Failure Mode and Effect Analysis - FMEA

Failure Mode and Effects Analysis (FMEA) is a model used to prioritize potential defects based on their severity, expected frequency, and likelihood of detection. An FMEA can be performed on a design or a process, and is used to prompt actions to improve design or process robustness. The FMEA highlights weaknesses in the current design or process in terms of the customer, and is an excellent vehicle to prioritize and organize continuous improvement efforts on areas which offer the greatest return.

The process is very straightforward, and begins by identifying all of the probable failure modes. This analysis is based on experience, review, and brainstorming, and should use actual data if possible. New designs or processes may not have actual historical data to draw from, but "proxy" data may be available from similar designs or processes.

The next step is to assign a value on a 1-10 scale for the severity, probability of occurrence, and probability of detection for each of the potential failure modes. After assigning a value, the three numbers for each failure mode are multiplied together to yield a Risk Priority Number (RPN).

The RPN becomes a priority value to rank the failure modes, with the highest number demanding the most urgent improvement activity. Error-proofing, or poka-yoke actions are often an effective response to high RPN's.

Following is an example of a simplified FMEA for a seat belt installation process at an automobile assembly plant.

As you can see, three potential failure modes have been identified. Failure mode number two has an RPN of 144, and is therefore the highest priority for process improvement.

FMEA's are often completed as part of a new product launch process. RPN minimum targets may be established to ensure a given level of process capability before shipping product to customers. In that event, it is wise to establish guidelines for assessing the values for Severity, Occurrence, and Detection to make the RPN as objective as possible.