Document Your Findings
Once you have collected the data, compile it into a usable format to make decisions. Figure 3-3 contains a worksheet for recording the process information. You can use this form to record your data.
Figure 3-3: Process Information Summary Sheet.
In Chapter 1 we alluded to value stream mapping as a product that would help in the selection of a target process. If your plant has not done value stream mapping, then this might be a golden opportunity to do so after you have collected all the above data. The value stream mapping process asks an organization to assess its current situation in the form of a "Current State" map. The organization then asks itself what it wants to become by developing its "Future State" map. When done properly with real data, the Future State map becomes a powerful roadmap to the future.
For the purposes of this book, we have not used value stream mapping to document the current production process, because value stream mapping asks the organization to look at and utilize all the Lean tools. Kanban is just one of those tools. We will assume that the organization has decided to implement kanban as part of its strategic plan (regardless of how the organization arrived at the plan).
Analyze the Data
Once you have the data consolidated, then it's time to take a look at the data. Analyze the data for:
By consistency , we mean:
Are the units correct and are the numbers consistent with your expectations?
Is everything recorded in the same measurement system, i.e., English or metric?
Are the units mixed?
Are the entries in the same dimensionse.g., seconds versus minutes?
Nothing slows down the kanban design process like trying to decipher unit problems.
When considering accuracy , consider these items:
Are these the true scrap numbers, changeover numbers, downtime numbers?
Have we overinflated the numbers?
Have we lumped all the parts into the same category by giving them all the same changeover time or scrap rate?
Also, look at the data for realism. Does the data really reflect how we run our process?
Make sure the numbers are consistent, accurate, and real. If you have misstated the numbers, it will become evident in the next phase when you calculate the kanban. You will see the errors when the calculated numbers tell you to increase inventory, or worse , when the errors tell you that you can cut your inventory in half. As the old computer saying goes: garbage in, garbage out!
Let's work through two examples to make sure everyone understands the data collection process and how to use the forms. The first example, shown in Figure 3-4, is a simple two-part number kanban for two plastic injection molded parts . Figure 3-5 shows the summarized data. The second example, shown in Figure 3-6, is a process line that makes ten-part numbers ”some that require changeovers and some that do not. Figure 3-7 shows the summarized data.
Example 1: A plastic injection molding machine makes plastic rings for carnival games . Two different colors are produced ”red (part A) and blue (part B). The machine makes 2 rings at a time with a mold cycle of 30 seconds. The process has a system scrap of 3%. The mold is not changed when switching between colors, but the raw material hopper must be cleaned out between colors. This color change takes 15 minutes.
The plastic molding pellets are delivered daily and the supplier has never missed a delivery. The rings must cool for 15 minutes. They are made to order for 5 different customers.
The company has steady weekly orders of 2,500 red rings and 4,000 blue rings. The company promises next -day delivery of any order.
The company produces the rings 1 shift per day, 5 days per week. The plant has two 10-minute breaks and a 30-minute lunch . Each day the shift has a 10-minute shift starter meeting and 5 minutes/day is allocated for the monthly kaizen meeting, 5s activities, and a safety meeting. The plant does not have a preventive maintenance program so they experience 1 hour of downtime each day for minor breakdowns and adjustments.
”Use this information to fill in the process information summary sheet.
Figure 3-4: Example 1: Two-Part Number Kanban.
Figure 3-5: Example 1: Process Information Summary Sheet.
Example 2: A plastic extrusion process line extrudes 10 different profiles. Each profile is unique and is assigned a unique part number ”3502 through 3512. The profiles are used in a downstream process where they are mold spliced to create environmental seals for the company's main consumer product ”electronic component carrying cases. The extrusion process experiences 5% scrap and 2% scrap in the mold splicing process. All profiles have a changeover time of 35 minutes. The demand and speeds for each profile is as shown:
Component Case Demand
Part Length (in mm)
Production Rate (ft/min)
The process line averages 95% uptime. The process line is scheduled 5 days per week for 1 shift. All PMs and 5s activities are handled while the operation is running or on overtime.
The plastic molding pellets are delivered daily, but the supplier misses about 15% of their deliveries. The extruded profiles require no further cure time once they are extruded and cooled in the production process.
Figure 3-6: Example 2: Ten-Part Number Kanban.
Figure 3-7: Example 2: Process Information Summary Sheet.
Take a close look at Figure 3-7 and perform the recommended analysis. Did you find the incorrect numbers? Did you see flags for follow-up? What about the mixed units? Figure 3-8 highlights the errors and flags in Figure 3-7. Figure 3-9 shows the corrected table. This example shows the necessity to carefully analyze your numbers to make sure you have the correct data to correctly calculate the kanban quantities .
Figure 3-8: Example 2: Errors in the Original Data Collection Form.
Figure 3-9: Example 2: Corrected Process Information Summary Sheet.