Home >

Complete a Thorough Root Cause Analysis

Contents[Hide]

TOYOTA’S CORPORATE slogan is “Yoi shina, yoi kangai,” which means, “Good thinking, good products.” This applies particularly well to the analysis portion of problem solving. Toyota places high value on the ability to think logically and creatively because a solid thinking process will produce the best results.Every Toyota manager understands, above all, the value of human creativity—that it is the single thing that will set them apart from their competition.

The analysis phase of problem solving should be an exploration into areas previously not understood. It’s a bit of detective work, a bit of scientific experimentation, and an opportunity to discover new things. Analysis is the “Ah-ha” stage, the time to gather evidence, the time to repeatedly ask “Why?” and to find the source of an issue, its root. When the root causes are discovered, the “answers” to solve the problem become obvious. At this time “good thinking” will generate the best solutions—highly effective, simple yet elegant, and low cost, but not shoddy.

As Albert Einstein once said: “The important thing is not to stop questioning.”

Principles of Effective Analysis

Effective analysis is crucial for finding and understanding the many potential causes of the problem. From those potential causes, it’s necessary to narrow the field and focus on the most significant ones. Much of Toyota’s great success stems from the ability to fully analyze a situation and understand the many causes of the problem beyond the most apparent. The following principles are a crucial part of the Toyota approach:

  1. The analysis must not be clouded by preconceived ideas of the problem causes. If the cause is assumed, it will preclude a useful analysis and most likely lead to poor results.
  2. Always follow the genchi genbutsu principle to verify the source of the problem. Do not depend on others, or on data, to find the cause. Use information to point toward the location to “go see.” The point of cause must be observed firsthand.
  3. Analysis is continued until it is certain that the true causes, or root causes, of the problem are discovered (using the “Five-Why” method).
  4. In nearly all situations there are multiple causes for problems, and thus the analysis must be comprehensive. Toyota evaluates causes through the 4Ms: Man, Method, Material, and Machine.
  5. Since there are many possible causes, it’s necessary to narrow to the most significant ones. Narrowing allows the focusing of efforts to generate greater results.
  6. During the analysis, the goal is to identify problem causes that can be corrected by the problem solver. This avoids the tendency to defer the problem to others and forces the question, “What can we do?”
  7. A thorough and complete analysis will yield root causes that will clearly indicate specific, corrective actions. There is an observable and obvious trail leading from the problem to the causes and to the solutions.
  8. Thorough and complete analysis provides factual data, allowing precise prediction of potential results when the causes are corrected. Determining the exact result is an important part of the process since it forces the evaluation of capability and effectiveness in examining a problem.

As with many aspects of the Toyota Way, the thought process is critical to success. Notice that during the following conversation, people will jump to preconceived conclusions rather than recognizing the simple but true answer to the question. Using the example in the problem statement below, we would begin the Five-Why process as follows:

Problem statement: “The fabrication units per hour is below goal.” Upon asking our group “Why?” we might get the following answers:

1.Because the machines break down.
2.Because operators are absent.
3.Because we run out of parts.
4.Because operators are not trained.
5.Because the setup times are long.

Each of these answers may be “true,” as in the conversation between the engineer and the lean sensei described in chapter 14, but they are further down the Five Why chain. The first challenge is to focus solely on the direct question: “Why are the fabrication units per hour below goal?” Then the true answer would obviously be: “Because we do not make enough parts each hour.” Knowing where to focus is crucial in order to train our minds to understand the complete chain. Skipping what appear to be obvious links in the chain will cause jumping to preconceived causes, thus overlooking other possibilities. This is one of the greatest risks and also the greatest challenges in thinking.

TRAP

In many cases we see people attempting to force the Five-Why process into five boxes by trying to “figure out” the correct chain with five “answers.” This process does not fit a predeveloped template format. The causal chain may branch at any level and yield unknown quantities of answers at each. If you are struggling to find Five Whys, most likely you’re jumping across links in the chain. Take time to reflect on the simpler, more obvious answer in order to allow the discovery of all possibilities.

Proceeding with our questioning, we would ask, “Why don’t we make enough parts each hour?” Again the tendency is to skip to the obvious answers, but by approaching this with a different thought process, we would see this answer: “Because we lose opportunities to make good parts.” The production of any product is accomplished by utilizing the time of people and machinery, and available material. In this case there are only two main causes for a shortage of production—loss of time and loss of material (scrap). Note that this line of thinking also maintains a narrow focus that will isolate the most significant causes from the less significant ones. In the example above, the first question led immediately to a lengthy list. Once a long list is established, it’s extremely difficult to narrow the focus. It is much easier to maintain a narrow focus and divide the possibilities gradually through effective questioning. At this point the Five Why chain would look like Figure 15-1.

Problem statement: The fabrication units per hour is below goal.

Why?

We are not able to make enough parts each hour Why?

We are losing production opportunities

Why? Why?

Losing time Losing parts (scrap)

Figure 15-1. Initial Five-Why analysis

At this level the Five-Why chain has developed the first branch. Prior to asking “Why?” for both branches, it’s important to understand which is the most significant. This understanding will maintain a narrower focus. For the sake of this demonstration, we will assume that the data show that scrap is very low and time is the greater loss, and proceed to show the continuation of the causal chain from this level. It is imperative to actually confirm the overall impact of each item, rather than to assume. The scrap quantity data may be available and fairly easy to quantify; however, the time losses will require a visit to the workplace (genchi genbutsu) to verify the amount of time loss.

When asking “Why?” do not jump down the chain to the deeper issues. Carefully consider the loss of time in a production process and try to keep the focus narrow by answering the direct question. Look for the broad categories under which the detailed answers will fall. Remember to use the “Therefore” method if you find yourself answering further down the chain. If the answer “Setup time is too long” arises, state “Therefore” and find the answer. In this situation it would be: “Therefore the machine is not running for a long time.” The following step would be: “Therefore we are losing time.” If the “Therefore” method was used on some of the other issues, it’s likely that “The machine is not running” (or “The line is not running”) would be a consistent theme. This is the common category we are looking for. In addition, our questioning may lead us to understand that loss of time due to excessive process cycle time is also a primary category. Now the Five-Why chain will appear as shown in Figure 15-2.

Again the causal chain is branching. At this time a visit to the workplace is

absolutely necessary. In order to improve your observation ability you must learn to “look with intention.” Based on the analysis thus far, what is the intention of your observation? The intention is to look to see whether there are cycle time losses or situations during which the process is not operating.

The general thinking within Toyota is to consider the cycle losses first. Cycle time losses are those losses that occur every cycle as the operation is performed;

Problem statement: The fabrication units per hour is below goal.

Why?

We are not able to make enough parts each hour Why?

We are losing production opportunities

Why? Why?

Losing time Losing parts (scrap)

Why?

Why? Cycle time losses

(Not most significant)

Process not running

Figure 15-2. Second pass Five-Why analysis

therefore, they have a “high tendency” of occurrence. The cumulative effect of these small losses can be very great. In addition, the reduction will generate an immediate and continuous payback. A small payback that can be captured immediately and will continue to pay forever is a preferred result. Small cycle time losses are also generally easy to correct. They may include excessive operator or machine motion, delays due to waiting, or overprocessing (doing more than necessary). Of course, these are all forms of muda (waste), and the removal of muda is a primary objective.

Visiting the workplace, you will probably see many other examples of cycle losses and process stoppages. You’ll need to gather facts to understand the total impact of each issue—the importance, urgency, and tendency—and a simple way to do this is to use a value-added/non-value-added1 breakdown list as shown in Figure 15-3. The example is from a sawing operation, but the list generated is fairly typical in most manufacturing operations. Remember, the links of the causal chain were related to losses of time, either through cycle losses or due to losses of time when the operation is not running or not adding value. The list that is generated will include both cycle and run-time losses. Since the ultimate objective is to find causes that are linked through the causal chain to the original problem, we’re looking only for those activities that take time away from the value-adding task. In other words, if the operator is performing a non-value-adding task but the machine is adding value while the operator does the task, improving this item will not lead to reducing the problem, and thus is not a beneficial improvement. The first priority is to address the issues that directly reduce the time available to add value and therefore cause a loss of production.

Value-Added Task Blade is cutting wood

Non-Value-Added Task Load saw

Unload saw Change blade Clean up Break down Inspect parts

Move finished parts Meetings

Waiting for wood Handling wood

ALL ACTIVITY OTHER THAN CUTTING WOOD IS NON-VALUEADDED ACTIVITY

Figure 15-3. Value-added/Non-value-added analysis

For further information on the case see: Bill Costantino, “Cedar Works: Making the Transition to Lean,” in J.K. Liker (ed.), Becoming Lean, Productivity Press, 1997.

Continuing with the causal analysis (Five-Why) process in this example revealed the chain in Figure 15-4. Follow the bold text chain to the root cause in the outlined box.

Toyota uses this process of continually narrowing, isolating (using the 80/20 rule), and focusing efforts on the items that will provide the greatest benefit. Continuing to dig until the root causes are discovered also provides causes that are both easier to improve and, when improved, will solve the original problem. We can think of it as a funnel as shown in Figure 15-5.

Problem statement: The fabrication units per hour is below goal.

Why?

We are not able to make enough parts each hour

Why?

We are losing production opportunities

Why? Why?

Losing time Losing parts (scrap)

Why? Why?

(Not most significant)

Cycle time losses

Why?

Process not running

Loading the machine takes too long

Why?

Figure 15-4. Final pass Five-Why analysis

Operator walks 5 feet for material (Root Cause)

L A R G E P R O B L E M

Many Possible Causes Most Likely Causes Point of Cause

Five-Why Process

Root Causes

Figure 15-5. The narrowing and focusing process

Seeking Problem Causes That Are Solvable

During any process of analysis there will be a tendency to jump to predetermined causes. Predetermined conclusions are often based on issues that are not

within the ability or responsibility of the person developing them. A critical thought process of the Toyota Way is the assumption of finding causes that are in the direct control of the problem solver. In any problem analysis it is always possible to find causes that originate outside the control of the problem solver. For instance, it’s common to find fault with a supplier of material, or with a support group such as maintenance, or engineering (this is jokingly referred to as the “Five Who’s” and the objective is to find “root blame” rather than root causes). Also, there is a tendency to accept certain causes as “the way it is,” and therefore preclude the possibility of change. The following example demonstrates this phenomenon.

During the analysis of the sawing operation shown in Figure 15-3, it was determined that clean-up time was resulting in loss of production. The saws operated for three shifts, and each shift was assigned 30 minutes to clean up, resulting in a loss of 90 minutes per day. Following the Five-Why chain in Figure 15-2 above it is apparent that the operation is experiencing problems meeting the daily production requirement. There are lost time opportunities, and therefore the goal would be to capture the lost opportunities. The causal chain would appear as shown in Figure 15-6.

The cleaning activity is the “point of cause.” Finding the point of cause will provide both the time and place that the problem occurs. At this stage the root causes have not been determined, and the “Why?” questioning continues.

The leader asks: “Why do we clean up?” to find the root causes. Likely responses will be:

“It helps safety.”

“It makes the work area look better.”

Problem statement: The fabrication units per hour is below goal.

Why?

We are not able to make enough parts each hour

Why?

We are losing production opportunities

Why? Why?

Losing time Losing parts (scrap)

Why?

Why? Cycle time losses

(Not most significant)

Process not running

Why?

Figure 15-6. Identifying point of cause

Clean-up

Why?

(Point of Cause) Leads to root causes

“We like to work in a clean environment.” “Because the boss said we need to.”

“It helps the quality of the product.” “It is part of our 5S program.”

Each of these answers is true and valid, but tend to indicate issues that cannot be challenged: “good reasons,” and therefore “the way it needs to be.” Who could argue, for instance, that a clean workplace will provide a safer environment? Who would challenge the boss’s request? But none of these answers support the effort to resolve the problem! They’re a dead end. They presume lack of ability and responsibility for improvement. The answer to the question “Why?” must be related to the goal of capturing lost time opportunities and must be solvable. Think about the issue in terms of the goal of reducing the time required for cleaning. The current loss of time is 90 minutes per day. What would the possible benefit be if the total time could be reduced to 45 minutes per day—a reduction of 50 percent? It is very simple to calculate the additional production possible with 45 extra minutes of production time available. It is possible to set a goal of reducing the total cleaning time by 50 percent and establishing a new production target. This is a key point of the analysis—the result must be quantifiable, and there must be a clear understanding of the impact of the cause on the problem, which is the goal.

Notice what happens when the answer to the question “Why do we clean up?” is changed to “Because it gets dirty.” Continuing to ask “Why does it get dirty?” at this level would begin to yield the root causes. The objective is to reduce the time spent cleaning, so the perspective must be that of preventing the dirt, or minimizing its impact, thereby reducing the cleaning time required and enhancing the time to produce the product. A visit to the work area to see first hand how and where the dirt is generated will provide a clear understanding. Is the dirt being contained effectively? Does it leak from equipment? Identify areas that accumulate dirt: Is it possible to keep the debris from accumulating? Certain areas, such as under machines and tables, may be enclosed to prevent the accumulation. Observe the method of cleaning: Is it effective? Could the method be improved to reduce cleaning time? So you can see how the answer “Because it gets dirty” provides a perspective that yields numerous possibilities for improvement.

When the true point of cause is understood by following the genchi genbutsu method, many opportunities will surface that are well within the control of the problem solver and produce great results. Careful consideration of the causes, and by answering “Why?” in a way that will produce answers within the control of the problem solver, will generate tremendous opportunities.

Distill Root Cause Analysis to Simplest Terms

Note that any problem has many possible causes, and therefore many root causes. Attempting to list them all using the Five-Why causal chain would be tedious and time consuming. Though it’s important to understand the thought process and the flow-through to the root cause level, for the sake of focusing efforts, it’s better to return to the beginning and place real values on each cause along the chain, effectively isolating the most significant issues and providing tangible data that will indicate the degree of improvement possible.

A key to the Toyota process is the ability to be extremely concise in presenting massive amounts of information. Employees can then cut through the information available and communicate simply and clearly in a way that is understood by everyone. The Toyota Way forces the distillation of the information to only the most relevant details. It is always an important part of the process to review information with superiors, subordinates, and peers so guidance and support may be given. Providing reams of information that have to be interpreted or read by many individuals is considered a disservice to the readers. Imagine the waste if 10 or 20 people each had to read all the information and sift through the data to reach the appropriate conclusions!

A Picture Is Worth a Thousand Words

To be concise with the information, the analysis of the problem should be depicted graphically. This is aligned with the “visual workplace” philosophy of Toyota. To explain the loss of production capability, see the analysis depicted in the bar chart in Figure 15-7. In this situation the planned capacity would be the output if the process were operated 100 percent of the time. This may not include the hidden opportunity in cycle or scrap losses, but it captures the basic idea that losses are occurring.

Planned Capacity

Actual Capacity

Figure 15-7. Bar chart depicting production losses

Figure 15-8. Rank order chart of production losses by category

As shown in the causal chain in Figure 15-2 above, the causes for production losses are time losses and scrap losses. This results in multiple causes, and thus a branching of the causal chain. Visually this is depicted with a rank order chart as shown in Figure 15-8.

This rank ordering analysis shows that cycle losses have the largest potential for reduction in lost opportunity. A visit to the production area for a firsthand look (gemba), and a brief work method analysis (utilizing the standardized work process described in Chapter 6), would reveal the basic steps of the job:

1. Pick up material (walk to machine).
2. Load machine.
3. Start cycle (walk to start button).
4. Perform inspection (walk back to inspection area).
5. Place finished part in bin (walk to bin).
6. Unload machine (walk to machine).
7. Return to start of work cycle (walk to material).

Maintaining the pattern of graphically depicting this information, the work element times are presented in a stack chart format (also called a yamazumi chart) in Figure 15-9.

In addition to the time required for each element, it’s useful to understand the problem visually by depicting the flow of work (Figure 15-10). This pictorial flow is captured on the Standardized Work Sheet (Chapter 6).

Figure 15-9. Stack chart of work element times

Machine

Start

3

2

Machine

Figure 15-10. Visual depiction of work flow

Putting It All Together: The A3 One-Page Report

The analysis phase is typically where most of the problem-solving time is spent. The primary purpose of the analysis is to understand the causal relationships and to find enough causes that, when corrected, will yield an improvement sufficient to solve the problem. It is important to convey the basic findings in a way that clearly solves the problem. One method of doing this is to present them on an “A3,” the name Toyota uses to describe the single-page presentation of problem-solving activities (A3 is the European designation for an 11-by-17-inch sheet of paper).

Depicting the entire process on a single sheet of paper requires concise information. Obviously, every aspect of the problem discovered during analysis could not be explained on one sheet of paper; the causal chain alone would typically fill more than one sheet of paper. Figure 15-11 shows a completed Analysis section on the A3. This would follow the problem statement shown above. (For a detailed view of the A3 report-writing process, see Chapter 18).

Dig Deeply into Possible Causes

As mentioned earlier, the problem-solving process within Toyota is a collaborative activity. Initially, the question “Why did you pick up this problem?” was used to build consensus on the need for solving the problem, as well as to ensure a clear, shared understanding. Upon completion of the analysis, the collaboration between the problem solver, the superior, and the team includes a review to make sure that all aspects of the problem were considered. During this review a common question is, “Did you consider this item?” or “What led you to that conclusion?” Questioning is especially prevalent if obvious links in the causal chain have been skipped. Often the problem solver returns to the analysis to consider additional possibilities. Toyota managers intuitively understand the importance of carefully and completely analyzing a problem before leaping into potentially fruitless “corrective” activities.

Completion of the analysis phase should provide a clear grasp of the myriad possible causes, a narrowing to root causes, and clear understanding of necessary solutions, including specific details relating to the proposed benefits of implementation.

Here’s a review of the key concepts covered in this section:

  • Analyze each issue with a fresh perspective and follow the genchi genbutsu principle of going to the actual work area and looking for yourself.
  • Approach analysis with the intention of finding causes that can be corrected by the problem solver.
  • Continually narrow to the most significant causes, and refocus the analysis accordingly.
  • Determination of root causes should provide a clear and obvious understanding of the necessary solutions.
  • Analysis should be fact and data based. The root causes should be quantifiable, and the effect of improvement should be predictable in advance of implementation.

Reflection Activities

Following the problem you identified in the “Reflect and Learn from the Process” section of the last chapter (remember, we told you to keep it in mind?) complete the following activities:

1.Make a list of possible causes for the problem, then narrow it to the three “most likely” causes.

2.Select one of these three causes and investigate further to determine if this is an actual cause that will lead to the “root cause(s).” Determination of root causes is the most important element in the problem-solving process. Make sure your analysis is thorough and complete prior to proceeding to the corrective actions.
a.Go to the area where the problem occurs to see it firsthand.
b.Observe the situation and use the Five-Why method to follow the likely cause through to the root cause. The answer to each “Why?” should be based on factual information that is seen firsthand. Do not do this based on “speculation.”
c.Use the “Therefore” method to track back to the problem statement to verify the accuracy of your logic.
d.Is it possible to prove a connection between the problem, the most likely cause, and the root cause(s)? (If you can make the problem occur, or stop at will, you’ve proven a root cause.)

3.Locate the “point of cause” (the actual location where the root cause occurs and the problem originates).
a.Many large problems have several root causes and thus several points of cause. Identify the three most significant ones.
b.Keep evaluating until you find the actual point(s) of cause (you can see the problem occur).

4.Avoid the temptation to identify solutions until you have identified and confirmed the root cause(s) of the problem.
a.Thoroughly test your conclusions and prove the root causes.
b.Confirm that controlling the root causes will actually resolve the problem.