Use Poka-Yoke to mistake proof your processes

Last Updated on 9 September 2023

This methodology was developed by Toyota, and was originally called baka-yoke (fool proofing), but was changed to the less demeaning poka-yoke (mistake proofing). Personally I like the term idiot proofing and use it e.g. when talking about anything stopping me doing something wrong before my morning coffee, it depends on who you’re discussing it with as colleagues can take it personally. You may want to just refer to it as mistake proofing, which is easier to spell than the Japanese term.

The idea is that ALL defects are stopped, whether by making it impossible to make the error, or that an alert activates if a defect occurs.

There are two types:

• Warning – Lets you know when an error occurs with an alert
• Control – Stops the error from occurring or stops the process when errors occur

Why we need Poka-Yoke

People will always make mistakes. I used to work in audit, and the motto “if there’s a way someone can do something wrong, at some point someone will do it wrong that way” served me well.

You will naturally do a process a little different each time, and sometimes that will be outside the acceptable parameters, and that’s not mentioning all the times you don’t get enough sleep, are distracted or (surely not) go to work a little worse for wear after the night before. That is not even mentioning times where the machine or parts go wrong.

Errors will occur, and at best will lead to waste, at worst will be found by the customer or cause a major issue. These errors can be prevented though. With enough training and the correct production methods, operating as if all errors can be prevented or easily detected will significantly reduce the number of errors that get through.

How do you perform Poka-Yoke?

As with all things, it’s easier if you implement mistake proofing using a predictable method:

1. Map the process

Create a detailed process map of the area you are trying to perform mistake-proofing for. Depending on the process this could be a flow chart, value stream map or similar.

2. Identify where the errors can occur

If the process is currently being used, you can use a simple check sheet to track where the errors are occurring. This could catch the commonly found issues.

If it is less obvious or you are also looking for rare high impact issues, you can brainstorm with the team to find where the errors are likely to be made.

3. Prioritize the errors

In an ideal world, we would make all errors impossible, however for reasons of time and budget you will likely need to prioritize. FMEA provides a good framework to do this efficiently. This grades to errors according to how likely and how severe the consequences of the error are. This is important as otherwise it is easy to just focus on the low impact frequent issues.

4. Think of solutions

Think of ways to mistake proof your errors, starting with the highest priority. Again, you may want to brainstorm or use NGT on this one, as often more heads are better.

You are usually trying to find solutions in the following order (if you can’t manage number 1, try 2 etc):

1. Design the process so you can’t make the mistake, e.g. you can’t turn on a microwave oven with the door open.
2. Make it so that the next activity can’t start if the previous one wasn’t performed correctly. E.g. the next part won’t fit on unless the previous one is there to attach it to.
3. Create a process where it is both very difficult to make the error, and a signal will flag to the operator that an error has been made

If all else fails, just make it as difficult as possible to make the error, e.g. making a form multiple choice.

5. Cost your solutions

You need to make sure that your solutions are possible from a budget / time perspective. You can use cost-benefit analysis to help you here if the solution may cost a lot in resources (time or money).

Examples

An obvious example is right next to me as I type this – my filing cabinet would tip over if too many full drawers are opened, but the mechanism inside stops me opening more than one drawer at once, saving me from getting crushed. At lunch today, I couldn’t start the microwave unless I closed the door, so that I didn’t accidentally cook myself.

Types of Poka-Yoke

The methods you can use are limited only by you and your workforce’s creativity. There are some examples to get you started:

• Alarms that detect errors and notify you when something isn’t correct
• Checklists that you go through each time to make sure no steps have been missed
• Guides in the machinery so that the material can only be put in the correct place
• Limit Switches that won’t let you turn on the machine until everything is correctly aligned and placed (such as the microwave example above)

How to implement

It’s easy to start using Poka-Yoke to improve your processes, and work towards a zero defect workplace:

• Start from assuming you can make the mistake impossible – you have to work from the assumption that all mistakes and defects can be removed
• Stop doing it wrong – just because you’ve accepted things up to now, doesn’t mean that you have to accept it going forwards
• Good ideas come from everywhere – from my experience the best ideas have come from the most unlikely sources; get the team together and have a brainstorming session
• Make removing errors part of the process, not an afterthought
• If you think you’ve got a good idea that may work, try it out – it may work better than you expect