Never Skip a Step in a Chaku Chaku Line!

Chaku Chaku lines and many other manufacturing cells and manufacturing lines have one (or more) operators working on different processes throughout the cycle. However, with this multi-machine handling, a worker must not skip a process if there are technical problems. Doing so—even with the best intentions—will increase fluctuations and cause waste. Let’s have a look at an example of why skipping steps is not so good.

Introduction

Chaku Chaku lines are popular semi-automated production lines for small parts with fast cycle times. One or more workers circle the line, insert the part into the machine (since positioning a part is a bit tricky for fully automated machines), and then move to the next machine. The machine automatically processes the part and then ejects the part once it is done. (Kicking a part out of the machine is much easier than adding one. This also frees up the machine for the next part). An animation of a Chaku Chaku line is shown below.  You can also read more on this in my post on The Chaku Chaku Line.

The Issue: Problems at One Station

Any machine has hiccups. Hence, sometimes there may be problems at one process in the Chaku Chaku line. The worker comes to this station, but the station is not yet ready. Maybe there was a problem with starting the process during the last round. Maybe there was a problem with the process itself. In any case, the process is not completed and no part is there to take to the next station. If it is a smaller issue, the worker may be able to fix it himself and get the machine going again.

But now the worker has a dilemma. Since the machine just started the process, there is no part ready for the worker to take along to the next station. Assume the problem is in process 2 in the image above. The worker now has two options:

• A) Wait for (almost) a cycle until the part in process 2 is completed and can be taken to the next process; or
• B) Continue to the next process 3, which can’t be started due to the lack of a part from the previous (troubled) process. However, he can take the completed part from the next process 3 and bring to the next process afterwards, 4, and continue the loop. This way the worker would avoid idle times.

Since there is usually pressure by management for the worker to work, chances are the worker takes option B and continues working with the next possible machine. Often, the worker also has an intrinsic motivation to work and contribute to the success of the company. However, despite the workers’ best intentions, taking option B and continuing to work with the next available machine is not good. Let’s see how this plays out.

Worker Continues with the Next Available Machine

The worker continues with the next available machine. In our example, he would take the completed part after process 3 and add it to process 4. Similarly, he would put the part from process 4 into process 5, and a new part into process 1 again.

However, in front of process 2 (our problem machine), he would find a non-standard situation. Instead of one part completed after process 1, he would now find two parts completed. Since the process 2 “skipped a turn” due to the problem, one part that normally should have been processed during the previous turn is still waiting, as illustrated below.

Okay, well, he simply takes one part (hopefully the older one to maintain FIFO) and adds it to the machine. While this feels like no problem, there is now additional waste in the system. The first one is inventory. You now have two parts where one part would have been enough. Second, depending on your material handling, he now has to handle two parts (e.g., remove one part from the machine, put it down, grab the other part, and insert it into the machine). This “put it down, grab the other part” is extra motion and extra transport, another type of waste (see my article on Muda for more). Overall, by “saving time” and continuing to work with the next available part earlier (option B), you now have introduced waste into the system. And, as long as there are new parts in the inbox, this will never go away.

Anyway, the worker continues, adds the part from process 2 into process 3… and then finds no part from process 3 that he could add into process 4. Since in the previous cycle, process 3 did not get a part from the problem machine, it had to skip a step and now has no parts ready for process 4. The worker has no choice but to skip process 4 and continue with process 5 by adding the part that was just completed by process 4.

Adding the next part to process 1 is just fine, but then there are still two parts between process 1 and 2, adding waste to the process as described before. He loads a part into process 2 with this extra effort. Loading a part from process 2 into process 3 is according to the standard, as is loading a part from process 3 into process 4. But then you have the next problem. Since process 4 did not get a part in the previous step (i.e., skipped a round), there is now no part ready for process 5. Hence, this time process 5 has to skip a round, and the worker again has no choice to continue to the first process again.

Skipping process 2 due to a problem forced the worker to skip process 3 in the next round due to a lack of parts, which then forced him to skip process 4 in the round thereafter, and finally also process 5. Hence, for every cycle afterwards he had to skip one process, until he had to skip the last process on the line. Since in this example we assumed an unlimited supply in the inbox, he never had to skip process 1. After skipping every process after the problem machine (the first “skip”), the system is now stable again, and every process will work every cycle (assuming no further problems with the processes). However, for as long as the system is running, there will always be two parts before the original problem process, causing waste through inventory, movement, and transport.

So, how would you get rid of these extra two parts? You could just take one out, but that would be wasteful, as you now have a perpetually incomplete part. The smarter solution would be to—maybe you guessed it—skip process 1 once too.

After that, you will have only one part between process 1 and 2, as it should be. Now, after skipping each process once (with the forced first skip of process 2 due to the problem), you are back to the original system again…

The Better Option…

…or—hear me out—instead of skipping each step in sequence and having additional inventory, transport, and movement, how about… just not skip the problem process in the first place? If there is a problem with one process in a Chaku Chaku line, maybe do NOT continue to work until the problem is resolved AND the part from the problem process is ready! Yes, option A would have been the better choice, even though this is not very intuitive. This similarly applies to many production lines with very small buffers, even if there is more than one operator in the line or cell (but not an operator at every station). If you have larger buffers, these buffers could decouple such fluctuations… but then you again have more inventory, transport, and movement. Now, go out, make sure the best intentions of your workers do not increase your fluctuations, and organize your industry!