(EN) What If Your Customer Takt and Your System Takt Do Not Match—Small Fluctuations

One of the key parameters for setting up and running any production system is the customer takt (i.e., the customer demand). This is measured as the average time between the demand of a customer for one product, based on the available working time. In the grand picture, the customer takt (the customer demand) should match the line takt (the production speed of the line). But what should you do if it doesn’t? Let’s investigate this in a small series of posts. This first post looks at small fluctuations.

Introduction

The customer takt is the average time between the demand of a customer for one product, based on the working time of the production system. To calculate it, you divide the available working time by the customer demand during that time.

\[ Customer \; Takt = \frac{Available \; Work \; Time}{Customer \; Demand \; during \; available \; Work \; Time}\]

(I have plenty of posts on customer takt with more detail, like How to Determine Takt Times, Pitfalls of Takt Times, The Relation Between Inventory, Customer Takt, and Replenishment Time, and more.)

The line takt is then simply the work time divided by the number of good parts produced during that work time. This includes all breakdowns, delays, and other problems.

\[ Line \; Takt = \frac{Available \; Work \; Time}{Produced \; Good \; Parts \; during \; available \; Work \; Time}\]

Hence, it is easy to see why the customer takt and the line takt should match: What you produce and what the customer wants should match!

Small Differences Between Customer Takt and Line Takt

Customer takt and line takt should match. However, in a real-world scenario, they sometimes do NOT match. Actually, if you look very closely, they rarely do match exactly, albeit they usually are close enough. Hence, before we go into detail about bigger mismatches that require intervention, let’s look first at the small mismatches that do not require intervention, or at least not immediately.

Then customer takt is the work time divided by the customer demand. Hence, it is a long-term average. Companies often use a week, a month, a quarter, or longer to estimate the demand. This long-term average averages out the demand fluctuations. If you would go to the extreme and actually measure the time between the customer ordering one item and the next item, you would have huge fluctuations. The duration may range from no time (if the customer orders two at the same time), to quite long time (e.g., at four o’clock in the morning when nobody is ordering, even if your factory may produce).

A production system is simply unable to match such fluctuations. But, fear not, we have a great way to handle these fluctuations, and it does so almost automatically (albeit not necessarily free): Inventory! (for make-to-stock items). Or, if you produce make-to-order: Time!

Small Fluctuations with Make-to-Stock

If you produce make-to-stock, your inventory goes up when the customer temporarily orders less, and it goes down when the customer temporarily orders more. That is what inventory is for, to decouple such fluctuations. You merely have to watch whether you are in danger of running out (if the customer orders more than expected for longer than expected) or are in danger of having excess inventory.

For plant managers, running out is usually the bigger worry, because they get yelled at by the customer. However, in lean manufacturing, excess inventory is also a concern, because it is bloody expensive and also a lot of headache to handle. Hence, lean manufacturing tries to limit the inventory, and one of the key points of a pull production is to have an inventory limit at which you stop production.

But having this inventory to decouple fluctuations in the first place gives you not only a bit of leeway for small fluctuations, but also time to react before you run out or exceed your inventor limit. While inventory buffers almost automatically, running out or reaching the upper limit means you have to take additional measures, which I will talk about in the next post.

Small Fluctuations with Make-to-Order

If you produce make-to-order, you cannot use inventory to decouple fluctuations. Instead, you have time (i.e., you simply let the customer wait). If you have a lot of orders, the lead time and hence the waiting time for the customer will go up. If you have fewer orders, the lead time and hence the waiting time for the customer will go down, which sounds good.. .except that your system utilization will also go down, which is not so good.

These fluctuations in lead time and the subsequent fluctuations in delivery time to the customer will also help you to decouple small fluctuations in demand. But, like inventory, if the lead time becomes too large, or if the utilization becomes too small, you may have to take additional measures to adjust your system, which I will also talk about in the next post.

Summary

So, overall, small fluctuations between the customer takt and the system or line takt cannot be avoided, but are also no problem. Inventory or waiting time will take care of this, and you merely have to watch that your system is not getting lopsided (stock-out or to much inventory for make-to-stock; and too long lead times or too low utilization for make-to-order).

In my next post I will look in more detail at what you can do if your system is getting lopsided and your customer takt does not match your system demand on a larger scale (i.e., your system output does not match your customer demand). But now go out, make sure your system fluctuates only within an acceptable range, and organize your industry!