What is Poka-Yoke?
Akinori Hyodo, former Factory Manager, Toyota HiAce Factory | #AskSensei Event 12 Summary | 22 July 2020
Poka-yoke is a well-known Japanese term in the world of lean. It means “error-prevention” or “mistake-proofing” and the implementation of poka-yoke on the shop floor is regarded positively and encouraged in general. But, should this really be the case? Lean sensei Akinori Hyodo discusses the idea of poka-yoke and explains a potential pitfall around its implementation.
Error Prevention and Poka-Yoke
Today I want to talk about error prevention and poka-yoke. Although we separated the two words here, poka-yoke is essentially a form of error prevention. Why is poke-yoke necessary? The basic premise is that we do not want to pass on any defects to our customers. We don’t wish to cause any problems for our customers and users of our products by supplying defective products.
Ultimately speaking, it would be better if you don’t have to rely on poka-yoke. In an ideal state, you will be better off without using poka-yoke. However, in the case where we have quality issues and defective products being produced, we need to investigate them, identify the root cause of those issues and put a countermeasure in place.
Alongside this, we also have standard work in place that our operators are performing. However, as I often point out, anything we humans do is never perfect. Therefore, even when operators are following the standard work that we have put in place, due to some reasons such as an individual not feeling well on a certain day etc., operators may make a mistake whether it is on one in 1000 or even one in one million units. We humans tend to make errors which we often refer to as “poka” (mistake) in Japanese. And, when a mistake is made, we cannot just dismiss this by saying “oh well, it was just one mistake out of 10 million, therefore we should be forgiven.”
If you see this from your customer’s perspective who received that one in 10 million defect, for them it is a defect rate of 100 per cent. On the condition that we don’t want to create that circumstance where a customer receives a defect, we work to eliminate the chance of defective products from being produced and passed on to our customers by putting standard work in place for every operator to follow and by constantly improving the level of standard work. However, we may not be able to cover everything through standard work alone because of the human error factor as I mentioned earlier. As this is the case, there is an opportunity for us to use poka-yoke to prevent defects from being made and flowing through to our customers.
When it comes to poka-yoke, you shouldn’t regard it as the first and foremost solution to your problems. It isn’t something that you should jump to straight away. You should consider poka-yoke only after you take appropriate measures by going through a root cause analysis, and once you are satisfied that the defects are due to human error. That is when a physical solution is put in place to prevent errors from being produced again. It is important for us to consider poka-yoke as our final solution after making various other attempts to prevent the errors from happening.
If you don’t take this approach, it is easy for you to end up with a process filled with costly, large-scale poka-yoke devices. If you think through the issue and apply good thinking to the problem to start with, you can spend little or no money to eliminate the chance of defective products being produced.
Based on my experience, I can tell you what’s going to happen when you have too many poka-yoke devices in your processes. Because they are all built around and into the same machine in a process, and work in combination with each other, once the process has a problem, fixing the problem becomes a complicated exercise. When you need to reset the machine, for instance, you also need to check all those poka-yoke devices you have built into the process and reset all of them to start the production again.
That is why I want all of you to think of poka-yoke as something to use as a final measure only after you have considered and tried various other methods and measures to prevent errors from happening.
As for the types of poka-yoke, there are mainly two types that are often used in manufacturing. One is a poka-yoke that prevents the manufacturing of defective products in the first place. This is a work-error prevention poka-yoke and with this type of poka-yoke in place, the defect cannot be produced. The second one is a poka-yoke that prevents defects from flowing to the next step in the process or to customers. In this case, rather than stop the defect from being created, this type of poka-yoke catches the defect such that it can’t proceed to the downstream process.
But, to be honest with you, if these issues are considered and addressed in the original design of a product, a process or a machine to begin with, you wouldn’t need to rely on any poka-yoke devices to prevent these errors. However, when these issues are handed over to production and factory side and making necessary changes to the original design is difficult and too cumbersome, then the factory side needs to think through various ways as to how they can prevent errors from being made through standard work and other means, and then consider introducing poka-yoke as a last resort.
Finally, as I have said regarding various points throughout this webinar series, the important thing you should remember when approaching poka-yoke is that your objective here should not be implementing poka-yoke. I wish to stress the point that poka-yoke is something that you can use as a final step when you haven’t been able to achieve quality error issues from happening. So, please don’t be thinking of introducing poka-yoke from the beginning and as an objective in and of itself.
How do we apply poka-yoke in the service industry?
In the case of applying poka-yoke in the service industry, we first need to consider what kind of issues exist in the service industry. As I have discussed in my opening remarks, there are various things you can do and implement to overcome the issues before applying poka-yoke. As such, I’d recommend you go through all the thinking process by looking at all the various different ways to prevent the issues from occurring and only use poka-yoke for the very final part that you cannot prevent due to human error.
For example, mistakes may be made in a workplace where things are not organized and the level of 5S is poor. Errors may occur when there are no standard or rules in place and all workers are performing the same process in various different ways. Or, those in the supervisory role are not properly checking the workplace their subordinates are working. In all those instances, simply introducing a poka-yoke device or measure won’t fix the problem.
When you are in the situation where you don’t have the fundamentals established properly but jump straight to solutions such as poka-yoke, this won’t give you a good result. So, regardless of whether you are in the service industry or other sectors, please think through all the various different things you can do to eliminate errors before applying poka-yoke. The thinking and approach around poka-yoke is the same in any industry.
Most automotive industries rely on using camera systems and laser checks for poka-yoke. How could we achieve the same level of error proofing and poka-yoke in the current climate without spending a lot of money?
Those camera systems and laser checks are used as an easy solution. But, oh boy, they cost a lot of money.
The most important thing is to analyze what the problem is. You must get into the detail of the analysis of the problem. Identify the key issues surrounding the problem. In doing so, you can find that there are many different things you can think through, apply and implement that do not cost any or much money. As I was saying earlier, if the final piece of the puzzle is something that you couldn’t achieve through low cost or no cost solutions, then the cameras and lasers may be your answer. If you can go through this detailed process, you may find that rather than having cameras and lasers everywhere, you only need to cover that last few per cent of issues where cameras and lasers are imperative. In this way, you will be able to minimize the number of expensive devices that you have in your workplace quite significantly.
Another important thing I want you to think about in this poka-yoke conversation is the role of management that play a part in applying poka-yoke. When looking at a workplace where expensive poka-yoke devices are put in place, you’ll often find that the manager approves those expensive solutions too easily. When the paperwork to approve those cameras, lasers and robots etc. is submitted, a lot of managers would say “oh, that looks good” and sign off. But, the challenge that more managers around the world need to take is to say to their people, “no, you need to have a look at your proposal again and work out a way to achieve the same outcome for half the cost”. Managers need to push back against making simple decisions around spending and costly solutions, and direct their subordinates to implement a cheaper solution. But, in reality, too many managers are simply signing off without questioning or pushing back against expensive proposals. In the end, everything we do and implement at work are all linked to the profitability of our company which is the key thing we are trying to achieve through all this work. So, what I have outlined here is also an important role of management.
In the case of a vehicle, the selling price of that vehicle is set. Therefore, there is a budget for the investment we can make into equipment and machinery and we cannot exceed the budget. So, when the factory side decides to implement all those poka-yoke devices around the existing processes, they are adding extra cost to the overall manufacturing cost. But, in doing so, the question I have is whether the management side is genuinely looking at the cost benefit analysis of implementing poka-yoke into their processes. Is the company truly getting the payback from implementing these poka-yoke devices? If too many of these devices are approved to be implemented without really checking on the effect of the implementation, the company can be on the way to bankruptcy.
Something I often ask people about poka-yoke and various other investments is that if they would consider spending their personal money on the very investment they are recommending the company to consider. And, the answer I get most of the time is “no”. I think this provides a clue as to whether the investment is genuinely necessary at the time or not.
So, at the risk of repeating myself, I wish to reiterate that poka-yoke is the last solution you should look into only after you have thought through and applied various other low cost and no cost solutions. It is also important that management creates the environment where employees are thinking more thoroughly through the problems. And, management can do that by pushing back against some of those investment ideas that are coming up to them.
Are electronic records of the information that goes on the huddle boards (visual daily management boards) maintained at Toyota?
The quick answer to the question is, yes, we are maintaining that information via electronic records. In the old days, it was all paper-based and we had to hold onto paper records. But regardless of in what form we maintain the records, there have always been rules around the detail of the information we maintain as records. For example, there is a certain amount of information we need to hold onto in order to calculate and accumulate the information and data required monthly, quarterly or annually, etc. That is why we maintain a certain level of information electronically so that it can be fed into those required numbers.
What can be the problems with reducing changeover time?
Just like I always say about everything, you must ask yourself, “what is the objective behind reducing the changeover time?”. If things that must be in place are not properly in place, this will create problems.
I am not quite sure as to what type of situation this question is specifically referring to, but I will talk through one idea here. The first thing I’d like to recommend is revisiting the objective behind reducing the changeover time. Ordinarily, the end objective of changeover time reduction is to reach one-piece-flow production. But, when you are producing many different types of products, although it is ideal that you try to produce those various different types of products in a one-piece-flow manner each time, economically it doesn’t always necessarily make sense. That is why you need to batch items up into a little bit larger volume than just one. Of course on the other hand, the larger the batch size becomes, the larger the inventory becomes.
So, I encourage you to reflect on the indicators and KPIs that you are using around changeover time. Obviously, changeover time is one KPI but you also need to be looking at the inventory levels as well. If you are decreasing the changeover times but your batch sizes and inventory volumes remain the same, this is not making any improvement at all. Therefore, you need to be looking at both sides of things to make sure you are actually making an improvement and linking it with your overall objective in the first place.
In reality, there is no such thing as achieving a perfect improvement the first time around. So, you need to make a change, check if you are achieving the target, and make other changes as you go along in order to get closer and closer to the target of your objective.
The problem in manufacturing or in any business is not that you have problems. There are always going to be problems and every workplace has problems. However, the key thing is that you recognize those problems as problems and attack them one by one to make sure you overcome them.
#AskSensei is a regularly-scheduled webinar held together with Shinka Management Senior Lean Consultant and former Toyota HiAce Factory Manager and Director Akinori Hyodo. Each event we cover a different topic related to lean, with participants invited to put their questions to Hyodo Sensei.
To be notified of up-coming #AskSensei webinars and other events, sign up to our eNews mailing list at the bottom of this page.
Up-coming #AskSensei Events and Registration
For further information about the #AskSensei webinar series, please refer to our #AskSensei overview page.