Knowing how often your production line is stopping is one of the most important single things a production manager can do to start improving productivity. This comes before OEE, for two simple reasons: downtime is an actionable metric and easy to understand.
There is a lot of theory as to how to do that. It is covered by Total Productive Maintenance (TPM), Lean Manufacturing and Six Sigma. Although it is important to understand the theory and we encourage you to study it, what we see in reality is that what most companies need is an easy-to-use and effective system. Therefore, we divide this into 7 steps.
The reliability of information is the number one priority when it comes to data. People do not use the number they do not trust. And time is a difficult one to track without an automatic system. It’s easy to measure (you just need a stopwatch, right?), but it’s difficult to track.
Many companies ask their operators to log downtime on a spreadsheet. “Every time the machine stops, you ‘just’ have to write down when it stopped and for how long.” That is a very hard thing to do. And it’s a boring task! No one can do it with precision in the long run. Not forgetting to look at the watch every time the line stops and remembering to look again when it runs and then remembering to write it down is not an easy task. Some operators will do it (mostly wrongly and it’s not their fault), some will guess and some will not do it at all and in the end, all you have is a bunch of data that you cannot use for anything because the information is unreliable.
There are inexpensive systems nowadays that can help you to automate this downtime logging. Use one and your operators will thank you.
In packaging manufacturing, we are mostly talking about production lines, with fully automated transfers. So, in a line with 5 machines, if the machine number 2 stops for 5 minutes and the other machines keep running (it is possible when you have accumulators between the machines, which is the case in most lines), is this a downtime of the line? There are endless discussions about this, but we understand that the correct answer is ‘NO’.
In a production line, the machine that pays the bills is the last one, so if it is still running, the line is running. If it stops, we understand that the line has stopped. It is still important to keep track of single machine stoppages if you have a system to do that, but only downtimes in the last machine should be considered as downtime of the line and only for that machine should we ask for downtime reasons.
After knowing when the line has stopped and for how long, in order to be able to take improvement measures, you need to understand why. Downtime caused by lack of production orders is totally different from downtime caused by a crash, which is totally different from downtime caused by a changeover. All of them require different measures to improve productivity and should, therefore, be counted separately.
There are three big categories under which we can categorize downtime:
- Planned downtime
- Unplanned downtime
- Changeover (Tool change)
Under category 1, we put everything that is not caused either by the operation or crashes, such as:
-lack of order
-world cup soccer match (yes, some production lines stop to watch their national teams playing in the world cup)
Under category 2, we should put all possible failures, trying not to feature too much. Some examples:
-Power outage (if this is frequent enough in your region)
It is important to mention that, although we are only tracking the downtime in the last machine, it can be caused by another machine which is upstream of it. This information, about which machine caused the downtime of the line, is very important and should join the downtime reason, especially when it is unplanned downtime.
Under category 3, we should take all possible types of changeover and group them into categories, including grouping them based on the task that takes the longest time to be performed. In plastic tubes manufacturing, for instance, we would have:
-6-8 color change
-Resin change dark to transparent
A 6-8 color change can take up to 3 hours, while a length change can be done in 20 minutes. It is important to tell them apart.
How many downtime reasons would fit the best?
Having only 3 or 4 downtime reasons is bad because we cannot know what is happening. Having 30 downtime reasons or more is just as bad. Since this information will have to be input by the operator, it must be easy to understand and easy to remember. Having too many items can overwhelm the operator and can use whatever comes to mind. It is like going to the Chinese restaurant and find 100 options on the menu. It’s difficult to choose!
A good number is between 10 and 15 downtime reasons. With more than that, operators will get confused, while with less than that, you will not have the details you need to run an analysis and take improvement actions.
The categories and parameterized downtime reasons allow us to analyze the data more easily, to make Pareto’s that will allow us to take corrective and preventive actions to improve the process. However, since we do not want to have 100 downtime reasons in the list, the operators should have the possibility of adding comments. This will help the maintenance team to really track what happened and don’t be surprised if many fascinating improvements ideas appear in the comments.
Knowing how much time your line has been down and understanding, is the raw material you need, to improve your productivity.
Micro stops are those that normally last less than 5 minutes. They are especially difficult to track manually and pass almost unnoticed. However, it is not uncommon to have a production line losing a full production day per month only in micro stops. Therefore, it must be tracked and analyzed. The analysis of the root causes of micro stops is rather difficult, because it may have multiple different reasons. Since we are not asking the operators to justify every single downtime reason (if we did, he would spend more time reporting than operating the machine), the micro steps will be seen as a block. To better understand what is happening, in case the micro stops are too significant in your production line, a production analyst will have to spend some time at the line observing the behavior of the machines.
Before introducing this system into the production floor, you will have to make sure that all operators have the same understanding of when to use each downtime reason. There must be a unification of the concepts within the production team, otherwise, you will end up comparing apples with bananas.
After you are running this system for one month or two, you’ll realize that some downtime reasons are never used, while others are used all the time. Now it is time to fine-tune the downtime reasons. How to do that: you will break down the ones that are used all the time and delete those that are never used.
As most production improvements, this is a living, never-ending process and making the first step is the most difficult part. The rest will come naturally.