not carry out a simple trial. Walk down your production line and think to yourself “What can happen to the product here that will impact the customer? What are we doing about it?” Ask the production staff what routinely goes wrong with the process and think “How can that affect the product”. The FMEA will capture all this knowledge and experience in a structured way and will allow you to take actions which will minimize the risk of a disappointed customer.

A brief history of FMEA
FMEA was developed by the American military at the end of the 1940’s. I understand their frustrations with munitions malfunctioning led them to develop a methodology that would eliminate all the potential root causes. A detailed method was documented: MIL-P-1629.
It worked and so it was adopted by the nuclear and aerospace industry. NASA credited the success of the moon landings to its use. NASA was also concerned about contaminated food on space missions that they developed a very similar technique called HACCP(Hazard Analysis and Critical Control Points) specifically for the food industry.

In the 1970’s slow speed shunts involving the Ford Pinto were resulting in fatal fires if the petrol tank split. As you can imagine this was a public relations disaster. To eliminate a re-occurrence for this and other serious problems the Ford Motor Company implemented FMEA in their design process.

More organizations were using FMEA as a risk analysis tool and in 1993 the AIAG (Automotive Industry Action Group) incorporated it into the QS9000 standard for automotive production and its suppliers. QS9000 became TS16949 and since November 2016 IATF16949. While FMEA was specifically aimed at automotive, other industries which demand the highest levels of reliability, for example, semiconductors and oil and gas, have implemented it. These reliability standards are now extending to white goods and everyday electronic products. Strictly the current format is FMECA (Failure Modes and Effects and Criticality Analysis), the Criticality of the defect for the end user is incorporated into the method.

For practical purposes it is a risk analysis and defect reduction technique that takes into account three things; the severity of the defect for the end user, the occurrence of the potential root cause and our effective detection or elimination of this root cause.

Getting started with Risk Analysis
An important part of any risk analysis is the scope of the process. This may sound trivial but it is essential to know if your analysis should include, for instance, the incoming materials store. Then we need to make a complete process flow. This is a crucial task and should include every step in the process including rework steps. At each of the process steps, we need to identify the possible sources of variation and potential problems. We need to document defects that may occur on the product.

We also need to list equipment and process problems that may cause defects in the product. In a risk analysis process, the creation of a complete process flow and a complete list of potential problems can take 50% of your time.
Where do the ideas come from?
Traditionally a Brainstorming technique has been used to generate a list of possible problems at each process step or design step. While this is useful it is often more useful to incorporate all your current sources of variation into your initial FMEA. These include:

• Inputs from scrap incidents
• Inputs from customer returns
• Ideas from the designers, machine operators and maintenance team
• If you have SPC then consider out of control points
• Incidents on similar equipment
• Advice from industry experts
• Advice from your material and equipment suppliers

Using these sources will capture a lot of problems and marginalities that already exist and make your FMEA practical and useful from day one.

From IATF16949 FMEA to VDA AIAG
When assessing potential risks in a process it is useful to document the information in a structured FMEA format. The IATF16949 format from the Automotive Industry was widely used and an example is shown in the following graph;

In the first 4 columns, you document the process steps, the process requirements, potential failure modes and the effects of the failures. In the next step, you identify the process problems which can cause the failure and the controls you have in place to prevent or detect the problems. The final columns contain the corrective actions taken and the results of these actions.

In june 2019 the new VDA AIAG format was published. This format supports a 7 step approach:

The 7 steps are as follows:

Step 1: Planning and Preparation. In this step the Header information is filled out and the scope of the FMEA is decided.

Step 2: Structure Analysis. A more detailed breakdown of the manufacturing process is added. Focus Element of the PFMEA: the process step station number and name under review. Next Higher Level: process item system (the overall manufacturing process). Next Lower Level process: work element 4M type (based on Ishikawa approach). This encourages the users to consider the categories of Man, Machine, Material, Method, etc., leading to a more complete list of Failure Causes (FC.)

Step 3: Function Analysis. Added the description of functions and requirements related to the Next Higher Level and Next Lower Level. This supports a clear and complete description of the Failure Effects (FE) and Failure Causes (FC).

Step 4: Failure Analysis. Potential Failure Mode is replaced with Failure Mode (FM) of the Focus Element. Potential Effect(s) of Failure is replaced with Failure Effects (FE) to the Next Higher-Level Element and / or Vehicle End User. Potential Cause of Failure is replaced with Failure Cause (FC) of the Work Element.

Step 5: Risk Analysis. Classification is replaced with Special Characteristics and Filter Code. Occurrence is replaced with Occurrence of the FC. The Occurrence rating now is based on “prediction of FC occurring”, which leads to determining the actual robustness of the Prevention Controls (PC). Current Process Control – Prevention is replaced with Current Prevention Control (PC) of the Failure Cause (FC). Current Process Control – Detection is replaced with Current Detection Control (DC) of the Failure Cause (FC) or the Failure Mode (FM). Detection is replaced with Detection of the FC or FM. Detection is now based on three factors: detection method maturity, opportunity for detection, and ability to detect. RPN is replaced with AP.

Step 6: Optimisation: Recommended Action replaced with Preventive Action and Detection Action. Added the columns: Status (planned, decision / implementation pending, completed, discarded), Action Taken with pointer to evidence, Special Characteristic, and Remarks.

Step 7: Results Documentation: Internal reporting to management and customer reporting.

The Severity, Occurrence and Detection are all scaled on a 1 to 10 basis. 1 is the best rating and 10 is the worst. Multiplying them together gives the Risk Priority Number (RPN). Originally FMEA was aiming at lowering the RPN number but detection is less important than severity and occurrence so severity and occurrence are getting more important. All this information is stored in the FMEA. When a cycle of learning is completed the FMEA will be released. The FMEA is a living document and it will be reviewed and updated regularly. New causes, failure modes and will be added as you learn more about your process.

FMEA software
Most companies start applying FMEA in Excel. Although Excel is really flexible it is not really suited to apply FMEA. The user has to do a lot of extra work to follow the FMEA rules, create the documents, create links between process flow, FMEA and Control Plan but especially maintain the documents.

To get started quickly with FMEA, you can download a trial version of our FMEA Software. DataLyzer FMEA is very cost effective compared to implementing your own system in Excel. The trial version can be used for 30 days so the least it will do is help you get started with FMEA implementation at no cost. This introduction is also available as a white paper. On our Whitepapers page, you can find more background information about FMEA. DataLyzer offers very cost effective online FMEA training. For more information please download the brochure here.