Quality Management

Overview

Whether a systems engineer delivers a product, a service, or an enterprise, the deliverable should meet the needs of the customer and be fit for use. Such a deliverable is said to be of high quality.

Over the past 80 years, a quality movement has emerged to enable organizations to produce high quality deliverables. This movement has gone though four stages, each discussed below. First , acceptance sampling was developed to apply statistical tests to decide whether or not to accept a lot of material based on a random sample of its content. Second, statistical process control was developed to determine if production processes were stable. Instead of necessarily measuring products, processes were measured. Processes that departed from a state of statistical control were far more likely to develop low quality deliverables. Third, design for quality focused on designing processes that were robust against causes of variation, reducing the likelihood that a process would go out of control and accordingly reducing the monitoring requirements. Fourth, six sigma methods applied the tools and power of statistical thinking to improve other aspects of the organization.

Definitions

The American Society for Quality provides the following definitions:

Quality: A subjective term for which each person or sector has its own definition. In technical usage, quality can have two meanings: 1. the characteristics of a product or service that bear on its ability to satisfy stated or implied needs; 2. a product or service free of deficiencies. According to Joseph Juran, quality means “fitness for use;” according to Philip Crosby, it means `conformance to requirements.'

Acceptance sampling: Inspection of a sample from a lot to decide whether to accept that lot. There are two types: attributes sampling and variables sampling. In attributes sampling, the presence or absence of a characteristic is noted in each of the units inspected. In variables sampling, the numerical magnitude of a characteristic is measured and recorded for each inspected unit; this involves reference to a continuous scale of some kind.

Statistical process control (SPC): The application of statistical techniques to control a process; often used interchangeably with the term “statistical quality control.”

Six Sigma: A method that provides organizations tools to improve the capability of their business processes. This increase in performance and decrease in process variation lead to defect reduction and improvement in profits, employee morale and quality of products or services. Six Sigma quality is a term generally used to indicate a process is well controlled (±6 s from the centerline in a control chart).

Quality Attributes

Quality attributes, also known as quality factors, quality characteristics, or “ilities”, are a set of system’s non-functional requirements that are used to evaluate the system performance. There are a large number of system quality attributes that you can find within the literature (http://en.wikipedia.org/wiki/List_of_system_quality_attributes). Of course, depending on the type of the system you deal with, some of these attributes are more prominent than others. Ideally, you would want to optimize for all the quality attributes that are important to the system, but this is an impossible task. Therefore, it is important to conduct a trade off analysis to identify the relationship between the attributes, and whether a change in one attributes would affect positively or negatively any other attribute. An example of such trade off table is shown bellow.

-
	Flexibility	Maintainability	Reliability
Flexibility		"+"	"-"
Maintainability	"+"		"+"
Reliability	"-"	"+"

Finding the right set of quality attributes is the first step in quality control and management. In order to acheive high quality, quality have to be measured, monitor, manage, and improve on. Therefore; in order to increase the overall system quality, you should be able to;

- Identify the quality attributes

- Prioritize these attributes

- Identify the metrics that can be used for these attributes

- Measure and monitor the attributes

- Validate the measurements

- Analyze the result of those measurements

- Based on the analysis, establish processes and procedures that result in improved system quality

Quality attributes for products

Quality attributes for services

Throughout the SEBOK majority of the discussion concentrates around the products. However, quality of survives also plays a major role in the customer satisfaction, which is the measurement of the overall system quality. Services can be divided to two major categories, primary and secondary. For example, the city public transportation system, the United States postal service, or the medical services provided by the hospital are all examples of the primary services. On the other hand, the services that provide help to the customer, in order to assemble a BBQ grill is an example of a secondary services, which are typically referred to as the customer service. Identifying and appropriate quality attributes is the key step in the quality management for services. Some examples of service quality attributes include; affordability, availability, dependability, efficiency, predictability, reliability, responsiveness, safety, security, usability, etc. Again depending on the type of the service, some of these attributes are more prominent than the others.

For example, in the case of services that is provided by the hospital, one may be more interested in the availability, reliability, and responsiveness than potentially the security (typically hospitals are assumed to be safe) and the affordability (typically insurance covers the majority of the cost). Of course, if the patient does not have a good insurance coverage, then the importance of affordability will increase.

Quality attributes for enterprises

An enterprise typically refers to a large complex set of interconnected entities that includes people, technologies, processes, financial and physical element. It is obvious that a typical enterprise has a number of internal and external stakeholders, and as a result there are a large number of quality attribute that will define its quality. Identifying the right set of attributes are typically more challenging in such a complex system. An example of an enterprise is the air traffic management system that is mainly responsible for the safe and efficient operation of the civil aviation within a country or collection of countries. There is a large number of stakeholders that are concern about the overall quality of the system, some example of these stakeholders and some of the primary quality attributes that they are concern with are identified in the following table.

-
Stakeholders	Primary Quality Attributes
Passengers	Safety, affordability, reliability, etc.
Airlines	adaptability, efficiency, profitability, etc.
Air traffic controller	Safety, reliability, usability, etc.
Hardware & software developers	Reliability, fault tolerance, maintainability, etc.
Government/regulatory agency	Safety, reliability, affordability, etc.

Measuring quality attributes

As previously mentioned, you cannot achieve quality, if you cannot measure. The Measurement System Analysis (MSA) [Wheeler and Lynday 1989] is a set of measuring instruments that provide an adequate capability for the team to conduct appropriate measurements in order to monitor and control quality. The MSA is a collection of;

Tools: measuring instruments, calibration, etc.

Processes: testing and measuring methods, set of specification, etc.

Procedures: policies and procedures and methodologies that are defined by the company and/or regulatory agency

People: personnel (managers, testers, analysis, etc.) who are involved in the measurement activities.

Environment: both environmental setting and physical setting that best simulate the operational environment and/or best setting to get the most accurate measurements

Once the quality attributes are identified and prioritized, then the MSA will help in the monitoring and controling the oversll system quality.

Additional detail about measurement are presented in the measurement section.

Quality management strategies

Acceptance Sampling

In acceptance sampling, a lot of products is presented for delivery. The consumer samples from the lot. Each member of the sample is then either categorized as acceptable or unacceptable based on some attribute (attribute sampling), or measured against one or more metrics (variable sampling). Based on the measurements, inference is made as to whether the lot meets the customer requirements.

There are four possible outcomes of the sampling of a lot.

**Truth Table** - Outcomes of acceptance sampling
	Lot meets requirement	Lot fails requirement
Sample passes test	No error	Consumer risk
Sample fails test	Producer risk	No error

An sample acceptance plan balances the risk of error between the producer and consumer. Detailed ANSI.ISO/ASQ standards describe how this allocation is performed. [ANSI/ISO/ASQ A3534-2-1993: Statistics—Vocabulary and Symbols—Statistical Quality Control.]

Statistical Process Control

Statistical process control (SPC) is a method was invented by Walter A. Shewhart, which adopts statistical thinking to monitor and control the behaviors and performances of a process. It means using statistical analysis techniques as tools in appropriate way to estimate the variation in the performance of a process, to investigate the causes of this variation, and to recognize from the data when the process is not performing as it should.(Mary et al, 2006, p.441). By performance here we mean is ‘how well the process is done’.

The theory of quality management emphasizes to manage processes by fact and to maintain systematic improvement. All product developments are a series of interconnected processes, which have variation in their results. Understanding variation with SPC technology can help the process executors understand the facts of their processes and find the improvement opportunities from a systematic view.

The key tools in SPC are control charts. The control chart is also called Shewhart 3-sigma chart. It consists of 3 limit lines, called center line which is the mean of statistical samples, Upper and Lower control limit lines which are calculated by mean and standard deviation of statistical samples. The observed data points or their statistic value (mean as process behaves) are drawn in the chart with time or other sequence order. Upper and lower control limits indicate the threshold at which the process output will be considered as ‘unlikely’.

There are two sources of process variation. One is common cause variation, which due to inherent interaction among process components. Another is assignable cause, which due to events that are not part of the normal process. SPC stresses that bringing a process into a state of statistical control, where only common cause variation existed, and keeping it in control. The usage of control chart is to distinguish between variation in a process resulting from common causes and assignable causes.

If the process is in control and if standard assumptions are met, more than 99.73% points will plot within the control limit. Any points outside the limits, or appear some systematic patterns, imply a new source of variation would be introduced. A new variation means increased quality cost.

The control limits of control are based on the understanding of the past of process. So it is also called the natural bound of the process. It presents a graphic display of process stability or instability over time, controls the current of the process, and predicts the future of the process.

More advanced control charts exist. Cumulative Sum Charts detect small, persistent step change model departures. Moving average charts, with different possible weighting schemes, also detect persistent changes.

Design for Quality

Variation in the inputs to a process usually results in variation in the outputs. Processes can be designed, however, to be robust against variation in the inputs.

Response surface experimental design and analysis is the statistical technique used to assist in determining the sensitivity of the process to variations in the input. Such an approach was pioneered by Taguchi.

Six Sigma

Six sigma methodology is a set of tools to improve the quality of business processes; in particular, to improve performance and reduce variation.

Six sigma methods were pioneered by Motorola and came into wide acceptance after they were championed by General Electric.

Problems are addressed by six sigma projects. These projects follow a five stage process:

Define the problem, the stakeholders, and the goals
Measure key aspects and collect relevant data
Analyze the data to determine cause-effect relationships
Improve the current process / Design a new process
Control the future state / Verify the design

These steps are known as DMAIC for existing processes and DMADV for new processes.

An extensive literature exists on six sigma.

A variant of six sigma is called lean six sigma, where the emphasis is on improving or maintaining quality while driving out waste.

Standards

Primary standards for quality management are maintained by ISO, principally the IS0 9000 series. ISO 9000 home site. The ISO standards provide requirements for the quality management systems of a wide range of enterprises, without specifying how the standards are to be met, and have world-wide acceptance. The key requirement is that the system must be audited.

In the United States, the Malcolm Baldridge National Quality Award presents up to three awards in six categories: Manufacturing, Service company, Small business, Education, Healthcare, and Nonprofit. The Baldridge Criteria have become de facto standards for assessing the quality performance of organizations.