ETQ Root Cause Analysis

Root Cause Analysis (RCA)

Root Cause Analysis: Ensure Lightning Doesn’t Strike Twice

Process failures, complaints, nonconformities—these quality issues can all happen in any manufacturing environment for many reasons. And like lightning, you don’t want the same issue to strike twice. Or better yet, strike at all!

Incorporating a thorough Root Cause Analysis (RCA) process into your quality management system is essential to minimize or eliminate the probability of repeat issues.

Let’s get familiar with the details of a Root Cause Analysis:

  • What is Root Cause Analysis?
  • Why conduct a Root Cause Analysis?
  • How do you perform an RCA?
  • The 5 steps of a Root Cause Analysis
  • Common Root Cause Analysis tools
  • Most common RCA mistakes
  • Why quality management software can make a big difference
  • How analytics from ETQ Reliance makes RCA easy

 

What Is Root Cause Analysis?

Root Cause Analysis is a problem-solving method to isolate the originating factor, or factors, of quality issues.

The term encompasses a range of techniques to define, understand, and solve issues across the spectrum of people, processes, and tools that make up your quality management system. For example, a manufacturer might face supplier material nonconformances, a manufacturing defect, an accident at a job site, or a failure to reach production goals. Though dozens of factors may contribute to these problems, they typically cascade from one or two sources at most.

The root cause is typically a single core issue or factor that triggers the cause-and-effect chain of events leading to the problem. However, depending on the complexity, several factors may constitute the root of a problem.

As a rule of thumb, a factor is a root cause if removing it prevents the problem from recurring. A contributing factor is a condition that influences the likelihood of that problem.

Why Do Root Cause Analysis?

Conducting thorough and unbiased Root Cause Analysis is your ticket to less waste, more productivity, and better product quality. You can explicitly measure the incentives by the cost-of-poor-quality (or cost of failure) outcomes that matter to your organization.

The costs could be time and money lost in scrap and rework, time and money lost on customer complaint handling, lost time and revenue getting products to market, time and money lost on recovering your brand’s reputation, or myriad others.

Root Cause Analysis is integral to the corrective action and continuous improvement concepts outlined in ISO 9000/9001 and all-around good practice even if you aren’t seeking certification.

Establishing an RCA process is a stepping stone toward resolving quality events, deviations, and failures consistently. You’ll also gain insights that lead to increased productivity and higher employee morale.

Here’s a quick guide to getting started with RCA.

How Do You Perform Root Cause Analysis?

There are several methods for conducting an effective RCA. Choosing which to use depends on your organization and the complexity of the problem you are trying to solve. For example, an isolated workstation training issue will likely have a more obvious cause and straightforward path to resolution versus handling a product recall.

Root cause analysis is also multidisciplinary because apparent problems with a product or process can originate in almost any part of the organization. That can initially feel daunting—there’s much ground to cover across the product lifecycle.

The most popular analysis methods, discussed in more detail below, give you options to handle almost any level of complexity behind the quality events you encounter.

The following 5-step process will chart your course and help you set expectations with your stakeholders.

5 Steps To Plan Your Root Cause Analysis

Step 1: Define the problem

Begin at the beginning by identifying and documenting the specific symptoms and any warning signs related to the problem. To accomplish this, assemble a balanced, cross-functional team with members who can provide various perspectives on what happened or is happening. This proactive collaboration is important for information discovery and the buy-in you’ll need later for the action plan.

Step 2: Measure the impact

Next, you’ll want to understand how the issue affects the organization. Identify the duration and severity of any quantitative or qualitative impacts and risks, like unplanned downtime, diminished customer trust, compliance violations, worker safety, or all the above. This step will help you set measures to evaluate the effectiveness of your action plan later in the process.

Step 3: Identify contributing factors

By asking probing questions (who, what, where, when, why, how), you’ll have a list of possible factors to address. Ideally, also pull in historical data on similar issues or complaints from global company data or data from other sites. The more relevant information you gather, the better you can contextualize the conditions that led to the quality event and assess whether it’s a single incident or part of a pattern of similar issues.

A single problem can have multiple causes, and the goal of the next step is to identify the root one.

Step 4: Isolate the root cause

Place the collected data into your chosen root cause analysis framework, such as a Five Whys, Fishbone, or Pareto chart. The output should help determine the quality event’s primary cause and shape your action plan. When failures occur, it’s easy to lean on assumptions or attribute the cause to a perceived correlation. As the saying goes, correlation is not causation—more on this in the Five Whys section below.

Step 5: Develop a solution

“Develop” is the keyword here. You’ve gathered lots of data and performed the right level of analysis to propose a well-informed solution that eliminates the recurrence of the root issue. Even so, the action plan may need to evolve or change as you test and implement a new process. Also, key to your solution is monitoring its effectiveness. You’ve come a long way in this analysis and don’t stop now. Be sure to check back on the results. Is the issue resolved? If not, cycle through the RCA process again.

Common Root Cause Analysis Tools

There are a handful of industry-standard tools that help you identify the root cause of an incident. Depending on the scope of your analysis or requirements to track and trend data, you may use just one or a combination. ISO 9000/9001 guidance does not prescribe a particular method, so choose the appropriate tools for your situation.

The most important thing is to put these time-tested methods to work to improve your quality programs steadily over time.

Five Whys

Five Whys is by far the most popular root cause analysis framework. It’s excellent for simple to moderate problems and combines well with other analysis tools. As the name suggests, the process invites your cross-functional review team to ask why an incident occurred until no other possible causes are left.

The method helps cut through the blame game and objectively zero in on a single causal factor. The replies to why must be factual, backed by data, and focused on process or system errors.

Keep applying the process as many times as you need. As you question, you may reveal factors that need their own Five Whys branch. Keep in mind that new branches may signal it’s time to shift your analysis into a Fishbone diagram.

By the way, you can ask why more or fewer than five times!

Fishbone Diagram

Fishbone diagrams are structured to handle complex problems or systemic issues. This method allows you to capture, organize, and prioritize your data to create a comprehensive plan addressing every aspect of an important issue.
Place the problem from Step 1 above at the end of an arrow (or head of the fish). Each branch is a category of contributing factors, with individual factors placed along the corresponding branch.

Fishbone Diagram Root Cause Analysis

Pareto Chart

The Pareto chart applies the 80/20 rule to help you identify the vital few areas where you can achieve the most significant quality improvement. For quality professionals, the rule postulates that 80% of the problems come from 20% of the issues. The sorted histogram chart depicts the number of occurrences in each issue category in descending order. The cumulative percentage line helps mark what categories fall in the 80%.

When your issue categories are well-defined, you will see a steep cumulative percentage line that quickly tapers to horizontal.

Pareto Chart Root Cause Analysis

Additional RCA Methods

If you’re ready to up your RCA game, then consider these additional methods:

Scatter Diagram

A scatter diagram is a graphic presentation of the relationship between two variables where you suspect a cause-effect correlation. For example, you might plot data from nonconforming material incidents against hours of unplanned downtime. From the resulting pattern of data points, you can discern if there is a strong, weak, or complex correlation – or no correlation at all.

Failure Mode Effects Analysis (FMEA)

Although not a root cause analysis tool per se, FMEA is a systematic method for quality managers to proactively analyze all possible paths to product or process failure and assess the relative impact of those failures. The result is discovering areas with the most risk, which you can act on to prevent potential issues.

The best method for performing and documenting your Root Cause Analyses depends on your organization. Keep it simple with a Five Whys approach if you’re just starting out. Over time, you can experiment to determine what method or combination of methods work best for your teams. You’re already ahead if you are thinking about the value RCA can bring.

Most Common RCA Mistakes

Root Cause Analysis can be complex and involve many stakeholders inviting some common pitfalls into the process. Here are a few of those mistakes to avoid:

  • Rushing the process. Resist the temptation to lean on assumptions or attribute the cause to a perceived correlation. Taking just a few minutes for a Five Whys discussion can go a long way towards uncovering the TRUE root cause and solving the problem for good.
  • The wrong people are involved. It’s good practice to conduct your analysis as soon as possible, but don’t sacrifice expertise or unbiased perspectives over immediate availability. Memories can fade quickly, and one countermeasure is to record or transcribe interviews with involved personnel within days or hours of a severe incident for later review.
  • Accepting “operator error” as a root cause. Although it’s possible, simple operator error is probably not the root cause. Keep digging. Is there something wrong with the equipment? Is it something in the environment? A bad procedure? If so, why is the procedure bad?
  • Not following up on solution effectiveness. You can use good old-fashioned SMART* goals to help define success. Think of it like a coin. One side is your analysis and proposed action plan; the other is ensuring your actions solve the problem or don’t introduce new ones. For management buy-in, communicate the value of your corrective actions in terms of business impact, such as cost savings on rework or scrap. (*Specific, Measurable, Actionable, Realistic, and Time-bound.)

 

60 second snippet – Learn when can root cause be less effective. Watch entire webinar.

RCA Toolkit

Root Cause Analysis Toolkit

Download our new guide to learn how to perform a root cause analysis and put the three essential root cause tools into practice

The Role of Quality Management Software

Maintaining a closed-loop quality process for root cause analysis and corrective action is your best defense against recurring quality issues. It’s impossible to manage this activity and more through spreadsheets and email in today’s complex business environments. QMS software is the best way to drive process improvements across teams, secure your documentation, and maintain a central repository for quality data.

Technology like Reliance NXG allows you to build, connect, and automate quality processes across the product lifecycle – from product design to production to customer use. As a quality manager, you become a strategic partner with teams across your organization, providing data-driven insights to improve quality, increase efficiency, and reduce costs.

One of the key requirements when we bought ETQ was that we could use it for supplier management. Our supplier quality engineering group is extremely happy with ETQ. We are going after a market that requires zero defects (automotive market). It's a different standard.

Steve Bergeron
Director of quality for US manufacturing and
materials and QMS infrastructure, Cree Corporation
Fichier 1