Process Analysis

Statistical process control applies various statistical techniques to the measurement and analysis of the variations that occur in any production process and to monitor the consistency with which the processes used in manufacturing result in products that are within their design specifications. To control a process, one must understand how the control input affects that process. The first step in analyzing a process is to operationally define what that process is. Once the process is defined in detail, it can next be analyzed. This set of activities helps determine where in a process an error or failure is most likely to occur or where the process is most in need of reengineering. In addition, process analysis helps determine if there are any unnecessary, redundant, or irrelevant activities in the process that can be eliminated.

Keywords Business Process; Business Process Reengineering (BPR); Control Charts; Globalization; Mean; Quality Control; Sample; Standard Deviation; Statistics

Statistics > Process Analysis

Overview

The growing trend toward globalization in the twenty-first century brings with it both advantages and disadvantages. In many instances, globalization means that businesses have larger marketplaces in which to sell their products or services. However, the same global marketplace that offers increased opportunities to one company in an industry offers the same opportunities to the other companies within that industry. As a result, globalization also brings with it increased competition. To have a competitive advantage, therefore, companies need to offer their goods and services not only at a competitive price but also with a high level of quality if they hope to be competitive.

For manufactured products, the effects of globalization can be seen all around. The hookup instructions for the new television or stereo system come not only in English, but also Spanish, French, German, Korean, Japanese, or any number of other languages commonly spoken in the industrialized world. If we look at the box or the identification sticker on the machine, in many cases we will also find that the product was not even manufactured in the United States but halfway around the world. The semiconductor, automobile, and television production markets in particular have been deeply affected by foreign competition. In the mid-twentieth century, the phrase "made in Japan" signified a product of poor quality; in the twenty-first century, in many instances it has become a standard of excellence. It has been argued, in fact, that Japanese manufacturing techniques have surpassed those of the West and have cut deeply into the market share.

Statistical Process Control

Arguably, one of the primary reasons for this change is the increased emphasis on statistical process control. This approach to manufacturing processes emphasizes doing things right the first time rather than having to do things over to correct for faults in the first process. Statistical process control includes analysis of both short-and long-term capabilities, applying statistical methods to keep processes (and product quality) within tolerances, and continually analyzing and improving processes. Statistical process control applies various statistical techniques to the measurement and analysis of the variations that occur in any production process and to monitor the consistency with which the processes used in manufacturing result in products that are within their design specifications. These processes have nothing to do with the quality of the product itself: They only monitor whether or not that product is within specification. Statistical process and quality control are not tools for improving the quality of the design, but help monitor whether or not a product is being manufactured as designed.

Characteristics Common to Manufacturing Systems

Most basic manufacturing processes hold certain characteristics in common. As shown in Figure 1, these include a measurement of the state or condition of the process and a controller that calculates an action based on a comparison of this measured value with a set point (a predetermined or desired value). In addition, basic process control systems typically have an output signal that results from the calculation and that is used to manipulate the process through an actuator. The process itself reacts to this signal and changes its state or condition (e.g., correcting for errors or drifts in the process). Two of the most important signals used in process control are the process variable and a manipulated variable.

• The process variable — or output of the process that is to be controlled — is automatically measured by a device in the field. The value of the process variable is used as an input for an automatic controller that takes action based on this value (i.e., either correct the process or let it continue as it is).

• To control the process variable, the manipulated variable (also referred to as the control input) is adjusted. For example, if one needed to control the flow of a liquid into a tank, one would manipulate the position of the value (i.e., the manipulated variable) that controls the flow (i.e., the process variable).

Control InputTo control a process, one must understand how the control input affects that process. For example, if the input conditions are changed, it needs to be determined whether the output will rise or fall, the level of response that can be expected in the process, the length of time necessary to observe a change in the process based on the manipulation of the input, and what response curve or trajectory of the response is expected. Capability studies are used to determine the success or failure of a process and whether or not the use of statistical process control techniques or other action is appropriate. However, simply having a process under control is no longer sufficient in many situations for a company to maintain its competitive advantage. Modern organizations need to continually improve their business processes in order to stay ahead of the competition. Statistical quality control uses statistical techniques to measure and improve the quality of processes.

Analysis of Processes

Process Definition

The first step in analyzing a process is to operationally define what that process is. This activity is followed by process analysis, which is leveraged into an examination of ways to make improvements in the process. Defining the process involves observing the flow of manufacturing or service. The development of a process chart can be helpful in this activity, particularly when defining service processes whose flow may not be obvious at first glance.

• The first step in developing a process chart is to capture the total process from start to finish.
• Once this high-level description of the process has been captured, the next step is to fill in the details at the task or process operation level. This includes the specific actions taken at each stage in the process including any exceptions or subroutines.
• This definition is then further fleshed out to the procedure or process detail level that captures the individual steps within a task (e.g., individual hand movements). A sample total process chart is shown in Figure 2.

Process Analysis

Once the process is defined in detail, it can next be analyzed. This set of activities helps determine where in a process an error or failure is most likely to occur or are most in need of reengineering. In addition, the process analysis helps determine if there are any unnecessary, redundant, or irrelevant activities in the process that can be eliminated. As with all quality control engineering endeavors, it is important that a systematic approach be taken to analyzing the process. Preconceived notions about the process should be set aside, and actions should be recorded as factually as possible. The analysis should examine every aspect of the process, including small details. Problems in small details can often be compounded and affect larger systems. It is important also not to rush to judgment in process analysis. Hunches and premature conclusions should be avoided and new methods or approaches not considered until the entire process is analyzed in detail so that all problem areas can be exposed and dealt with simultaneously. Some of the key questions to ask during process analysis are shown in Table 1.

Table 1: Process Analysis Critical Examination Procedure: (Adapted from Beckford, 2002, p. 217)

Applications

Controls

Before a process can be improved, it must be first analyzed to determine whether or not it is in control. One of the primary sets of tools used by quality control engineers to determine whether or not a process is in control (i.e., stable) is Shewhart control charts. These tools are simple graphing procedures that help quality control engineers and managers monitor processes and determine whether or not they are in control. Control charts are based on the statistical ideas that random noise is naturally occurring and will be evident in any manufacturing process and that, within a random process, there is a certain amount of regularity. Because of this regularity, a variable will differ from its mean by more than two standard deviations only 5 percent of the time (i.e., one occurrence in 20). A process is considered to be within statistical control if it performs within the limits of its capability according to these parameters.

Shewhart Control Charts

Shewhart control charts are used to make measurements and to check for compliance. The X-bar chart (so named because the mathematical symbol for the arithmetic mean is X¾ ) is a graphical display of the means of some characteristic of the product (e.g., acceptability of solder joints) of small random samples taken from the production line over time. The means taken from the random samples are plotted over time on a chart that contains a center line (i.e., the mean for the process) and upper and lower control limits. The center line on the chart shows the location of the arithmetic mean of the means of the samples. The upper control limit on the chart is a line showing the location of three standard deviations above the center line, and the lower control limit shows the location of three standard deviations below the center line. If all the points plotted on the chart fall between the upper and lower control limits, the process is considered to be in control. If computed sample means fall outside the control limits, the process is considered to be out of control, and the process is stopped so that an assignable cause can be determined. Some assignable causes are easily explained by passing phenomena that are unlikely to occur again (e.g., a power outage, breakdown of a machine, new employee). Other assignable causes, however, are more serious and require corrective action (e.g., replacing a defective part or machine, retraining of employees, switching suppliers). A sample X-bar chart is shown in Figure 3. Other types of Shewhart quality control charts include R charts that keep track of the range, p charts track the proportion of defective products, c charts track the number of defects, and s charts that examine sample variance.

Additional Control Charts

In addition to Shewhart control charts, there are more sophisticated charting methods available to help the quality control engineer determine whether or not a process is in control. Multivariate charting methods are available that allow the quality control engineer to monitor several related variables simultaneously. Other methods are available for charting a single measurement rather than a sample (e.g., moving average charts, exponentially weighted moving average charts) and cumulative sum methods that are more sensitive than Shewhart control charts for detecting small, consistent changes.

Uses of Control Charts

Quality control engineers examine process control charts in order to help determine how to continually improve the process. There are a number of different statistical approaches that can be used to help determine various control problems in processes. These include unstable patterns of variation, short- and long-term variation, changes in the average value of the output, erratic averages, measurement accuracy and variation, cyclic occurrences, individual points that are out of control, trends, and stratification. Unless a process is in control, capability of the process cannot be statistically calculated.

Case Study: Process Analysis within a Bakery

Beckford (2002) discusses a case study of a process analysis performed at a commercial bakery that supplies baked goods to various supermarket chains. As part of the process analysis, a map of the finishing process (Figure 4) was created to act as a guide to understanding the process and to form a basis for both quality and productivity measurement. Although the map appears to show a simple process, it shows only the high-level "total" process activities and does not include details shown in lower-level maps. Supporting maps were also created to show the details of the individual subtasks with the process. For example, the "decorate" subprocess actually comprises between five and seven subprocesses depending on the product being made. Each of these subprocesses in turn comprised multiple subroutines all of which were directly related to the product. The maps provided management with information about the processes that they had not previously understood. For example, "Inspect 1" was actually an informal process automatically performed by supervisors. Before the process analysis was undertaken, management had assumed that the reject rate for baked goods going through this process was 10 percent. This figure was derived by taking the difference between the number of cakes entering the "cut" process and those exiting the "pack" process. However, there was no measure of intra-process rejects or rework, both expensive items for the bakery. After the process analysis was performed and these things were taken into account, it was found that the actual reject rate was closer to 35 percent. The process analysis helped management obtain a clearer picture of how the production process actually works and enabled them to quantify a problem that they had only previously suspected. As a result of the analysis, it was found that the "decorate" subprocess was causing the most difficulty and needed to be reengineered.

Conclusion

To gain or maintain a competitive advantage in the twenty-first-century global marketplace, it is imperative that the processes of the organization be under control and working as efficiently as possible. Statistical process control applies various statistical techniques to the measurement and analysis of the variations that occur in any production process and to monitor the consistency with which the processes used in manufacturing result in products that are within their design specifications. Before these techniques can be applied, the process needs to be operationally defined. Once the process is defined in detail, it can next be analyzed to determine where in a process an error or failure is most likely to occur or if there are any unnecessary, redundant, or irrelevant activities in the process that can be eliminated. Process analysis can show where reengineering efforts are most needed so that the organization's processes can be continually improved.

Terms & Concepts

Business Process: Any of a number of linked activities that transforms an input into the organization into an output that is delivered to the customer. Business processes include management processes, operational processes (e.g., purchasing, manufacturing, marketing), and supporting processes, (accounting, human resources).

Business Process Reengineering (BPR):

A management approach that strives to improve the effectiveness and efficiency of the various processes within an organization.

Control Charts: A family of quality control charting techniques that help determine whether or not a process is under control. X-bar charts keep track of processes by examining the means of samples, R charts keep track of the range, p charts track the proportion of defective products, c charts track the number of defects, and s charts examine sample variance. Also called “Shewhart control charts” after their originator.

Globalization: Globalization is the process of businesses or technologies spreading across the world. This creates an interconnected, global marketplace operating outside constraints of time zone or national boundary. Although globalization means an expanded marketplace, products are typically adapted to fit the specific needs of each locality or culture to which they are marketed.

Mean: An arithmetically derived measure of central tendency in which the sum of the values of all the data points is divided by the number of data points.

Quality Control: A set of procedures or processes that help to ensure that products or services comply with predefined quality criteria or otherwise meet the requirements of the client or customer. Quality control activities include the collection and statistical analysis of data to determine whether the process includes systematic (i.e., nonrandom) variation in quality. Quality control activities include monitoring and inspecting products or services vis à vis predefined specifications or quality standards, determining the cause of variation, and developing and implementing changes to help meet target quality goals.

Sample: A subset of a population. A random sample is a sample that is chosen at random from the larger population with the assumption that such samples tend to reflect the characteristics of the larger population.

Standard Deviation: A measure of variability that describes how far the typical score in a distribution is from the mean of the distribution. The standard deviation is obtained by determining the deviation of each score from the mean (i.e., subtracting the mean from the score), squaring the deviations (i.e., multiplying them by themselves), adding the squared deviations, and dividing by the total number of scores. The larger the standard deviation, the farther away it is from the midpoint of the distribution.

Statistics: A branch of mathematics that deals with the analysis and interpretation of data. Mathematical statistics provides the theoretical underpinnings for various applied statistical disciplines, including business statistics, in which data are analyzed to find answers to quantifiable questions. Applied statistics uses these techniques to solve real-world problems.

Bibliography

Alsup, F., & Watson, R. M. (1993). Practical statistical process control: A tool for quality manufacturing. New York, NY: Van Nostrand Reinhold.

Altinkemer, K., Ozcelik, Y., & Ozdemir, Z. D. (2011). Productivity and performance effects of business process reengineering: A firm-level analysis. Journal of Management Information Systems, 27, 129–162. Retrieved November 27, 2013 from EBSCO online database Business Source Premier. http://search.ebscohost.com/login.aspx?direct=true&db=buh&AN=60148145

Altmann, W., & Macdonald, D. (2005). Practical process control for engineers and technicians. Boston, MA: Newnes.

Beckford, J. (2002). Process analysis. In Quality (pp. 211–219). London, London: Routledge. Retrieved September 11, 2007, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=17444164&site=ehost-live

Damij, N., & Damij, T. (2013). Process management: A multi-disciplinary guide to theory. Berlin, Germany: Springer. Retrieved November 27, 2013 from EBSCO online database eBook Collection (EBSCOhost). http://search.ebscohost.com/login.aspx?direct=true&db=nlebk&AN=654191&site=ehost-live

Sherman, P. J. (2012). Smart charting. Quality Progress, 45, 64. Retrieved November 27, 2013 from EBSCO online database Business Source Premier. http://search.ebscohost.com/login.aspx?direct=true&db=buh&AN=78296264

Suggested Reading

Alwan, L. C. (1999). Statistical process analysis. Boston, MA: McGraw-Hill/Irwin.

Kuroki, M. (2012). Optimizing a control plan using a structural equation model with an application to statistical process analysis. Journal of Applied Statistics, 39, 673–694. Retrieved 27/11/2013 from EBSCO online database Business Source Premier. http://search.ebscohost.com/login.aspx?direct=true&db=buh&AN=70708252

Petinis, V. V., Tarantilis, C. D., & Kiranoudis, C. T. (2005). Warehouse sizing and inventory scheduling for multiple stock-keeping products. International Journal of Systems Science, 36, 39–47. Retrieved September 11, 2007, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=15963761&site=ehost-live

Tsung, F., Zhou, Z., & Jiang, W. (2007). Applying manufacturing batch techniques to fraud detection with incomplete customer information. IIE Transactions, 39, 671–680. Retrieved September 11, 2007, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=24471429&site=ehost-live