Learn to Use the 5 Whys Root Cause Analysis Problem Solving Tool
Vložit
- čas přidán 30. 07. 2024
- The 5 Whys (academy.velaction.com/course/...) is one of the simplest problem solving tools used in Lean manufacturing and Lean offices. This presentation shows how to use the 5 Whys, and what to watch out for. Created and presented by Jeff Hajek of Velaction Continuous Improvement.
- Jak na to + styl
Jeff, very informative and useful. Thanks for the information. Kevin
thanks for this and it's very useful. it can be the reference in handling problem.
Great video very clear!
Good short video on a useful problem solving tool.
Thank you Jeff. Nice and clear.
verry good vdo easy to understand
Yam,
Common and special causes are really more about statistics than the actual event. Common causes actually have causes that, with a fine enough magnifying glass, look a lot like special causes and can be identified and removed. It is just that they fall within statistical control of a process--i.e. they don't spike. As for special causes, some are outside of the control of an operation, but far more are in the organization's control, or are at least manageable. Power outages can be managed with generators or batteries. Natural disasters and terrorist attacks can be mitigated with the choice of location, etc. I do agree, though, that many, if not most, problems do have multiple contributors. Quite often, though, eliminating the predominant cause will greatly improve a process. The 5 whys is a good starting point for most problems, and if the initial attempt fails, the problem solver can try again.
@AssistingHealth With just one why, people seldom get to the root cause. They end up treating that symptom and the underlying source of the problem is never addressed, so the problem recurs. And even if one symptom is prevented, new symptoms have have a way of popping up.
Thanks for the course
But when can i use Root cause, Taproot or causal tree analysis?
I'm not familiar with Taproot, so can't comment on using that software. Essentially, causal tree analysis is used when problems are complex and have multiple causes. The 5 Whys tool is what you would get if you followed just one branch structure through the causal tree. It is generally used as a simple tool to get people moving and to think a bit more critically during problem solving. Root cause analysis won't necessarily help you determine if the path you are on is the most likely, but it will help you make sure you don't treat a symptom of that path.
What is right ,to write 5 why first and then solve problem or solve problem and then write 5 why? practically writing 5 why will add huge amount of downtimes i cannot have such a stop for 5 why its is about high speed cigarette packing machines
+Ananth Kuikarni The 5 whys is a problem solving tool that is used as part of the problem solving process, so they are not a first-second sort of thing. One is a part of the other. With a high speed machine, you generally would just work to get the machine back up and running (a stopgap), but you'd then solve the underlying problem. That is where you would use the 5 whys--as an offline thing while the machine is running, not when the machine is stopped. The issue is that people often solve the symptom rather than the underlying problem. I would track the issues on your machine, then use a Pareto chart to go after the biggest issues first. Use the 5 whys to help solve those problems one by one to reduce your machine issues.
Very nice vid. Thanks for sharing this with us; Jesus Christ Bless! ;)
problem tree analysis
czcams.com/video/wCG2xKM7JbI/video.html
i would like to know the best 5 whys method you would use for call center
Hi Tram.
Sorry-I missed this comment when you posted it.
I'm not exactly sure I follow the comment, but I will give it a try to answer.
5 Whys is a specific type of root cause analysis. For call centers, it would be good for general issues. But because call centers tend to be so data intensive, you might be able to use a lot more sophisticated data analysis tools to go over the information.
The 5 Whys is more of a qualitative, gut feel tool than you'd probably need in a larger call center.
Hope this helps.
Jeff
how do you know that you reached the root cause? I.e in the video example, you could have continued to ask why. How do you know you need to stop here?
Good question. The number 5 is just a ballpark. In most cases, it will get you to the root cause. You have to apply some common sense and ask if removing the deepest cause will eliminate the symptoms you are facing. If the answer is no, keep going. If it is yes, make the change, and then check (as in PD"C"A). If the problem is still there, you need to dive deeper and try again.
Excellent question. Actually there is no way you CAN'T. You can't reach root cause repeating same question 4 times. What if you can not answer the first WHY?
You need a facts finding analysis approach to reach root cause
5Why is NOT a Root cause Analysis approach
@@eduardomuniz5314 The 5 Whys is one tool in a root cause analysis toolkit. It has the big drawback of not being scientifically repeatable, so for important problems, you want more hard data. It has the advantage, though, of being a tool that frontline team members can, and will, use. The whole point of this tool is to peel back the layers of the problem to make sure you are not treating symptoms. It is effective if you want to look at why dirt keeps accumulating around a machine. It is probably not the best approach to determine why sales are falling in the Midwest region.
You said some other practitioners won't even use it.. what other tools would they use?
Ahh..it's been a while since I made the video, so I hope I remember what I said. Generally, the issue with the 5 Whys is repeatability. If you don't get the same answer regardless of who uses it, you should question the results. For the 5 Whys, it will generally still yield good results on simple problems, and if you are wrong, you just PDCA the heck out of it. But when you want to be really sure, you need to get increasingly scientific using objective studies and formal data collection. Flow charts have some of the same characteristics, but less so. Pareto charts are more scientific because you are taking actual counts/measurements, but there is still some inherent repeatability issues in selecting categories. Run charts get more precise. FMEA can help isolate problems with more structure. A decision tree can guide act sort of like a 5 Whys tree, but with data added at each of the decision points. Control charts are even more scientific. ANOVA and other statistical analysis is even better. As you can imagine, the more scientific and precise you want to get, the more skill and resources you need to invest.
what if there are multiple causes
Hi Rose. That's one of the weaknesses of the 5 Whys. If there are multiple causes, you'd do the 5 Whys for each branch of the cause 'tree'. In the real world, each cause may also branch at the 2nd or 3rd why (i.e. multiple reasons for each cause.) And adding to the shortcoming, the causes may interact with each other. In those cases, you have to get more sophisticated than the 5 Whys. But for simple problems, the 5 Whys really helps cut down on recurring problems related to not addressing the root cause.
@Aimrehtopyh You blame the Brits!
@AssistingHealth I like your story. But I do recommend looking at the process before trying to hold people accountable for errors. In my experience, most problems are not caused by some fundamental lack of ability or motivation, but rather by a problem with a process.
Problems are unpredictable and can be caused by people too
5 Whys often called Root Cause, assumes every problem has a root cause. It’s an overly abused methodology which causes problems to go unresolved because of its misuse. Thinking of any manufacturing operation, Deming correctly instructs us that each process experiences variation based upon Common and Special Cause.
Common causes are those which are inherent to the process (such as those commonly associated with the Ishikawa diagram). Special causes are those which are not inherent to the process and over which the organization has little to no control (i.e. Power outages, Natural disasters, Terrorist attacks etc.)
Failure in using the 5 why method, occurs when folks performing analysis don’t consider Deming’s common causes and assume everything is assignable to a single or root cause. At a minimum Material-Manpower-Machines-Methods- and Management will all be acting upon a manufacturing process at any one point in time (common causes found in the Ishikawa or fishbone diagram).
Knowing there are usually several possible causes acting upon a single process at any one point in time indicates that the idea of a single root cause is not plausible. 5 why becomes valuable when it is used in concert with the Ishikawa diagram (fishbone) to allow the list of probable causes to be reduced by eliminating via 5 why those causes which are not probable.
"Let's say you show up for work one day and find a giant oil slick next to your offshore rig in the gulf of Mexico."
problem tree analysis
czcams.com/video/wCG2xKM7JbI/video.html
You mentioned that different groups will come up with different root causes. You should look for a system-based cause rather than a person-based one. And you can test your root cause by running it backwards with another team to see if you end up with the same problem.
There are two issues. The first is that this isn't a scientific tool. You can get different results simply because different people know different things about a process. For example, if you are dealing with bad gas mileage, a mechanic may focus in on getting the car tuned up, and the guy who owns a tire store might immediately gravitate towards proper inflation of your tires. Both are probably contributing factors, and both can probably be confirmed as you mention. Which leads to the second issue. This tool implies that there is one root cause, when in fact, most problems have several contributing factors. It is more like a tree than a straight line. The key with the 5 whys, though, is about now stopping at surface layer symptoms, but rather diving deep into the underlying issue.
@@VelactionVideos The point is that if you have different teams to "code" and "decode" and you get different answers, you need to check each of the "whys". Talk to the people who do the job. Check the material spec against the required spec. Look at the takt time: if it's shorter than previously defined, what has changed? You should CHECK for a different team with 5 whys. If all the answers are the same, then you may have a good result. MAY is the operating word. If you get a different cause, you need to use another tool. 5Y is a good tool to point you to the real cause, BUT YOU NEED TO VALIDATE IT. Talk to operators, inspectors, packing checkers, and ask them if and how they could or should have stopped them.
In short, 5Y is not, and should not, be the end of checking the root cause: you need DATA to validate it. Check everything. Could the night shift worker have been tired? Maybe: ask him or her. Look at the hours worked and the effort required: would that make YOU tired? Does the worker have a medical other problem? Do you talk to the worker at all? If the team decides in the first meeting to go to "operator error", do you take that as read, or do you talk to the operator?
I loathe "operator error" as a "root cause". Does the operator come to work one day INTENDING to make something wrong? Probably not. But do you ever talk to the operator?
The MAIN problem with fishbone/5WY's is that people try to solve it IN THE ROOM ON THE DAY. You don't just validate the root cause but "Y's" in the room. Get out into the factory or office and talk to the people who do the actual job. Look at what their documented procedures say. Look for facts, records, data to support the root cause identified. If there is no match, you need to look for another cause: but armed with new data this time. Take your time: you should have containment actions in place by this time.
Always check your guesses and assumptions with FACTS.
I'm not sure if you've seen the 5 Whys form I have, but it has exactly what you mention on it. It has a confirmation step at each 'why' to try to add some rigor to the process. But the truth is, the vast majority of little issues on the shop floor don't really need a lot of scientific rigor to solve the problem. They just need a little investigation. Maybe a painted part keeps coming in scratched because of a problem with the transport cart. If they look at the cart and find debris on it and just remove the debris, they still have the problem of how the debris got there in the first place. But there probably isn't really any need to do any data collection and number crunching to go figure out where the debris came from. It just takes a little legwork. There is a lot of utility in the 5 Whys for simple problems, but for more complicated ones, I make sure problem solvers are going deeper into the issue than the 5 whys allows.
@@VelactionVideos But you could at least put some kind of spacer in the cart to allow debris to fall to the bottom and not scratch the parts. I agree that for a really complex issue, it's a blunt tool. I would tend to go for an affinity analysis for those..