An Example of 5 Whys - Is this Root Cause Analysis? Let Me Know Your Thoughts…
Tailchi Ohno, the creator of the 5-Why technique, is quoted using the following example to demonstrate using 5-Why’s for root cause analysis:
1. “Why did the robot stop?”
The circuit has overloaded, causing a fuse to blow.
2. “Why is the circuit overloaded?”
There was insufficient lubrication on the bearings, so they locked up.
3. “Why was there insufficient lubrication on the bearings?”
The oil pump on the robot is not circulating sufficient oil.
4. “Why is the pump not circulating sufficient oil?”
The pump intake is clogged with metal shavings.
5. “Why is the intake clogged with metal shavings?”
Because there is no filter on the pump.
What do you think? Is “NO FILTER ON THE PUMP” a root cause?
Let me know your comments.
And if the inventor of 5-Why’s uses this as an example, should people call 5-Whys a root cause analysis technique?
August 21st, 2007 at 1:34 am
Well, I guess you have to know the why-why analysis by taichi emphasis on resching five levels which means that “No filter on the pump” may be the solution, but if it is taken to the sixth step then probably the answer or solution would be “Bad Design of the Pump”..which means back to the drawing board.
Sometimes solutions may be within your control on the 5th why, going on to further levels may lead you to a solution that may have implications on the cost, time and etc…we can keep on doing the why-why’s till there’s no end..so I guess taichi may have looked into solutions that can be handle resolved within constraints…
August 21st, 2007 at 8:59 am
Regarding the example: weak. Indeed. Maybe 5 more of these and you’ll be getting close to a root cause (maybe maintenance, maybe design, maybe something else, who knows…).
More generic: I’ve said this before (I believe even on this blog in a comment to Mark’s earlier root cause/5 Why article): people who think that 5 Why is actually a “technique” or “method” should wake up and smell reality. It’s sometimes a good help to get a bit further, but it should be applied with extreme care and NEVER be relied on as the sole tool.
September 11th, 2007 at 9:37 am
And why were their metal shavings?
Mark
September 11th, 2007 at 9:43 am
Is “NO FILTER ON THE PUMP” a root cause?
This depends if the oil condition (amount of metal shavings) is normal / reasonable to anticipate. If the oil is as expected, this would work as a root cause. Otherwise, you should continue the “whys”
September 11th, 2007 at 9:49 am
Don’t miss my point.
This is the guy who INVENTED the 5-Why technique and this is his example of how 5-Whys should be used.
I can’t imagine so many metal shavings that they clog a pump without a problem that causes the metal shavings.
But more importantly, this is really a troubleshooting example … not a root cause analysis example. Once we finish troubleshooting and find the cause of the shavings, we will be ready to start to identify the Causal Factors and then START the root cause analysis.
Actually if I knew more about the robot, this might be a good Equifactor® troubleshooting example.
September 11th, 2007 at 9:50 am
The 5-Why is a useful exercise to track a single line of cause and effect but it seems to lead to a causal factor rather than a root cause.
In the example we end up with “No filter”. Because of the linear model, we left the metal shavings halfway up. The metal shavings are indicative of a problem in the system (robot) that is not corrected by a filter.
If we only pursue the “filter” fix, we only correct a symptom.
There could be many hidden causal factors and root cause that have been bypassed in a single line of cause and effect.
Are there metal shaving because the robot is lifting beyond its capabilities?
The design without the filter may be adequate if the machine is not operated outside its design parameters.
The advantage of a technique that examines mutiple causal factors is the ability to capture unrelated symptoms of the actual root causes by asking more questions in parallel cause-effect chains.
September 11th, 2007 at 10:03 am
Yes! But one more thing to consider…
The 5-Whys don’t lead you beyond your current knowledge.
Current knowledge in this simple example might have been … use a filter!
Human error problems are even more challenging when all you have is current knowledge and 5-Whys!
September 11th, 2007 at 10:45 am
To quote from the Wikipedia: “The five iterations are not gospel; rather, it is postulated that five iterations of asking why is generally sufficient to get to a root cause. The real key is to encourage the troubleshooter to avoid assumptions and logic traps and instead to trace the chain of causality in direct increments from the effect through any layers of abstraction to the first or root cause.”
Additional thoughts: Before you are too critical of Toyoda you have to look at the context and the time frame. Toyoda didn’t invent asking Why. He may have been the first one to put write it down as a “tool” with a cut-off on the number of why’s that “generally” lead to a root-cause.
Asking “Why” -without a limit on the number of times - was taught to me as an anlysis tool almost 40 years ago. We were also taught to be critical of our own work, avoid assumptions, seek expert advice and to verify our findings, especially if the answer came too easily or too quickly.
September 11th, 2007 at 11:07 am
Thanks for the comment.
Yes - Sakichi Toyoda was the developer of 5-Whys and probably started it more than 40 years ago (at least by 1950).
I agree that for the age - 1930’s - 40s - 50s … 5-Whys wasn’t too bad.
The problem is that many people are still using 5-Whys as their main root cause analysis tool and think that results like “no filter on the pump” are root causes. Thus, use of 5-Whys doesn’t get people to root causes … rather it is a toubleshooting tool that gets them to a single Causal Factor (and maybe not the only Causal Factor).
These Causal Factors need to have a systematic analysis that gets people beyond their current knowledge to find real, fixable root causes.
That’s the reason that I posted this example. To demonstrate the limitation of the tool EVEN when used by a master (Taiichi Ohno). If he doesn’t get it right, what chance do others have?
September 11th, 2007 at 11:10 am
‘5 Whys’ is not an effective tool to use to determine systemic/process problems with an event/incident. Although for certain situations it does work - the example is probably not a good example. There is also a question about the definition of root cause. Root Cause for 5 Whys is the same as immediate cause for TapRooT.
I use 5 Whys as a back check to see if I’ve determined the correct investigative path for an incident. If it passes 5 Whys then it reinforces my position on the incident case.
September 11th, 2007 at 11:13 am
The fact that there is no filter on the pump is not the root cause. The root cause lies mainly in the fact that a pump was in place that was not built correctly and was not tested prior to use. How many other pumps are out there without filters? What is the quality control check process like before pumps are put out for use?
September 11th, 2007 at 11:30 am
Bob - I agree the example is not good.
Here’s another example (this time from Wikipedia):
My car will not start. (the problem)
1. Why? The battery is dead. (first why)
2. Why? The alternator is not functioning. (second why)
3. Why? The alternator has broken beyond repair. (third why)
4. Why? The alternator is well beyond its useful service life and has never been replaced. (fourth why)
5. Why? I have not been maintaining my car according to the recommended service schedule. (fifth why, root cause)
This example (reviewed by many on Wikipedia) still has problems…
I’ve NEVER seen a service schedule for replacing an alternator … But let’s assume there is one.
The question still begs the answer:
1. WHY would someone not follow a schedule?
2. Would following the schedule have prevented the problem?
3. Should an alternate last longer and never require replacement for the life of the vehicle?
4. What is causing the alternator to age and can that be prevented?
5. Why didn’t the operator (driver) notice that the alternator was not functioning before the battery was dead?
Again, we are back into troubleshooting … This example doesn’t cause us to understand WHY the failure occurred and yet this example is another “standard” example used to explain the 5-Why technique.
Again - If the standard examples don’t work … Shouldn’t people be looking for a better technique?
Does anyone have a good 5-Why example for anything except an extremely simple failure?
September 11th, 2007 at 3:25 pm
Posting for Lane LeClerc (The 5 why’s link did not function…)
There is more to the robot story.
I would have to ask question # 6 - Why was there no filter on the pump?
May lead to further questioning
Lane LeClerc
September 11th, 2007 at 3:39 pm
And one more comment …
Why are people still teaching 5-Whys (or asking a certain number of Whys (5-6-7-8-100) as a recommended root cause analysis practice?
Shouldn’t instructors (root cause experts) know better?
If asking Why 5 Times seldom produces a good root cause analysis, shouldn’t anyone who is knowledgeable teach people something more robust?
If someone is teaching 5-Whys as root cause analysis, should a reasonable person wonder about the teacher’s level of knowledge or sincerity about solving problems?
September 11th, 2007 at 6:35 pm
5 Why’s is being rolled out as a problem solving technique as part of Lean implementation on our site. The process is sound and leads quickly to a cause but can focus on one line of investgation.
I have highlighted to the consultants on our site that the advantage of Taprrot is that it does not discount lines of investigation early and can lead to a more thorough outcome that can truly prevent reoccurence.
September 12th, 2007 at 12:50 pm
OK I get the point, inventor, poor examples, however I tend to agree with Jim Conroy… It’s very easy for me to understand and even more importantly explain 5 Whys. I haven’t been to 5 Why’s training but my first exposure it was presented to me as a starting point for an organization who has no root cause process – it’s simple and requires no software. Perhaps asking just more than one why in some cases would be a good starting point. How many fuses you think were replaced before looking elsewhere? When do you think the pump problem was found? Before or after the robot bearing was replaced? When you are in reactive maintenance mode with pressure from operations and management to get it back running it is difficult to perform any root cause diagnostics, but to at least ask to one’s self why a few times may avoid you having to return. I see 5 Whys as an introduction to simple root cause for events that happen everyday. If I had invented 5 Whys or when my organization gets to a point where they determined it was/is flawed I would/will be ecstatic…they get it!!! Ok guys it is flawed, how do we fix it…let’s not limit it to 5 rather let’s get to the right answer and sometimes we may have to create teams, analyze data, review equipment history, put science to the investigation and yes maybe even use software to help document and illustrate the results.
One can argue the first example given is actually 7 Whys. Note: Why stopped? Blew fuse. Why Fuse blew overloaded ckt.. …Why? Locked up, Why locked up? Lubrication (it could have been several reasons). I think the conclusion of insufficient lubrication is a bit presumptuous but if accurate one would have to say the questions and answers given after that were wrong. Something happen before the pump clogged and this is a symptom not the cause. That is, unless normal operations calls for there to be filings produced in the bearing system…and design engineers forgot the filter… but I think not?)
In the second example, again wrong questions/ answers: 2. Why? The alternator is not functioning. (second why) The 3rd’s question’s answer should have been…3. Why? The alternator is well beyond its useful service life and had not been maintained according to my car’s recommended service schedule.
4. Why? Because I couldn’t afford replacing it and determine I would risk it? (fourth why) Stop or…5. Why? I have a crappy job and don’t go to work as much as I should. (fifth why, root cause)
Mark goes on but I would argue - Does this failure or even Root Cause matter? Can an applicable value added maintenance schedule, predictive or otherwise be developed? For me, it doesn’t matter and I have no motivation for getting a better job – run to failure. So why go on trying to figure out how to extend service, etc.? To an ambulance service, fire department maybe it does matter; depends if backups exist? I guess.
Lastly, I’ve been involved with only a few detailed root cause analysis investigations and look forward to further training and learning opportunities but I deeply believe that it would be very short sighted of me to say only performing a detailed RCA for every failure provides value as well as saying the 5 Why’s approach provides no value or is the answer in every situation. Each has its place it just depends and sometimes no root cause analysis is the best approach. Thanks for the discussion…
September 12th, 2007 at 3:22 pm
On Star Trek Spock said:
“An easly understood falsehood is better than an incomprehensible truth.”
However, this IS NOT the choice we are faced with.
The choice faced by people trying to solve problems is: HOW MUCH EFFORT IS NEEDED TO FIND A REAL, EFFECTIVE SOLUTION TO A PROBLEM? IS IT WORTH MY TIME?
My point is that 5-Whys are EASY and easy-to-understand but often don’t get us to an effective solutions. Thus no matter how easy the technique is, most people are wasting their time but they THINK they are doing effective analysis. What they are really doing is trying solutions and seeing what works.
I can’t remember who told me:
“Don’t mistake activity for progress.”
But I think that it could be applied to 5-Whys. People have activity (they are asking questions) but that doesn’t mean that the questions are helping them understand the problem and develop effective solutions. Instead, they are guiding the questions to get to answers THEY ALREADY KNOW. They then apply these answers which may work or which may appear to work (because the incident doesn’t repeat in the near-term).
The less likely the failure, the more likely people are to mistake activity for progress when the real problems have NOT been fixed (and are now dormant waiting for the right time to recur).
Another famous saying:
“Even a blind squirrel finds a nut now and then.”
Perhaps the people using 5-Whys are lucky that some of their answers work without actual root cause analysis and they make some progress without real understanding. That’s what I believe success with 5-Whys looks like.
Also, really smart people (experts) already know the answers to simple or familiar problems. They are just making 5-Whys get to the point that their brain has already arrived at. Thus, they make progress based on their knowledge. Unfortunately, this technique is NOT transferable to others without the knowledge.
Can anyone tell me how 5-Whys would get someone to an answer they didn’t already know? How would you ask the right WHY when you don’t already understand the cause-and-effect relationship?
That’s just one of the basic problems with 5-Whys.
September 14th, 2007 at 3:56 pm
Dear Mark:
First all the word “WHY ” must not be used in a Root Failure Analysis.
Use the words how, when and others not related directly to human behavior.
We must follow always the Root Failure Analysis tree and the Equifactor
October 2nd, 2007 at 9:54 pm
I know this is way behind the times, but here it is anyway…
The greatest power to the 5-whys is that it provides an easily comprehensible presentation to people who are not knowledgable regarding specific RCA methods. To try to ’sell’ an analysis with its charts and so on to a group of managers who’s last, or maybe only exposure to RCA methods was barrier analysis 25 years ago, is a waste of time and effort. You need to present something they can easily grasp and follow, and, regardless of what methods you use to do the analysis, they can usually follow 5-whys.
Trying to show them the snap chart and the tree is an exercise in futility. Unless, of course, the management staff has attended RCA training of sufficient depth to enable them to understand what you are talking about. I don’t find that scenario to be very likely.
So, use the RCA tools that get you a solid analysis that you can justify, then present it in a way your audience can understand. (I’ve got one management team that can’t understand anything but barriers, so after all of the analysis is done, it is translated to barriers so they can understand what happened and what needs to be done. They don’t care that the actual analysis is not barrier, that’s just how they want to see it.)
Does that make 5-whys a root cause method? Nope. But, it does make it a pretty good presentation tool.
October 3rd, 2007 at 10:57 pm
Greg
Thanks for your comment and I agree that 5-Whys and Why Trees aren’t very hard to present to management. I also agree that 5-Whys is not a good root cause tool. That said …
I’ve never had trouble presenting TapRooT results to management. SnapCharTs are easy for managers or factory floor folks to understand with zero training. In fact the usual response I’ve seen is … “We need to get our people to be able to investigate and present so clearly!” And we’ve have many on-site courses after managers saw a presentation and insisted that more people get trained.
Don’t underestimate the understanding of your audience. People on the factory floor and in the corporate boardroom are quick learners.
To me, your argument just proves that it would be a wise investment to get your management some advanced root cause analysis training. Otherwise, they might just decide to change the answers to the “Whys” on your 5-Why chart and get answers theat they like better.
And one more note … One of the things we teach in all our courses is presentation of a root cause analysis. We always teach NOT to use the tree as a presentation tool … that’s not what it’s made for.
Thanks again,
Mark
October 17th, 2007 at 11:59 am
While there has been much conversation on when and how to use the “5 why’s” method, one perspective seems to be missing. Are the questions meaningful and applicable? I have seen and been involved in projects where all the companies’ particular change and analysis tools were used but the true root cause was not found. By not understanding the true system, by not having the right subject matter experts, and by not following a structured “why process” with identical parameters agreed upon, the process failed. I can use a wrench for many uses correctly and incorrectly, but does that mean a wrench is a bad tool?
April 1st, 2008 at 6:00 am
It is realy good article and explained in in detail.
I would like to receive your articles periodiclly through my e-mail ID.
Thanks and best regards,
B.S.RAO