Design of Experiments (DOE): Applications and Benefits in Quality Control and Assurance

Abstract

This chapter explores the applications and benefits of Design of Experiments (DOE) in the context of quality control and quality assurance. DOE is a statistical methodology that enables researchers and practitioners to systematically investigate and optimize processes, identify critical factors affecting quality, and reduce variability and waste. This chapter begins by introducing the overview and definitions of DOE, covering topics such as the history of DOE, types of DOE, steps involved in conducting DOE, and key components of DOE. The specific applications of DOE in quality control and quality assurance were explored, highlighting their importance across various industries. It demonstrates how DOE can be effectively applied to optimize products and processes, reduce defects and variation, improve quality, implement Six Sigma, and validate and verify processes. It then delves into the specific applications of DOE in quality control and assurance, highlighting its significance in various industries and sectors. Furthermore, the book addresses challenges and considerations in implementing DOE in real-world scenarios, such as resource constraints, experimental constraints, and data analysis complexities. It provides basic information on software tools commonly used in DOE.

1. Introduction

Quality control and assurance are crucial aspects of any manufacturing or industrial process. Ensuring high-quality products and services is essential for customer satisfaction, brand reputation, and overall business success. Quality is a measure of the level of conformance of a product to design specifications or the ability of a product or service to satisfy user requirements. The duo of quality assurance and quality control helps deliver a defect-free product or service. Quality assurance focuses on preventing defects by ensuring the approaches, techniques, methods, and processes designed for the projects are implemented correctly. Quality control, on the other hand, focuses on identifying defects by ensuring that the approaches, techniques, methods, and processes designed in the project are followed correctly [1]. Quality assurance is process-oriented and a managerial tool, whereas quality control is product-oriented and a remedial tool, according to [1]. One powerful tool used in quality control and assurance is the design of experiments (DOE). According to engineers and technologists, they often make use of DOE methodologies for various applications ranging from the design of new products, improvement of design, maintenance, control and improvement of manufacturing processes, maintenance and repair of products, and several others [2, 3, 4]. This chapter aims to explore the applications and benefits of DOE in quality control and assurance. Design of experiments (DOE) is a statistical method for planning and conducting experiments. DOE is used to identify the factors that affect a process and to determine the optimal levels of those factors, as shown in Figure 1. DOE can be used to improve the quality of products and processes, reduce costs, and increase efficiency [5]. Businesses and manufacturing companies can use Design of Experiments (DOE) in a variety of ways to differentiate themselves from the competition by constantly redesigning their products or creating new products to establish a presence in other markets. First, DOE can be used to identify the factors that most affect the quality of a product. By understanding which factors are most important, businesses can focus their efforts on improving those factors. This can lead to a product that is more reliable, durable, and user-friendly than the competition’s products. Second, DOE can be used to reduce the cost of manufacturing a product. By identifying the most efficient way to produce a product by optimizing manufacturing processes using DOE methodologies, businesses can improve the quality of their products and save money on labor, materials, and other costs [6]. This can lead to a lower price for the product, which can make it more competitive. Third, DOE can be used to develop new products that meet the needs of a specific market. By understanding the needs of the target market, businesses can develop products that are more likely to be successful. This can help businesses gain a foothold in new markets and increase their market share.

Figure 1.

Design of Experiment (DOE).

1.1 Objective of the chapter

The objective of this chapter is to provide an in-depth understanding of DOE and its applications in quality control and assurance. We will explore various experimental designs, statistical techniques, and methodologies that are commonly used in DOE. Additionally, we will discuss its practical applications across various fields and the benefits and advantages that DOE offers in ensuring and improving quality standards.

1.2 An overview and definitions of design of experiments (DOE)

Definitions of Design of Experiments (DOE): DOE is a statistical methodology used to systematically plan, conduct, analyze, and interpret experiments to obtain valid and reliable results. It allows researchers to efficiently explore and identify the significant factors influencing a process or product’s performance. DOE is an important statistical method used in controlling input factors or variables in order to ascertain the level of relationships with the output (responses) according to Figure 1, so as to ensure product or process quality. DOEs are usually carried out in five stages [7, 8] as shown in Figure 2. They are:

Figure 2.

Five stages of DOE.

1.3 History of DOE

DOE has its roots in the work of Sir Ronald Fisher, who developed the basic principles of DOE in the early twentieth century. Fisher’s work was initially applied to agricultural research, but it was soon adapted for use in other fields, including manufacturing, engineering, and medicine [9]. Since then, many scientists and statisticians have contributed to DOE development and its application in different fields [9, 10, 11, 12].

1.4 Types of DOE

There are many different types of DOE, each with its own strengths and weaknesses [8, 13]. The best type of DOE to use will depend on the specific situation. Factors to consider include the number of factors, the number of levels for each factor, the desired level of confidence, and the time and budget constraints (Figure 3) [10, 14, 15, 16].

Figure 3.

Types of DOE.

The most common types of DOE are:

1.4.1 Full factorial designs

These designs involve testing all possible combinations of factors. For example, if there are two factors with two levels each, there would be four possible combinations (2 × 2 = 4). Full factorial designs are the most comprehensive, but they can also be the most time-consuming and expensive.

1.4.2 Screening/fractional factorial designs

Fractional factorial experiments are a type of factorial experiment that uses fewer experimental runs than a full factorial design. These designs involve testing a subset of the possible combinations of factors. Fractional factorial designs are less comprehensive than full factorial designs, but they can save time and money.

1.4.3 Response surface methodology (RSM) designs

RSM is a type of DOE that is used to fit a mathematical model to the response variable. RSM can be used to identify the optimal levels of the factors and predict the response variable for new combinations of factors. These designs are used to study the relationship between a response variable and multiple factors.

1.4.4 Mixture designs

These designs are used to study the relationship between a response variable and multiple factors that are mixed together. Mixture designs are often used in the food and beverage industry to understand how the different ingredients in a product affect the taste, texture, and other properties of the product.

1.4.5 Taguchi designs

These designs are a type of fractional factorial design that is specifically designed for quality improvement. Taguchi designs are often used in manufacturing, where it is important to produce products that meet the required quality standards.

1.5 Steps involved in conducting DOE

In order to obtain good results from a DOE, the following 8 steps shown in Figure 4 below are necessary:

• Define clear objectives

• Select Process variable

• A feasible experimental design must be selected.

• Execute the selected design.

• Ensure that the data are consistent with the experimental assumptions.

• Analyze and interpret the results.

• Results presentation, and application for decision making.

• Conclusions

Figure 4.

Steps required in DOE.

In the application of the concept of DOE methodology for quality control and assurance, the following terminologies, otherwise known as components of DOE, are commonly used:

2. Key concepts/components of DOE

1. Factors: Variables that may influence the outcome of an experiment.

2. Levels: The values at which factors are set during an experiment.

3. Response Variable: The outcome or output variable that is measured or observed.

4. Experimental Units: The entities or subjects on which the experiments are conducted.

5. Treatment: The combination of factor levels applied to an experimental unit.

6. Replication: The process of repeating the experiment to reduce variability and enhance reliability.

3. Applications of DOE in quality control and assurance

3.1 Product and process optimization

DOE enables the systematic exploration of various factors and their interactions to optimize product and process performance. By identifying the key factors and their optimal levels, manufacturers can improve quality, reduce costs, and enhance efficiency.

3.2 Defects and variation reduction DOE

Helps identify the root causes of defects and variations in a manufacturing process. By conducting experiments and analyzing the results, quality engineers can pinpoint the factors that contribute to defects and develop strategies to reduce or eliminate them.

3.3 Quality improvement and six sigma

DOE is an integral part of Six Sigma methodologies, which aim to achieve process excellence and reduce variation. By using DOE, organizations can identify critical process parameters, set optimal levels, and implement strategies to minimize defects and variations, thus improving overall quality.

3.4 Process validation and verification

DOE plays a crucial role in the validation and verification of manufacturing processes. By conducting designed experiments, organizations can gather data on process performance, determine critical process parameters, and establish robustness and reliability of their processes.

4. Key factors affecting quality optimization processes and waste reduction

Optimizing quality processes and reducing waste is a crucial aspect of many industries. While the specific factors influencing these processes can vary depending on the context, here are some key factors that commonly affect quality optimization and waste reduction.

• Process Design and Standardization: Well-designed processes with clear specifications and standard operating procedures (SOPs) play a vital role in optimizing quality and minimizing waste. Factors such as process layout, equipment selection, workflow efficiency, and error-proofing mechanisms can significantly impact the quality of output and waste generation [17, 18].

• Quality Control and Monitoring: Effective quality control measures, including robust inspection protocols, real-time monitoring systems, and statistical process control (SPC) techniques, help identify and rectify quality issues promptly. Monitoring critical process parameters and implementing quality control checks at various stages can minimize defects and waste [12, 19].

• Training and Skill Development: Well-trained and skilled personnel are essential for maintaining quality standards and reducing waste. Adequate training programs that emphasize quality awareness, technical skills, and problem-solving capabilities contribute to consistent quality optimization and waste reduction [20].

• Continuous Improvement and Lean Practices: Embracing continuous improvement methodologies, such as Lean Six Sigma, can drive quality optimization and waste reduction. Tools like value stream mapping, root cause analysis, and Kaizen events enable organizations to identify and eliminate process inefficiencies, defects, and non-value-added activities [21].

5. Software tools commonly used in DOE

There are several software tools available that can assist in the design, analysis, and implementation of Design of Experiments (DOE). These tools provide a user-friendly interface and various statistical capabilities to simplify the DOE process. The choice of software tool depends on the specific requirements of the experiment, the level of complexity involved, and the user’s familiarity with the software. Here are some commonly used software tools for DOE:

• Minitab: Minitab is a popular statistical software package widely used for DOE. It offers a comprehensive set of DOE tools, including factorial designs, response surface methods, and mixture designs. Minitab provides easy-to-use graphical and statistical analysis features, making it suitable for both beginners and experienced users.

• JMP: JMP is a powerful statistical software developed by SAS. It offers a range of DOE techniques, such as factorial designs, response surface methods, and mixture designs. JMP provides an interactive interface with drag-and-drop capabilities for designing experiments, analyzing data, and visualizing results.

• Design-Expert: Design-Expert is a specialized software tool specifically designed for DOE. It offers a wide range of experimental design options, including factorial designs, response surface methods, mixture designs, and Taguchi designs. Design-Expert provides advanced graphical and statistical analysis features to facilitate the optimization of processes and product formulations.

• R: R is a popular open-source programming language for statistical computing and graphics. It has a rich collection of packages that support DOE, such as the ‘DOE’ package and ‘rsm’ package. R provides extensive flexibility and customization options for designing experiments, analyzing data, and performing advanced statistical modeling.

• Excel: Microsoft Excel, though not specifically designed for DOE, can be used for simple experimental designs and analysis. It offers basic statistical functions, charts, and data analysis tools that can be utilized for conducting DOE experiments and analyzing results.

6. Some examples of specific fields where DOE found its practical application

6.1 Manufacturing industry

DOE can be used to improve the quality of products, to reduce costs, and to increase efficiency. For example;

• DOE can be used to optimize the process of manufacturing a part, identify the root cause of a quality problem, or reduce the variability of a process, which is a measure of quality. It can be used to identify the causes of defects in a product or to find ways to reduce the time it takes to manufacture a product. DOE can be adopted in the manufacturing industry by an industry that desires to manufacture a machine part (from Al-Si alloy material) with minimum surface roughness by combining three controllable variables (cutting speed, feed rate, and depth of cut). Due to the combinations of many variables, the Design of Experiments (DOE) could be used to carry out the study of the effect of the three machining variables (cutting speed, feed rate, and depth of cut) on the surface roughness (Ra) of Al-Si alloy [22, 23].

• In the automotive industry, DOE is used to improve the fuel efficiency of cars and trucks. By identifying the factors that most affect fuel efficiency, businesses can design cars and trucks that use less fuel. This can lead to lower emissions and a reduced cost of ownership for the customer.

• In the food industry, DOE is used to improve the taste and texture of food products. By understanding how different factors affect the taste and texture of food, businesses can develop products that are more appealing to consumers. This can lead to increased sales and a stronger brand reputation.

• In the pharmaceutical industry, DOE is used to develop new drugs that are more effective and less harmful than existing drugs. By understanding how different factors affect the effectiveness and toxicity of drugs, businesses can develop drugs that are more likely to be approved by the agency in charge. This can lead to increased profits and to a better quality of life for patients.

6.2 Engineering

DOE is used in engineering to improve the design and performance of products and processes. According to [10] in their research presented 77 cases of quality improvement through practical applications of DOE in Engineering. For example;

• DOE can be used to optimize the design of a car engine or to improve the yield of a chemical reaction [10]. DOE can be used to improve the design of products and processes, reduce costs, and increase efficiency. Furthermore, DOE can be used to optimize the design of a bridge, identify the root cause of a failure, or reduce the weight of a product.

• DOE can be used to optimize the design of composite materials for specific structural applications. Factors such as fiber type, fiber volume fraction, resin content, and curing parameters can be varied systematically to achieve desired mechanical properties such as strength, stiffness, and impact resistance [24].

• DOE is used to optimize process parameters to improve yield and quality in semi-conductor engineering. Factors such as temperature, pressure, etching time, and gas flow rates can be varied systematically to identify the optimal settings that result in minimal defects and enhanced performance [25].

6.3 Medicine

DOE is used in medicine for diverse applications. DOE found its applications in medical research and practice, including drug formulation, medical device manufacturing, and radiation therapy. Most of these applications widely demonstrate how DOE can be used to systematically investigate and optimize factors influencing key outcomes in medical applications. Some of the key examples are;

• DOE can be used to optimize drug formulations to improve the bioavailability of pharmaceutical products. Factors such as excipient composition, drug concentration, and manufacturing parameters can be systematically varied to identify the optimal combination that maximizes drug absorption and efficacy [26].

• DOE can be used to optimize the dosage of a drug, to identify the side effects of a treatment, or to reduce the risk of a disease. DOE is used in medicine to improve the effectiveness of treatments and to reduce the side effects of drugs. For example, DOE can be used to identify the best dose of a drug to treat a particular condition or to find ways to reduce the toxicity of a drug.

• DOE can be utilized to optimize treatment protocols in radiation therapy for cancer patients. Factors such as radiation dose, treatment duration, and beam angles can be systematically varied to identify the optimal combination that maximizes tumor control while minimizing side effects on healthy tissues [24, 27].

6.4 Agriculture

Agriculture: DOE is used in agriculture to improve crop yields and to reduce the use of pesticides and fertilizers. DOE can be used to determine the best combination of fertilizer and irrigation rates to maximize crop yields, optimization of plant growth conditions in controlled environments and many more. For example;

• DOE can be used to evaluate the effects of fertilizer formulations on crop yield. In the work of [17], DOE was used to investigate the impact of different fertilizer formulations on crop yield. The study involved varying factors such as nitrogen, phosphorus, and potassium concentrations in the fertilizer mix. By systematically designing and conducting experiments, they were able to determine the optimal combination of nutrients that maximized crop yield while minimizing the amount of fertilizer required.

• DOE can be used to assess the impact of irrigation techniques on water use efficiency. The effect of different irrigation techniques on water use efficiency in crop production was evaluated by [28]. Various factors, such as irrigation frequency, irrigation duration, and water application rate, were manipulated and studied. The experiments allowed the researchers to identify the optimal combination of irrigation practices that resulted in improved water use efficiency without compromising crop yield [28].

• Optimization of Plant growth conditions in controlled environments was conducted by [29] using DOE. The growth conditions of the plant was optimize in controlled environments such as greenhouses or growth chambers. Factors such as light intensity, temperature, humidity, and CO₂ levels were systematically varied to determine the optimal combination that promoted plant growth, development, and yield.

6.5 Marketing

DOE is used in marketing to improve the effectiveness of advertising campaigns and to increase sales. DOE is a powerful tool that can be used to improve the performance of products, processes, and organizations. It is a valuable tool for anyone who wants to improve their results. Here are some specific examples of how DOE has been used in practice:

• DOE can be used to optimize product packaging design elements such as color, shape, size, and labeling to understand their impact on consumer perception and purchase behavior. By systematically varying these factors, marketers can identify the optimal packaging design that maximizes consumer appeal and product sales [30].

• DOE can be employed to evaluate different pricing strategies and their impact on consumer behavior, purchase intent, and profitability. Factors such as price levels, discount offers, and promotional strategies can be systematically varied to determine the optimal pricing strategy that maximizes sales and profitability [31, 32].

• DOE can be utilized to test and optimize various elements of advertisements, such as visual design, headline, copywriting, and call-to-action. By systematically varying these factors and measuring consumer responses, marketers can identify the optimal combination that maximizes advertisement effectiveness and consumer engagement [33].

7. Benefits of DOE in quality control and assurance

• Efficient Resource Utilization: DOE allows organizations to allocate their resources efficiently by identifying the most influential factors. By focusing on these factors, companies can optimize their processes and achieve significant improvements in quality without unnecessary expenditure.

• Cost Reduction: By systematically exploring process factors and their interactions, DOE helps identify cost-effective solutions. By reducing defects, eliminating waste, and optimizing process parameters, organizations can save costs associated with rework, scrap, and material consumption.

• Improved quality: DOE can be used to improve the quality of products and processes by identifying and reducing the variability of the process.

• Reduced costs: DOE can be used to reduce costs by identifying the most efficient way to produce a product or by reducing the amount of waste.

• Increased efficiency: DOE can be used to increase efficiency by identifying the root cause of problems and by improving the design of products and processes.

• Enhanced Decision Making: DOE provides a structured approach to experimentation, resulting in reliable and statistically valid data. This enables informed decision making based on evidence rather than intuition or guesswork. By using DOE, organizations can make data-driven decisions to improve quality and minimize risks.

• Faster Time to Market: DOE facilitates the identification of critical process parameters and optimal levels, leading to faster process optimization. By reducing the time required for experimentation and process development, organizations can accelerate product development cycles and bring products to market more quickly.

8. Challenges and considerations in implementing DOE in real-world scenarios

• Resource Constraints: Limited resources such as time, budget, and availability of equipment or materials can pose challenges in implementing DOE. Conducting experiments may require significant time and financial investments. It is essential to carefully plan and allocate resources to ensure the feasibility and success of DOE studies [19, 34].

• Experimental Constraints: Some experiments may face practical constraints due to factors such as safety regulations, ethical considerations, or limitations in the process or system under investigation. Researchers must identify and address these constraints to design experiments that are feasible and align with regulatory requirements [21, 35].

• Data Analysis Complexities: Analyzing experimental data and interpreting the results can be challenging, particularly when dealing with complex designs or large datasets. Specialized statistical knowledge may be required to properly analyze and draw meaningful conclusions from the data obtained through DOE. Consideration should be given to the appropriate statistical methods and software tools for analyzing the experimental results [35, 36].

• Planning for Interactions and Confounding: Identifying and addressing potential interactions among factors and confounding effects can be complex in DOE. Interactions and confounding can affect the interpretation of experimental results and lead to incorrect conclusions. Careful consideration and appropriate experimental design strategies, such as fractional factorial designs, can help mitigate these challenges [16, 37, 38].

9. Conclusion

In this chapter, the authors focus on the applications and benefits of DOE in quality control and assurance. DOE is a very important statistical methodology that enables both scientists, engineers, researchers, and various other professionals to design, develop, and optimize high-quality products and services. DOE has many applications in various fields. Application of DOE ensures high-quality products and services, increases customer satisfaction, efficient resource utilization, cost reduction, enhanced decision making, improves brand reputation, and ensures overall business success. Having presented in this chapter the numerous applications of DOE in various fields, it can be concluded that DOE is a powerful research tool and methodology for quality control and assurance.

Acknowledgments

The opportunity to have this chapter review included in the main book, Quality Control and Quality Assurance - Techniques and Applications, is greatly appreciated by the authors. We also commend the rigorous peer review process.