Section 3. Too Much Variation in the Cycle Time of a Single Step

3. Too Much Variation in the Cycle Time of a Single Step

Overview

Processes are required to cycle at a fast enough rate to generate entities at a pace to meet customer (or market) demand. Takt Time represents the pace of customer (or market) demand. For more details see "TimeTakt Time" in Chapter 7, "Tools." Even if, on average, the Cycle Time (the actual rate of processing entities) of the process is shorter than the Takt Time and the process on average meets demand, the Cycle Time could be variable enough that in the short term it does not meet Takt and thus affects delivery performance. See also "TimeGlobal Process Cycle Time" in Chapter 7.

This category infers that some work has been done to relate the Takt Time to the Cycle Time. If this is not the case, then return to Chapter 3 to select the Problem Category for the process as a whole.

Measuring Performance

The measures used should have been put in place at the Global Problem Category level, but if not, then use the Process Cycle Time as the metric, calculating both its mean and standard deviation (the aim being to reduce both).

Tool Approach

If this hasn't already been done at a previous step, then:

	Focus should just be on measuring validity (a sound operational definition and consistent measure) of Cycle Time versus a detailed investigation of Gage R&R. See "TimeGlobal Process Cycle Time" and "MSAValidity" in Chapter 7 for more detail.
	For a baseline that takes into account the variability in the Cycle Time, it is best to capture at least 30 to 50 data points. Take the mean and standard deviation of the Cycle Time for the baseline.

Even though we often consider a process step of this type to be the smallest, indivisible unit of a process, there are usually sub-steps within each step. It is often within these sub-steps that there are NVA activities causing the variability in Cycle Time.

This involves an excruciating level of detail for some, but in the majority of cases identifies opportunities where the process step waits for something (usually information, materials, testing, and so on).

From the VSM, identify actions to remove any NVA activity from the process step. It is less likely to encounter variability in the VA portion of a step than in the NVA elements, so reduction of NVA activity usually significantly reduces variation in Cycle Time.

If from the VSM, significant sources of NVA activity were removed and the Team feels the issue might have been resolved, then repeat the Capability Study. If the variability is now at a manageable level, then proceed to Chapter 5.

If not, then continue in this roadmap.

The roadmap to solution from here on relies on the equation Y=f(X₁, X₂,..., X_n), where the Ys are the Cycle Time and the Primary Performance characteristic(s) of the process. For instance, if in a chemical production process we are interested in the assay of the product as a Primary Performance characteristic, then the Ys are assay and Cycle Time.

Go to Section C in Chapter 3 and focus on this single process step to determine which Xs can be manipulated to gain the best level of performance for the Ys (including Cycle Time) and minimize the variability.