Measuring and improving throughput.

MEASURING AND IMPROVING THROUGHPUT 

Throughput, the number of good units of an item produced per time period, is the product of three components: productive capacity, productive processing time and yield. Carole Cheatham, CPA, PhD, professor of accounting at Northeast Louisiana University, Monroe, discusses the relationship of these components and how each can be measured and improved, thereby increasing throughput. Three recurring themes in manufacturing are control operating expenses, decrease inventories and increase throughput. Concentrating on improving these areas is enabling many companies to compete effectively in world markets.

Of the three themes, increasing throughput is the most difficult for managers and accountants to deal with because most of them have a business, rather than an engineering, background. While controlling operating expenses and cutting inventories relate to their experience, increasing throughput puzzles them. They need a method for segmenting and measuring throughput.

An effective approach to measuring throughput is to break it down into its three components: productive capacity, productive processing time (PPT) and yield. Analyzing it in this way allows the manager to make improvements by increasing one or more of the components of throughput.

Various factors influence each component. For example, PPT is influenced by queue time, move time, down time, set up, inspection and rework. If these unproductive times can be reduced in relation to the productive processing time, throughput can be increased.

PRODUCTIVE CAPACITY

Productive capacity is the maximum a particular manufacturing cell or segment can produce with a given technology. As shown in exhibit 1 above, productive capacity is measured by dividing total units processed during a time period by the processing time. It is an optimistic view of what could be produced if processing time were all productive and all units produced were good.

In the example in exhibit 1, the manufacturing segment started 10,800 units. The actual time spent working directly on the product was 600 hours. Dividing units started by processing time results in a productive capacity of 18 units per hour.

Productive capacity is affected by a number of factors, some of which are shown in exhibit 2 on page 91. Batch size affects productive capacity. Many manufacturers have found that they previously used batch sizes that were too large. Not only were the lots too unwieldy to handle efficiently in the production process but they also unduly increased the investment in inventory.

Plant layouts also affect productive capacity. In many factories the layout is being shifted from large groupings of similar equipment to smaller "manifactories" or "mini-assembly" lines" in which some of each type of equipment is grouped. This speeds order processing.

In an effort to improve productive capacity, bottlenecks in production processes are being identified and eliminated. In addition, equipment and labor efficiencies are being examined.

There is a new emphasis on designing the product for ease of manufacture. In the past, other design features often overrode this consideration, sometimes resulting in products that could not be produced efficiently. Today's products are using more interchangeable parts than in the past and are less likely to require special operations.

PRODUCTIVE PROCESSING TIME

PPT shows how much time is spent on activities that add value to the product. As shown in exhibit 1, PPT is computed by dividing processing time by the total time the product spends in a particular segment of the factory.

In the example in exhibit 1, the manufacturing segment had 600 hours of processing time, while the total time the product spent in that segment was 1,000 hours. This resulted in 60% PPT.

The process time is measured in direct labor hours or machine hours or a combination of both. However, the emphasis of this measure is on time that actually adds value to the product. Therefore, process time does not include time spent moving the product, the time it is waiting in queue for machinery to become available or any time not devoted to "hands on" operations.

Total time is measured in terms of the clock hours the product spends in a manufacturing segment. It includes all the time spent in queuing, moving, down time, setup, inspection and rework, as well as process time.

Queue time is the time the product spends waiting for a productive operation. Queue time can be reduced if bottlenecks are eliminated and optimal batch sizes used. There is now increased emphasis on the "pull through" concept of production rather than on the older "push through" idea. The sales-driven company pulls its orders through production as orders are received, while the production-driven company pushes through large numbers of units to be inventoried. Using the pull through concept typically reduces both batch sizes and queue times.

Move times are being decreased by improving plant layouts. Down time is reduced by identifying machines or operations that are perpetual trouble spots. Setup times, though necessary, are being dramatically reduced through increased automation and efficiency and improved design features such as the use of more interchangeable parts. In a report of the American Accounting Association on just-in-time systems, one authority sets the standard for setup times at no more than five minutes and cites automobile manufacturers who can change from one car model to another in two and a half minutes.

Reinspection and rework time also are being decreased by an emphasis on quality throughout the production process. Designing a product that can be efficiently produced and purchasing quality materials can reduce reinspection and rework costs substantially by reducing defective products.

YIELD

Yield shows the percentage of good units. It is determined by dividing the good units produced in a particular time period by the total units started during the period. Yield is actually an indicator of quality output.

Exhibit 1 shows 10,000 good units produced out of a total of 10,800 units started, resulting in a yield of 92.6%.

Manufacturers used to emphasize quantity rather than quality of production. Meeting production goals caused manufacturing segments to produce many units that were rejected later, probably at a final inspection point. The costs of these units were not recouped until they had been reworked. In some cases they were scrapped or sold as spoiled units at a substantial loss. Today manufacturers recognize that quality has to be built in rather than imposed following an inspection process.

Quality is being built in by methods such as inspecting raw materials to eliminate the possibility of later rejection of processed units because of defects in materials. Products are being designed with quality in mind. Features that are likely to cause quality problems, such as parts that are difficult to manufacture or use of fragile materials, are modified or eliminated in the design stage.

Quality also is being attained by shifting responsibility for it from the inspection or quality control function to the production function. Workers are encouraged to identify and correct defects. In some plants, workers can stop the assembly line if the process is resulting in defects. Allowing a worker to stop the assembly line would have been unheard of in the old production-driven environment. Allowing it now means a process can be corrected before a mass of defective units is produced.

THROUGHPUT

Throughput, then, is the measure of how many good units are produced during a particular period. Throughput can be calculated in two ways: by dividing good units produced by the total time or by multiplying the other three throughput measures.

Exhibit 1 shows 10,000 good units were produced in a total time of 1,000 hours, resulting in a throughput of 10 units per hour. Throughput can also be computed by multiplying the productive capacity of 18 units by the 60% PPT and the 92.6% yield. In other words, this company has a basic capacity to produce 18 units, but this is modified because only 60% of total time is productive and only 92.6% of the units produced are good units.

The reciprocal of throughput (total time/good units) is average production time. Adding average production time for all manufacturing cells or segments gives cycle time, the time it takes to get the product through the entire process. Increasing throughput for the entire manufacturing operation is the same as decreasing cycle time. Managers also may want to think of decreasing cycle time.

Having short cycle times gives a company a competitive advantage in dealing with customers. It also means working capital needs can be minimized because the product is delivered sooner and cash is collected more rapidly. Increasing productive capacity through elimination of bottlenecks, optimizing batch sizes and increasing efficiency will decrease cycle time as well as increase throughput. Improving PPT by decreasing queue times, move time, setups and other unproductive time will have a positive effect on cycle time. Likewise, increasing yield by concentrating on quality of design, workmanship and materials will improve both throughput and cycle time.

EXHIBIT 2

Factors influencing throughput components

Productive capacity (total units/processing time):

Batch sizes

Plant layout

Bottlenecks

Equipment efficiency

Labor efficiency

Product design

Percentage of productive processing time (processing time/total time):

Queue time

Move time

Down time

Set up

Reinspection

Rework

Yield (good units/total units):

Quality of materials

Quality of workmanship

Inspection

Design features

Tolerances

Focus of responsibility