Takt time in a nonrepetitive manufacturing environment

Analyzing takt time makes sense in low-product-mix production, but it also makes sense in a high-product-mix environment—if you take the right approach with the lean manufacturing tool. Getty Images

In last month’s column we looked at the role of takt time in an operation where the same or similar products run all the time across a machine or cell. In these situations, the value of takt time is easy to see.

But what about an operation that has a variety in products, volumes, complexity of processing, and times to produce? This might sound like your fabrication shop.

For instance, suppose you run multiple products through a press brake. Some of the products are quick and simple, while others have more complex bends that take longer to produce. Maybe the volume mix for these products changes from month to month. In short, you must account for a great deal of variety. Even so, you can still use takt time for effective, objective resource management. Here’s how.

Planning Resources at the Brake

Let’s use an example to flesh out the takt time technique for planning and managing resources at the press brake. The volume of some products is high enough that you can plan for it as separate time in your takt time analysis. Other products are lower volume but exhibit similar processing characteristics (cycle times and levels of complexity). One way to handle this is to put these products in a family. The mix begins to shape up as follows:

Product No. 123: Single part number requires 200 pieces per day on a predictable basis. Cycle time is 10 seconds per piece.

Family No. 1: Comprised of different part numbers with short cycle times and simple bends. Cycle times are in a tight range around a mean of 15 seconds. The requirements for a typical day are 100 pieces from a combination of part numbers in this family.

Family No. 2: Comprised of different parts that are a bit more difficult to produce. They have tight tolerances and multiple bends requiring longer cycle times. Cycle times are in a range around a mean of 75 seconds. The requirements for a typical day are 150 pieces from a combination of part numbers in this family.

Family No. 3: This family of parts is a real challenge. Characteristics might include large parts requiring handling assistance, bends that are difficult to check, tight tolerances, and compound angles. Cycle times range around a mean of 300 seconds per piece. The requirements for a typical day are 45 pieces from a combination of part numbers in this family.

For purposes of this example, let’s assume that you run one eight-hour shift. You provide 30 minutes of lunch/break time, so the available time for the press brake is 450 minutes per day. The lean purist would say you have 450 minutes for producing parts, but the reality is you have to deal with some changeovers during the day.

Changeover is a form of non-value-added work. Of course, you have to do changeovers, but you should always strive to drive them down to as close to zero as possible. But for this example, you typically do six changeovers per day that total 60 minutes. Now your practical available time is 390 minutes per day.

Figure 1
This chart shows the required daily throughput for a press brake producing a variety of parts.

You can appreciate that this is not the tickety-boom process in high-volume production where you are making the same parts day after day. To apply our takt time analysis correctly, we need to take a different approach.

Set up a Model

As a refresher, the equation for takt time is the available time divided by units of demand. If you are producing only one part number during the day, the demand is simple. It is the number of that product you need to produce to meet customer demand.

In our mixed-model, nonrepetitive environment, the demand is a bit more complicated. Demand is focused on “equivalent units.” The equivalent unit absorbs all the cycle time variation from the mix of products and results in some quantity of units of demand.

Figure 1 shows the required throughput for a typical day at the press brake. Note that the chart shows the target cycle time for each of the groupings. You might have some variation among parts within a family, but this should be close enough to provide relevant conclusions. The number-of-pieces-produced column shows the expected customer demand for each grouping.

The final column shows the equivalent units in minutes. The 200 pieces per day for Product No. 123 with a cycle time of 10 seconds represents 33.3 minutes of work—200 × (10 seconds/60)—where the baseline for the equivalent unit is one minute. The same logic applies to each of the three families. The result is 470.8 equivalent units of demand per day on the press brake.

The takt time calculation would be 390 minutes divided by 470.8 units of demand. This gives you a takt time of 0.83 minute/equivalent unit of demand. The average cycle time per piece is the total number of minutes divided by total number of pieces: 470.8 minutes / (200 + 100 + 150 + 45) = 0.95 minute/piece.

The resulting chart in Figure 2 shows the takt time (pace you need to run to meet customer demand) as 0.83 minutes/equivalent unit of demand. The cycle time bar shows the actual run rate at 0.95 minute/equivalent unit. Since the cycle time is greater than the takt time, there is a problem. You cannot meet customer demand, so something has to give.

To shorten the cycle time to less than the takt time, you either have to create additional available time (increase the numerator) or reduce some of the demand (reduce the number of demand units).

Longer lead times often aren’t an option these days, but you might be able to shift some work to a similar press brake with available capacity. So let’s assume the demand can be changed.

Consider our new takt/cycle time chart in Figure 3. Notice that the demand for Family No. 3 is reduced from 45 to 30 units a day. Since this family has the longest cycle times, the reduction of a few pieces has a big impact on the equivalent units. The new equivalent unit value is 395.8 minutes. Now the cycle time is less than the takt time. This means the press brake will be able to meet the customers’ demand (see Figure 4).

Figure 2
The cycle time at the press brake that processes a specific mix of parts is 0.95 minute. That’s more than the 0.83-minute takt time, so there’s a problem. Something has to give.

Insights from the Takt Time Model

This analysis has demonstrated how you can use takth time in a nonrepetitive environment with a high mix of products that have varying cycle times. Using the “equivalent units” technique, you can model the operation to draw meaningful conclusions about the press brake’s ability to satisfy your customer.

Although you may have 30 or 40 (or more) different part numbers to produce on the press brake, grouping parts into families allows you to identify a few groups that are much more reasonable to manage and plan for takt time, capacity, and ability to meet customer demand purposes. The families take some of the noise out of the planning and analysis process.

You can perform the cycle time analysis periodically to take the pulse of the operation. The frequency depends on factors such as changes in demand for existing customers, gaining new customers, changes to cycle times (new tooling or modified process that shortens cycle time), or remixing the demand across other machines. If things are reasonably stable, then monthly or quarterly might be fine. If you have a lot of churn, you may need to do this analysis more frequently.

Note that this takt time and cycle time analysis is simply one more tool in the lean toolkit—and if you’ve just started your lean journey, it probably won’t be the first tool you use, especially if cycle times vary unpredictably.

That said, takt time analysis shouldn’t be ignored, even in a high-product-mix environment. If you apply it correctly in the right situation, you can use takt time to sharpen your understanding of the manufacturing process.

Slide Show | 5 Images