Using the Workbook
For individuals and groups that are using the CD-ROM Workbook , Chapter 3 on data collection contains a blank form for data collection. We have also included the two examples for reference purposes. Use the data collection form to collect the data for your kanban. In a meeting setting, you may want to make an overhead of the blank form and fill it in as you review the data.
The process of implementing kanban scheduling begins by collecting data on the target process. The data that needs to be collected includes number of parts to be produced by the process, changeover times, downtime, and scrap levels.
Use your team to collect this data. The team should follow the guidelines in the chapter to ensure you get the data needed to calculate the kanban's size .
Once you have collected the data, then use the data collection form to organize the data into a format for analysis. When you have the completed data sheet, then review the data to make sure the units are consistent, correct, and realistic. Look for errors in the data, such as tabulation errors, recording errors, and unit errors. Use the two examples provided at the end of the chapter to help illustrate the data summation and analysis process.
Chapter 4: Size the Kanban
In Chapter 3 you gathered data on your process in preparation for this point. In this chapter we will use the data from Chapter 3 to "size the kanban." Or, in layman's terms, you will determine how many containers of each product you will require to effectively operate the kanban (and keep your customer supplied). Figure 4-1 expands the model started in Figure 3-1 by adding this step.
Figure 4-1: Calculating the Kanban Quantities .
Once you start kanban scheduling, these quantities become the maximum amount of inventory held. Therefore, when all these containers are full, then you stop production. This rule is one of the major tenets of kanban scheduling ”you only produce when you have signals.
These quantities become the scheduling signals. To set up these signals, we will determine the emergency level (or red level), the schedule signal (or yellow level), and the normal level (or green level). In Chapter 5 you will develop a kanban design, which uses these signals to control the production schedule.
In this chapter we propose two methods for sizing internal kanbans. The first method calculates the container quantities based on the data collected in Chapter 3. This method allows you to optimize and potentially reduce the quantities based on the characteristics of your process.
The second method uses your existing production schedule and makes the current production quantities the kanban quantities. This method allows for quicker implementation and less math, but does not offer the potential for reducing inventory levels.
We will discuss the benefits of each approach as we proceed through the chapter. However, our experience tells us that the first approach is better since it calculates the quantities based on the processes' capability. This method also typically leads to reduced inventory. We focus mainly on the calculation method, but also provide the information necessary to perform the second method. We will let you decide on which method to use for calculating your kanban quantities.
This chapter also addresses supplier kanbans. The primary issues with these kanbans will be safety stock and establishing the lead time for replenishment. As part of this activity we discuss kanbans with vendors who have already developed this type of scheduling system as well as kanbans with vendors who have no such experience. During the discussions about supplier kanbans, we show you a methodology for selecting the vendor parts that should be placed under kanban scheduling and how to address variability in order quantity. Additionally, Appendix H contains a case study that is based on a kanban in use at a plant with 3,000 different finished-goods part numbers , with batches as small as fifty units.
All of these initial kanban quantities should be based on your current condition instead of how you want the future to look. By using current data, you can start the kanban now, rather than waiting for your downtime, your scrap level, or your changeover times to be reduced.
Further, by using current data you reduce the risk of failure. Essentially when you size the kanban with wishful or projected future data, you open yourself up to raw material stock outs or production stoppages. Just like the case of overcompensating with buffers, sizing the kanban with future data creates an unreal production condition. Managers who make this mistake often wonder why every product is in the red level all the time.
Don't make the same mistake. Size the kanban with the current data, then set continuous improvement goals that will reduce the kanban quantities. Chapter 9 specifically addresses the various ways to reduce the kanban quantities. Best of all, these recommendations will be compatible with other Lean, or continuous improvement, activities you may already have underway.