MEASURE FOR MEASURE
A Clearer Picture With RCA
Root cause analysis is vital to an effective quality assurance program
by Craig A. Niemann
The first part of this two-part article described how to lay the right foundations for an effective improvement-based quality assurance program (QAP) in a calibration laboratory (March 2008, p. 52).
Embracing the philosophy, choosing the sampling method and selecting the right personnel to manage the QAP laid the groundwork and set the stage for its success.
The next step is conducting root cause analysis (RCA) to examine the individual nonconformities, and then periodically performing trend analysis to determine the most significant nonconformities along with the associated root cause.
The aforementioned steps, in conjunction with the right foundations, should put your calibration laboratory on the road to a process-based approach to improve laboratory operations and manufacture a better product for the customer.
Don't forget your code
Using a system of codes to classify the nonconformity and the associated root cause will help in analyzing data and will assist management in determining and correcting the most significant trends in the laboratory. But be careful not to dilute the data by making the codes so specific that management cannot identify any laboratory trends.
For example, one nonconformity might have the wrong date on the calibration certificate, while another might have incorrect charted values—two different errors, but both relate to documentation. Many of the classifications will end up being a best-fit scenario instead of something that accurately describes the nuances of the individual nonconformity. Like the choice of how to sample the end-of-the-line product, this is laboratory and workforce dependent.
Effective RCA is a significant milestone in the development of an improvement-based system and is probably the hardest to achieve. Nobody likes to be told they are wrong, so QAP personnel need to make sure the focus is on the process and not the person.
This will help establish an open and honest atmosphere during investigation of the nonconformity. After all, the ultimate goal of an effective QAP is to prevent a recurrence of nonconformities. An effective RCA can make that happen.
Enter the observatory
The first step in conducting RCA is to gather data, which can be done effectively by observing the technician recalibrate the item or perform the suspect process without divulging the nonconforming condition. There are many things quality personnel can and should look at to determine whether an item has a quality nonconformity. A few questions to ask are:
- Is the item safe to use?
- Does the item have any physical defects that affect operation?
- Does the documentation have any discrepancies that affect quality?
Depending on the nonconformity and the unit being tested, QAP personnel can have the technician perform the entire calibration or just the measurement functions that did not meet required tolerances. The RCA process is enhanced if the technician repeats the normal calibration process as closely as possible. This will help determine why the out-of-tolerance condition occurred.
The technician might have used a different standard or substituted methods resulting in a faulty in-tolerance reading during calibration. This is one reason QAP personnel should not assume too much when they discover what appears to be a nonconformity. There is a possibility the unit is actually within specifications and the reviewer is performing the step incorrectly. Everyone is human, QAP personnel included.
Ideally, observation of the technician performing the calibration will answer why the nonconformity occurred, but this is not always the case. Conducting a preliminary RCA session immediately after the technician performs the process will help focus the efforts of data collection for the RCA meeting. Data such as item reliability, calibration procedure changes and reliability of the calibration standards might yield insight to the true root cause and enable a more effective corrective action.
Talk it over
Once all the needed facts are gathered, an RCA team consisting of the technician, supervisor and a representative from the QAP should meet to discuss the nonconformity, nail down the root cause and come up with a corrective action. Because time can make memories fade, the RCA meeting should be held as soon as possible after the discovery of the nonconformity.
At the meeting, the QAP representative should act as facilitator to keep the team focused on process improvement instead of placing blame. All the work done so far is wasted unless technicians, supervisors and QAP personnel find a way to look at the nonconformity objectively and determine the root cause.
Several different techniques—such as five whys, cause and affect diagrams, and brainstorming—will help the team reach logical conclusions and discover the root cause of the nonconformity. In the end, the technique used is not as important as the conclusions reached by all of the stakeholders. It is critical that the RCA team identify the reason for the nonconformity and develop effective corrective action.
Just as consistency is important when identifying a nonconformity, it is also important in RCA. When the team members identify what they think is the root cause, QAP personnel should make sure the root cause classification is consistent with previous decisions made as part of the QAP. Otherwise, management will use flawed data during trend analysis and will not be able to attack the most prevalent root cause.
It might help the RCA process to keep key leadership personnel separate until everything is complete and a corrective action is ready for implementation. Leadership should review all the documentation as a last sanity check.
Some teams can lose focus during the RCA meeting and end up attacking the symptom of the nonconformance as opposed to the root cause. A final review of all the documentation by leadership should prevent this.
Personnel should also beware of checklists. If your technicians are required to follow a checklist when calibrating items, do not turn the list into a dumping ground for every corrective action. Some checklists tend to be a two-sided, single-spaced document in eight-point font. An extensive checklist ceases to be a useful tool and becomes something the technicians pretend to follow only when someone is watching them.
Follow the trends
Trend analysis is the final step in a truly effective, improvement-based QAP. On a regular basis, management needs to review all the data collected by the QAP to look for nonconformity and root cause trends. Then it needs to come up with effective solutions to try to correct the most prevalent trends.
The focus of a trend analysis meeting is the bigger picture or macro-level view of the laboratory, similar to a management review. The QAP, in concert with the technicians and supervisors, will correct the individual nonconformities, so it's important not to get hung up on the corrective actions for each one.
Instead, look at the data collected by the QAP and use your imagination. It might help to have a standard breakdown of the data for every trend analysis session. That way, it is easy to see whether the previously applied corrective actions are having any effect.
After the standard breakdown, you are limited only by your imagination. For example, on which day of the week do most nonconformities occur? When breaking down that data, you might come up with Wednesday as the biggest day for nonconformities, not Monday or Friday as originally suspected.
Other things you can look at are the skill level, bad actors and calibration areas with the most errors. This process can and must be tailored to the workforce at your laboratory.
In addition, it is important to review the corrective action from the previous trend analysis meetings. You will find that some of the applied corrective actions have no effect. If that's the case, management should evaluate whether it wants to continue to spend the resources and adjust the corrective action or just abandon it all together.
Lastly, strive to make your corrective action SMART—specific, measurable, attainable, realistic and timely. Try to avoid corrective actions such as merely briefing the technicians, which in the long term will not curb current nonconformity or root cause trends. Stick to SMART actions, and an improvement-based QAP can lead to better processes and products.
Craig A. Niemann is a senior master sergeant with the U.S. Air Force stationed in Heath, OH, and has served as a quality process evaluator and quality program manager. He received his associate's degree from the Community College of the Air Force in 1997. Niemann is a senior member of ASQ and a certified calibration technician.