Root Cause Analysis

During the past weekend, I spent some time working on one of the rose beds around my house. Generally, this has been a beautiful bed with compact, dwarf shrub roses that seem to always be in bloom. However, this bed recently fell prey to several insidious plants, including Morning Glory, wild blackberry, and other types of weeds. Because of this infestation, I spent a while pulling these pesky varmints of the flora world out of the rose bushes, finally reclaiming an attractive bed. As I pulled these vines, I traced the tendrils back to their common root so I could eliminate the weed rather than break it off. With that, a thought occurred to me – this is what we do when we perform Root Cause Analysis or Lean Root Cause Analysis on a pesky problem that is facing us.

Problems—whether they are issues with quality, safety, or productivity—can be like the bane of the garden weeds and vines. They can grow among the good things that we are doing, seeming innocuous, and before we know it, these weedy issues choke out all the good progress we have made. Perhaps a line failed due to a major breakdown, or one of our employees got hurt in the plant. If we try to solve the problem without knowing and correcting the root cause, the problem will resurface, and sometimes can be worse than before. If we only solve the symptoms and not the root cause, the issue will return, and sometimes worse than the first.

Similarly, if we remove the vines of the weed without removing the root, the weed will regrow and re-infest the garden where it was found. So, how do we pull problems out by the root? We discover the root of the problem through Root Cause Analysis and dig the root out.

What is Root Cause Analysis?

The American Society for Quality (ASQ) defines Root Cause as a “factor or factors that cause[s] a nonconformance and should be permanently eliminated through process improvement.” ASQ then continues to define Root Cause Analysis as the “collective term that describes a wide range of approaches, tools, and techniques used to uncover causes of problems.”

Much like pulling a pesky weed out of a garden, root causes are often difficult to find and can be even more difficult to correct and eliminate. Every industry discipline, whether it is manufacturing, service, healthcare, warehousing, or transportation, faces problems. And each problem contains a cause—a root cause—which, when eliminated, will solve the problem at hand. Root cause analysis, as well as Lean Root Cause Analysis, is the first phase of solving problems. It is like finding the root of the weed in the garden.

When we pull weeds out of the garden, we use tools. These tools can be simple or complex, depending on how deep the root is or how complex the root system is. In Root Cause Analysis, the process is very similar. There are steps and tools that we use to discover the cause of the problem and find the solution.

Application of Root Cause Analysis in Manufacturing Performance Systems

Take this as an example: a worker slipped in a puddle of water on a loading dock and broke his arm. The group investigating the incident found that the cause of the broken leg was two-fold: First, the worker was not paying attention to where he was going. Secondly, there was a puddle on the floor. The solutions for the incident were: the worker was disciplined and retrained in proper walking procedures, and the puddle was cleaned up. A week later, another worker was walking through the area, slipped in a puddle in the same area, fell, and fractured his hip. The puddle was a symptom of a water leak in a pipe overhead – the REAL problem. Root Cause Analysis would have solved this problem and prevented the worker’s hip fracture.

Root Cause Analysis Approaches

Is Root Cause Analysis a highly technical, difficult task to master? It depends. Does Root Cause Analysis employ special skills that are possessed by only a certain few? No. Root Cause Analysis can range from the very simple to the very complex. Despite the existence of different difficulty levels, anyone can perform good Root Cause Analyses and get good solutions to most problems by following a systematic procedure.

Root Cause Analysis ranges from the very simple to the highly complex. The skilled analyst, like the skilled gardener, will have many tools in his or her toolbox, but the most effective tool is often the simplest. Just like a gardener would not use a power shovel to pull a weed or a trowel to dig up a tree, the tool the cause analyst uses to approach a problem should be the simplest tool to get the job done. Often, the most effective tool is the one that is most used. We will discuss in this article some of the more simple and effective tools.

Root Cause Analysis Process

The first step of Root Cause Analysis is defining the problem. Many times, we see a problem and only partially define it. Worse, we only define a symptom. When we use the term “problem”, we typically mean, “a matter or situation regarded as unwelcome or harmful and needing to be dealt with and overcome” (Various Authors, 2020).

Correctly defining the problem rather than a symptom is critical for effective Root Cause Analysis. We define the problem using a problem statement. A problem statement:

• is an effect, not a cause
  • Flat Tire, Broken Arm, Temperature out of spec
• is an output of a process
  • An overheated, worn part
• is a WHAT statement
  • WHAT is wrong
• should not be WHY biased
• should be observable (or measurable)
  • “I see the…” Not “I think…”
• should be able to collect data objectively

Problem statements describe the problem concisely and completely. Problem statements are specific – the more data, the better. For example, if a problem statement was “Product defects were found in Line A”, then we would not know what the defects were, how bad they were, or even where they were in Line A. If Line A was long and complex, we could spend precious time trying to find where the defects were originating. This time could cause Line A to make more defects.

A good problem statement contains (1) What happened, (2) Who discovered the problem or who was affected by the problem, (3) How bad was the problem, (4) Where it happened, and (5) When it happened. A better way of stating the problem is “B Crew found 5000 wrong size parts (too big) in line A, process #3, in December.” Specifically stating the problem helps the problem-solvers focus on the area and begin building the Root Cause Analysis.

Once we have clearly, concisely, and specifically stated the problem in the “Who, What, When, Where, and How Much” format, we can then begin the quest of finding the root cause, or WHY the problem happened. As I have stated earlier, there are two main categories of Root Cause Analysis tools: Simple and Complex.

Simple Root Cause Analysis tools include Process Flow Diagrams, Fishbone (or Ishikawa) Analysis, and Why-Why (sometimes called 5-Why) Analysis. Each of these analyses has its own particular strengths and weaknesses which we will discuss further in this article.
Complex Root Cause Analysis works to solve more complicated and deeper problems. The tools used in complex root cause analysis rely on more powerful and deeper statistical tools. These tools include scatter plots, histograms, Pareto’s Analyses, design of experiment, and multi-variate analyses. These tools can become very large and unwieldly making the solving of a problem seem out of reach of all but the most sophisticated of statisticians.

The vast majority of problems can be solved using the simpler tools in root cause analysis. For that reason, we will focus our discussion on the simple tools and their use.

Process Flow Diagrams

Process Flow Diagrams are a pictorial representation of the process that includes where the problem occurred. Typically, these diagrams show the components of the process, giving enough details to clarify the location of the problem and perhaps some of the mechanisms around the problem area. Some texts instruct to make the endpoint of the Process Flow Diagram at the point of the failure or error and look backward while keeping the focus on those processes that can be controlled internally (Okes, 2009). For example, if your problem was caps not going on bottles correctly, you would not want to look at the bottle manufacturer or cap manufacturer until all of the processes in the capper area were checked first. Below is a simple Process Flow Diagram of the capper process:

We can then dig deeper into the areas where the problem is likely to have its roots and ignore other areas that are not related.
One advantage a Process Flow map has is its ability to be used as a Measles Map (another Root Cause Analysis tool). With a Measles Map, the analyst places a dot where defects or issues are found. The focus of the problem-solving team should be on the area where the majority of the dots lie.

Fishbone Diagrams

Ishikawa’s Fishbone Diagram is another popular tool used in basic Root Cause Analysis. The Fishbone Diagram was created to group possible causes of a problem into categories (Ilie & Ciocoiu, 2010). The diagram, also called a “cause and effect diagram”, is useful in helping the problem-solving team see potential causes in a structured method.

Typically, we use the 4 (or 6) Ms as the categories. The Ms are: Man (or huMan), Method, Machine, and Material. The other Ms to complete the 6 are: environMent and Measurement. With these categories, the team can brainstorm potential causes, and place those causes in their respective categories using Post-It notes.

The above Fishbone Diagram references a Root Cause Analysis completed on the cause of a flat tire. The original 4Ms were used.

Why-Why Analysis

The third of the basic Root Cause Analysis tools is the Why-Why Analysis or the 5-Why Analysis. This tool, developed in the 1930s in Japan, uses a question-and-answer approach to determine the ultimate root cause(s) of a problem. This process begins with the problem and asks the question “Why” until the question cannot be answered further. On average, the Why question is asked five times until the cause is found.

This tool is most effective when the team can go to the floor and discover the answer to a “Why” question while the answer is not yet known.
This technique is not without drawbacks, however. To be effective, the Why-Why analysis should not have an arbitrary number assigned to its title. To use the term “5-Why Analysis” forces the problem-solver, subconsciously, into stopping at five levels rather than exploring more or fewer levels as needed. To insinuate that every problem will be miraculously solved after asking “Why” five times can be a bit irresponsible. Many problems have multiple layers to work through.

Another potential drawback of the Why-Why Analysis is confining the analysis to a single form or sheet of paper. By doing this, the analyst is confined to one linear way of approach where the problem could have several branches. A better option for this is the use of Post-It notes and a white board or flip chart paper on a wall. Once the notes are attached to the board, lines can be drawn to connect the levels. Below is an example of the Why-Why Analysis.

This simple example is 4 levels, and the causes of the lack of calibration. As such, the dryer not at the correct temperature could be addressed.

Sometimes when the question “Why” is used, the respondent can see the question as attacking. The respondent can become defensive, and this creates a barrier in the process. So, people often have had negative experiences when the question “Why” is posed to them.
Starting a question with the word “Why” can also imply fault. A way to overcome this is to change the question. The question “Why” can be changed to

“What caused that”. In our above example, instead of asking “Why is the quantity of potassium nitrate lower than the specification?”, the question can be better asked, “What caused the potassium nitrate to be less than specification?”. Changing the wording can improve the outcome and reduce the defensive barriers that may exist during this process.

The ultimate goal of the Why-Why Analysis is for the team to arrive at causes that are actionable for the creation of countermeasures.

Challenges of Root Cause Analysis

Root Cause Analysis can present several challenges in solving problems. First, analysts can be tempted to settle on the first or earliest potential answer that surfaces. When this happens, the analyst or team is likely finding the answer to a symptom and not the real problem. This is especially true in the Why-Why Analysis because we can become uncomfortable asking or answering the “Why” questions. As a result, the analyst can reach a potential answer too early. The remedy for this is to keep asking and digging until the full answer is known. This can take time and effort on the part of the analyst and team.

These tools can be used separately or used with other tools. Very small problems can be solved by using only the Why-Why or Fishbone Analysis. However, the real power of the tools and the overall power of root cause analysis is using multiple tools to solve a problem – especially a stubborn, thorny issue that has wound its way into the process. There is an adage that says that if we only have a hammer in our toolbox, we see each problem as a nail. If we rely on one tool, we may miss vital information that would help solve the problem better. Analysts should be adept at using all of the Root Cause Analysis tools available and employ them when solving problems.

Conclusion

These simple tools can be used separately or collectively to solve problems. Most problems can be effectively solved with these tools. Much like weeding a garden, a good problem-solving process gets to the root of the problem and digs the root out, eliminating the problem completely.
No problem can be solved in a vacuum, however. Teams must work together to fully define the problem and get to the cause of the problem. The team can then develop effective countermeasures to solve the problem and keep the problem from reoccurring.

References

Ilie, G., & Ciocoiu, C. (2010, March). APPLICATION OF FISHBONE DIAGRAM TO DETERMINE THE RISK OF AN EVENT WITH MULTIPLE CAUSES. Management Research and Practice, 2(1), pp. 1-20.

Okes, D. (2009). Root Cause Analysis: The Core of Problem Solving and Corrective Action. Milwaukee, WI: ASQ Quality Press.

Various Authors. (2020). Google’s English Dictionary, Oxford Languages.