Changeover Sequencing under Duress: Problems with Make and Deliver

In my last post I looked at strategies to manage changeover sequencing if your supplier gives you trouble. Basically, you can sometimes reduce the damage by fine-tuning the prioritization (i.e., by using the limited raw materials to make the most important parts). This second post in this series looks at similar situations if your customer acts up, or if your own system makes problems. In other words, after discussing “source” in the last article, we now look at the “make” and “deliver” part. Admittedly, some of the approaches are similar to the problems with “source.”

Problems with Make

Broken Machine in JidokaSometimes you may have problems with “make” (i.e., your own production). This usually means something went belly-up, a machine broke, workers are on strike, your key specialist quit and now works for the competition, or other reasons. Here, too, you should of course work on fixing the system to get up to speed again. A huge topic, but again not the focus of this blog post.

The question is which of the remaining products to ship to which customer. This problem is somewhat similar to the problem with source above. Prioritize the remaining completed stock to get the maximum benefit (or the minimum loss) from your customers. Here, too, you may take apart completed products to make other products, again with the high risk of the cure being worse than the disease.

Problems with Deliver

Finally, there are problems with deliver, where your customer orders too much or too quickly, or alternatively orders too little. While this looks like a customer problem, it is in fact usually a problem of your capacity – you are unable to match your supply to the customer demand. Of course, one major action here is to adjust your capacity. Add or remove extra shifts or generally working hours, get a new machine, make a machine faster, and so on. But again, the focus here is on sequencing, which is another option to mitigate the problem until you have brought capacity online (or offline).

Crowded Supermarket in NagoyaFirst, let’s look at the more beneficial situation where the customer wants more products. Here, we have to distinguish between a peak demand for a single product or an increased demand across the entire board. Assume you have 20 products, with roughly 500 each per month. Assume you make one batch of 500, with a different product every day (and yes, this is still a pretty big batch, but it makes an easy example). For some reasons you get an order for one of these products of 3000, while all other products drop slightly so that the total demand is unchanged at 10 000 products next month. Manufacturing happily makes one big batch of 3000 products, reducing their changeover time. However, this could be really bad. If you normally make 500 of each product every twenty days, then those 500 are the inventory to cover the remaining nineteen days until you make the product again. If you now make six days of one product, you may run out of inventory for the other products. It may be better to split this large order of 3000 products in smaller batches for production of maybe six batches of 500 each. What this could look like at your plant depends heavily on your situation, but please do try to level demand peaks into smaller batches so as not to run out of stock for other products or raw materials for the peak product.

If your demand goes up across the board, then besides adding capacity you also have to prioritize your limited capacity. This is again similar to the limited supply example above. What products and customers are most important, where can you generate the largest benefit and the smallest loss. Prioritize your remaining capacity accordingly. Produce the most urgent parts.

For changeover sequencing, you can still optimize the sequence of the prioritized parts to reduce changeover time and effort. For problems with the supplier, the capacity was not the constraint when deciding on the changeover sequence. However, now it is the constraint, and you should not waste precious capacity for a wasteful changeover sequence. Hence, figure out all the parts you want to make, and then create a good changeover sequence. You also have to consider whether you want to use larger lot sizes, which means less changeovers but more fluctuations and a higher likelihood of running out of stock. Yet, even then I would still prefer the smaller lot sizes; otherwise there is a risk of running out of stock of the parts that you may need the most.

A problem could also be that your customer does not order enough products. You have more capacity than you need. Here, too, there are some things to consider for your production plan while you reduce your capacity. Often, the feeling on the shop floor is to use the available capacity and simply produce even though there are no orders. Do not produce more than your inventory limit! Do not create more inventory than you need just because you have the capacity! This is overproduction (which is the worst of the seven types of waste), which leads to inventory (which is the second-worst type of waste). This excess inventory can cost you more than the labor to make it. I would be very careful not to increase inventory beyond the target unless I am highly confident that a proportional increase in demand is just around the corner. And even then, if you use a pull system, adjust your target inventory rather than producing in excess of your inventory limit!

Sometimes, changes in the customer demand come unpredictable, but sometimes you know these changes beforehand. Seasonality especially does not surprise seasoned planners. Adjust your capacity beforehand, or even switch parts temporarily from make-to-order to make-to-stock for your season.

Of course, ideally you should be able to prevent any such problems from happening. In reality, however, despite best efforts sometimes your demand does not match your supply, or your machine breaks, or your supplier fails you. Besides fixing the root cause of the problem, you can sometimes use prioritization and changeover sequencing to mitigate the damage. Basically, make what is most important with the constraints you have, and do not overproduce. Now, go out, improve your production sequence, and organize your industry!

3 thoughts on “Changeover Sequencing under Duress: Problems with Make and Deliver”

  1. I loved the your article and the ideas that you brought up. This past summer, I read a book called “The Goal” which covered the topic of the theory of constraints. This post was a reminder of the key fundamentals that book preached. First, I love the message you brought up about not increasing inventory. People often think more inventory means more goods for when a rush occurs, but little do they know you are using resources on goods that may not be needed and using up precious space and machine capacity. The second part was that as supply chain professionals, the unexpected often happens. Individuals must be ready to adapt at a seconds notice. In regards to your comment of an order of 3000 of one SKU popping up. Would a potential idea for the future is see the popularity/rate of goods being ordered and create a safety stock to prepare for this 3000 product purchase? Or would stopping certain products productions and flash producing the high number order be a risk worth taking?

  2. This past summer, I read a book called “The Goal” which covered the topic of the theory of constraints. I loved the your article and the ideas that you brought up. This post was a reminder of the key fundamentals that book preached. First, I love the message you brought up about not increasing inventory. People often think more inventory means more goods for when a rush occurs, but little do they know you are using resources on goods that may not be needed and using up precious space and machine capacity. The second part was that as supply chain professionals, the unexpected often happens. Individuals must be ready to adapt at a seconds notice. In regards to your comment of an order of 3000 of one SKU popping up. Would a potential idea for the future is see the popularity/rate of goods being ordered and create a safety stock to prepare for this 3000 product purchase? Or would stopping certain products productions and flash producing the high number order be a risk worth taking?

  3. Hi Cristian, thanks for the comment. As for the 3000 of one SKU, in my example this was merely done to reduce changeover time, and served as an example that you should rather go for a batch of 500 and change over more often to reduce inventory. If you have batches of 3000 less frequently you need more inventory than if you have batches of 500 more often. Of course, both have to match the demand, and you should have some safety stock anyway to cover fluctuations You just need more for batches of 3000.

Leave a Comment

Cookie Consent with Real Cookie Banner