The Lean manufacturing, PPT Plus,
"Back-to-Basics" e-Training Package 
Lean Manufacturing Solutions

Increase the effectiveness of your
Lean Manufacturing Training Program
Lean Manufacturing Simulation Game 

Control Group Cycle Counting

Control group cycle counting is a very special type of cycle counting. It is the only one of the four discussed that is not a measure of inventory record accuracy, except for the specific parts and records involved in the control group, but rather a measure of the inventory process. (Note: The inventory process is the physical movement of materials into and out of "stock," and the record keeping process that tracks the material's movement.)

Control group cycle counting entails counting the same parts in the same locations repeatedly within a very short interval. The parts counted are referred to as the "control group." Its purpose is to verify a change in the process by reducing the time between cycle counts of a part, thereby facilitating the sleuthing or finding of error causes. When an extended period of time elapses between cycle counts for a part and an error is found, there may be hundreds or thousands of candidates for the cause of the error. If this time is reduced the error exposure time is also reduced and the number of candidates can be more easily analyzed and corrected.

Random Sample Cycle Counting

Random sample cycle counting is a sampling technique that selects its sample from the population at random. That is, every part in the population has an equal opportu­nity to be a member of the sample. There are two types of random sample cycle counting techniques: the first with a constant population and the second with a diminishing population.

Constant population can best be illustrated by visualizing a large bucket of Ping-Pong balls, one for each part number in the population, and the part number printed on the ball. After thoroughly mixing the Ping-Pong balls, the practitio­ner chooses the parts to be cycle counted—the sample—by randomly selecting Ping-Pong balls from the bucket. Upon completion of cycle counting activities, all Ping-Pong balls would be returned back into the bucket. Each new sample would be drawn from the same original bucket. This process is analogous to selecting numbers for a lottery. Statistically, each number has the exact same opportunity of becoming part of the next sample. This method of cycle counting is generally accepted as the best measure of inventory record accuracy, as it is totally non-biased.

Cycle counting with a diminishing population uses two buckets, one for 'counted parts' and one for 'non-counted parts." Samples are always selected from the 'non-counted parts' bucket until it is empty. When this occurs, the process is reversed guaranteeing that all members of the population will be sampled periodically.

ABC Cycle Counting

An alternative or variation to random sample cycle count­ing is ABC Cycle Counting. This method uses the Pareto Principle, which states: where there are a large number of contributors to a result, the majority of the result is due to a minority of the contributors. The ABC cycle counting process puts a bias on the samples to reflect the monies invested in the inventory. Simply explained, the 'A' class parts (the 20% of the inventory that represents 80% of the dollars spent per year for inventory) are counted more frequently than the 'B' and 'C' class parts that have lesser annual expenditures. A typical distribution of inventory and cycle counting frequencies for a stockroom of 10,000 part numbers is shown:

In sampling the 10,000 part numbers, the population is not treated as homogeneous as in random sampling cycle counting. That is, all the parts are not equally weighted in a single population. Rather the total population is broken into three separate populations, one representing each cost classification. A daily sample size of 32 is then taken from the 'A' population, 24 from the 'B' population and 20 from the 'C' population. By design this method of cycle counting will count the equivalent of 190% of the total parts within the stockroom each and every year, and it is biased towards those parts with the greatest annual inventory expenditures.

The practitioner must also realize 42% of the cycle counting effort (8000 -f-19000 = 42%) is spent on counting only 20% of the total part number population, while 26% is spent on 50%. Arguments can be made both for and against ABC cycle counting. For obvious reasons, those responsible for reporting the financial position of a company prefer it. From an operations point of view, it clearly shortchanges 50% of the parts. And when a product is being built, all parts are needed to finish it regardless of the part's cost classification or annual usage. In practice, the 'A' cost classification parts do not offer and have never really offered the practitioner a major operational problem with respect to inventory record accuracy. The problems that surface are more likely to arise with the 'B' and 'C' cost classification parts. Thus this method favors accounting and shortchanges operations.

To be Continued