Design of a Disinfection and Epidemic Prevention Robot Based on Fuzzy QFD and the ARIZ Algorithm

Abstract

In the present era, against the background of normalized measures of epidemic prevention and control, a multi-function intelligent robot that can completely replace human beings in disinfection and epidemic prevention work needs to be designed in order to contain the spread of viruses at the source, to efficiently and conveniently serve the controlled regions, and to meet the needs of the current context. Existing disinfection robot designs only address the problem of disinfection; there is no disinfection robot that can achieve full-domain and full-space disinfection (including stairs), as well as having multiple epidemic prevention functions. The fuzzy QFD tool and the ARIZ algorithm were used here to solve the problems related to disinfection robots. By using fuzzy QFD as the research framework and combining it with fuzzy AHP to calculate the weight of demand for the disinfection and epidemic prevention robot, the authors further translated the demand into the technical and part characterization of the robot. The conflicts in the fuzzy QFD model were solved using the ARIZ algorithm. Finally, based on the solving principle of the ARIZ algorithm and the qualitative and quantitative conclusions of fuzzy QFD, a robot for disinfection and epidemic prevention was designed in a comprehensive manner. A robot that can completely replace human beings in the tasks of disinfection and epidemic prevention was designed in this work, solving the problems associated with the robot’s ability to climb stairs and realizing the innovation of the robot’s comprehensive disinfection, distribution of materials, real-time monitoring, temperature measurement, and other functions. This study provides a theoretical reference for the design of related products.

1. Introduction

At present, many countries are in the phase of epidemic prevention and control and will remain in such a period for a long time. Under the concept of normalized epidemic prevention and control, questions of how to stop viruses at the source, transport supplies quickly and efficiently, and carry out real-time monitoring, in addition to other problems, need to be comprehensively resolved. It is an important and necessary task to carry out periodic and quantitative disinfection treatment in public areas, transport supplies to high-risk areas, and complete other epidemic prevention work. These tasks are often carried out by human staff; the staff themselves are at risk of infection, so a robot device is required that can replace them in performing disinfection and other related epidemic prevention work. Many scientific research institutions at home and abroad have stepped up in their research on the applications of mechanized devices for disinfection measures against the novel coronavirus, aiming at containing this virus’s spread within the sources of infection. The headquarters of Hangzhou Xiaoshan District Zhongkai Robot Co., Ltd. (Hangzhou, China) has developed a spraying disinfection robot named “Thor I”. This robot can complete approximately 2500 square meters of disinfection work in 15 min as a scientific and technological tool supporting the “anti-epidemic” team in Wuhan, replacing the manual epidemic prevention and disinfection work. The Danish Company Blue Ocean Robotics has developed a UVD disinfection robot that uses ultraviolet light (UV-C) to kill harmful microorganisms. At present, more than 40 countries and regions are using this robot, including those in Asia, Europe, and the United States. This robot can be used in various enclosed spaces, such as offices, large shopping malls, schools, airports, and production factories, as well as other environments. In the current research on disinfection and epidemic prevention robots, most of the works have studied disinfection methods, such as the method of spraying disinfectants, ultraviolet disinfection, etc. However, there is no research or discussion on how to achieve the comprehensive and full-domain disinfection of indoor and outdoor spaces. The problems of disinfection in stairways and other indoor public areas with various obstacles have not been resolved. Moreover, these robots are relatively simple. In addition to not being able to finish the disinfection work, they cannot complete other important and necessary epidemic prevention tasks, such as material distribution, temperature measurement, and real-time monitoring.

The remaining sections of this paper are organized as follows. In Section 2, the authors briefly review the literature on related methods, the shortcomings of past studies, and areas to improve in this field of research. In Section 3, the authors present a research framework for the design of a disinfection and epidemic prevention robot. In Section 4, the structure and effects of the design of the disinfection and epidemic prevention robot are shown and the research method used in the design of the disinfection and epidemic prevention robot is discussed. In Section 5, the conclusions are drawn, summarizing the innovative points of this paper and providing some suggestions for future work to expand on this study.

2. Related Work

Quality Function Deployment (QFD) is a type of planning tool that transforms customer demand into product quality assurance at the production stage through qualitative and quantitative analyses. As a research framework, it is often used in combination with other design methods. Many scholars have combined QFD with other relevant decision evaluation methods and problem-solving methods [1,2]. For example, Sireli et al. [3] imported the Kano model into QFD to accurately comprehend users’ needs and carry out the design of many products. Chen et al. [4] integrated the grey relational approach of Quality Function Deployment (QFD) and Quality Engineering (QE) to solve problems related to product design. Hartono et al. [5] incorporated the Kano model and QFD into KE for luxury hotel services. Wudhikarn et al. [6] proposed a decision-making method combining the Analytical Network Process (ANP) and Quality Function Deployment (QFD) to improve the objectivity and scientificity of decision making, as well as to identify the intuitive or illogical flaws in decision making. Murali et al. [7] combined QFD with theoretical approaches such as SWOT, BSC, and AHP; a comprehensive model was developed, and the efficiency of the model was verified using the After-Sales Service (ASS) business of a home appliance company as a research case. To summarize, some achievements have been made in integrating the calculation methods of different users’ demand weights into the development process of QFD. Traditional QFD decision making is based on fuzzy or unclear judgment and evaluation, and the input of the variables is often assumed to have a fixed value, which ignores the uncertainty of subjective evaluation, judgment criteria, and value requirements due to individual differences in perception. In order to solve this deficiency, many scholars have combined QFD with fuzzy mathematics to make decision making more objective [8,9,10]. Vinodh et al. [11] applied the fuzzy QFD framework to the sustainable design of consumer electronic products. Kang et al. [12] took the design of a mini car as an example to verify the feasibility of this method based on fuzzy QFD and EGM. Li et al. [13] used fuzzy QFD to determine the importance weights of ECs according to the context of CRs in an open design. Gundogdu et al. [14] proposed a novel fuzzy QFD method and described its application to the development of linear delta robot technology. Wang et al. [15] integrated fuzzy QFD and a grey decision-making approach for the supply chain collaborative quality design of large complex products. Lin et al. [16] used the fuzzy QFD model to solve the problems related to environmental production requirements (EPR) and sustainable production indicators (SPI), combining fuzzy set theory and analytical network processes to propose a systematic analytical procedure. The ARIZ algorithm, as the major component of the TRIZ theory, integrates several tools of TRIZ and considers the whole process, from problem analysis to solution; it is used by many scholars to solve engineering problems. Fey et al. [17] used the ARIZ algorithm to find solutions to real-life engineering problems. Krasnoslobodtsev et al. [18], based on the ARIZ algorithm, developed a robot that can be used to clean and organize any surface in space. Zhang et al. [19] applied QFD, TRIZ, and other theories to the innovative design of ergonomic products, namely kitchen stove products.

However, the need to solve practical technical problems while also reducing subjectivity and bias in decision making has not been adequately considered in past studies. In this paper, fuzzy mathematical theory is combined with the AHP and QFD methods. Fuzzy AHP is used to calculate the social demand weight values of disinfection and epidemic prevention robots; fuzzy QFD is used to establish the mapping model between the social demand of robots and technical characteristics and part characteristics, and to derive the contradictions and conflicts between technical indicators; and the ARIZ algorithm is used to solve the related contradictions and conflicts.

3. Research Framework

3.1. Design Process

Through the use of questionnaire surveys and in-depth interviews with relevant experts, the authors collected the current society’s requirements for disinfection and epidemic prevention robots and then selected, decomposed, combined, and built the requirement level model [20]. In contrast with the traditional AHP analytic hierarchy process, by adopting the fuzzy AHP method, one can optimize the complicated consistency test and maintain the primacy of data [21,22]. The fuzzy AHP tool has better accuracy and objectivity in complex evaluation systems to calculate the demand weights of disinfection and epidemic prevention robots in all aspects. Then, the robots’ demand weight is imported into the fuzzy QFD house of quality model. The triangular fuzzy value is used to judge the correlation matrix and the robots’ requirements are transformed into the robots’ technical characteristics; thus, the conflicts are obtained. The contradictions and conflicts in the technical characteristics need to be solved by the ARIZ algorithm, thus finding the ideal solution principle. Finally, fuzzy QFD is used to transform the technical characteristics of the robots into the characteristics of the parts. According to the importance of the characteristics of the robot’s parts and the solution principle of solving conflicts, a disinfection and epidemic prevention robot that can meet the requirements of the present context is designed, as shown in Figure 1.

3.2. Analysis of Infection and Epidemic Prevention Robot’s Requirements

The authors surveyed industrial design students and teachers and conducted interviews with experts in the field, focusing on the following aspects: 1. the type of machine or equipment that is needed to assist in the prevention and control of epidemics in the current epidemic period; 2. the functions that the machine or equipment should have in order to meet the needs of today’s society; and 3. the aspects that should be taken into account when designing such a device. Several demand indicators for disinfection and epidemic prevention robots were derived, which included disinfection treatment, fast forward movement, movement up and down stairs, distribution of supplies, temperature measurement, real-time monitoring, voice communication, high automation, easy operation, smooth running, stable center of gravity, disinfection safety, low failure rate, structural stability, material safety, good color, good shape, and several other indicators. The authors summarized, organized, classified, and expanded on the needs for the disinfection and epidemic prevention robot’s aspects in terms of the current requirements, which were divided into three levels: target level, criterion level, and index level. The second criterion level divides the social demands of disinfection and epidemic prevention robots into the main function A1, auxiliary function A2, man–machine performance A3, safety A4, and appearance A5. The third level divides the second level into more specific demands, as shown in Figure 2. After defining the disinfection and epidemic prevention robot’s demand levels, the authors conducted the following fuzzy matrix analysis.

After establishing the demand level of the disinfection and epidemic prevention robot, fuzzy AHP is used to construct the judgment matrix, which can process multi-objective and complex problems hierarchically, and we further determine the weight of social demand through the fuzzy consistency test so as to reduce the decision deviation. The calculation steps are as follows.

Step 1: The fuzzy analytic hierarchy process usually adopts the 0.1~0.9 point scale method; then, each expert is invited to construct the fuzzy judgment matrix by comparing the importance of various demand factors of the disinfection and epidemic prevention robot, $A = \left(a_{ij}\right)n\times n$. The scoring rules are shown in

Table 1. The 0.1~0.9 scale method.

Scale	Implication
0.5	By comparison, the two elements are equally important.
0.6	By comparison, one element is slightly more important than the other.
0.7	By comparison, one element is obviously more important than the other.
0.8	By comparison, one element is strongly more important than the other.
0.9	By comparison, one element is extremely more important than the other.
0.1, 0.2, 0.3, 0.4 (complementary)	If the element $r_{ij}$ is obtained by comparing the element $a_i$ with the element $a_j$, the element $r_{ji}=1-r_{ij}$ is obtained by comparing the element $a_j$ with element $a_i$.

.

Step 2: We construct the fuzzy judgment matrix. The fuzzy judgment matrix is denoted as follows:

$$A = \left[\begin{array}{cccc}{a}_{11}& a_{12}& \cdots & a_{1n}\\ a_{21}& a_{22}& \cdots & a_{2\mathrm{n}}\\ ⋮& ⋮& \ddots & ⋮\\ a_{n1}& a_{n2}& \cdots & a_{nn}\end{array}\right]$$

Here, $0 \le r_{ij} \le 1$

$$a_{ii} = 0.5, i = 1,2,3\cdots n,n\in N^{*}$$

$$a_{ij} + a_{ji} = 1, j,i = 1,2,3\cdots n,n\in N^{*}$$

Step 3: We perform the fuzzy consistency test. In order to verify the scientificity and rationality of the constructed fuzzy judgment matrix, the fuzzy consistency text needs to be carried out. The test procedure is shown in Formula (1).

$$r_{11} + r_{kk} = r_{1k} + r_{k1}$$

$$r_{11} - r_{k1} = r_{1k} - r_{kk} = b \left(b \mathrm{is} \mathrm{constant}\right)$$

$$r_{12} - r_{k2} = b\left(b \mathrm{is} \mathrm{constant}\right),\hspace{1em}{r}_{13} - r_{k3} = b,\dots r_{1j} - r_{kj} = b,\cdots \left(j = 2,3\cdots n,n\in N^{*},j\ne k\right)$$

(1)

Step 4: To calculate the weight of the criterion level and the weight of each index level, we use Formula (2).

$$w_i = \frac{{\sum }_{k=1}^n{a}_{ik}+\frac{n}{2}-1}{n\left(n-1\right)}$$

(2)

According to the above steps, the weight of each demand of the disinfection and epidemic prevention robot is calculated one by one, and the fuzzy consistency test is carried out. Firstly, the fuzzy judgment matrix of the disinfection and epidemic prevention robot’s target level A is constructed and the scoring and calculation results are as follows:

$$A = \left[\begin{array}{lllll}0.50\hfill & 0.85\hfill & 0.75\hfill & 0.65\hfill & 0.85\hfill \\ 0.15\hfill & 0.50\hfill & 0.40\hfill & 0.30\hfill & 0.50\hfill \\ 0.25\hfill & 0.60\hfill & 0.50\hfill & 0.40\hfill & 0.60\hfill \\ 0.35\hfill & 0.70\hfill & 0.60\hfill & 0.50\hfill & 0.70\hfill \\ 0.15\hfill & 0.50\hfill & 0.40\hfill & 0.30\hfill & 0.50\hfill \end{array}\right]w = \left[\begin{array}{l}0.255\hfill \\ 0.167\hfill \\ 0.192\hfill \\ 0.217\hfill \\ 0.167\hfill \end{array}\right]$$

Then, the fuzzy judgment matrix of each criterion level is constructed, and the scoring and calculation results of main function $A_1$ of the fuzzy judgment matrix are as follows:

$$A_1 = \left[\begin{array}{llll}0.50\hfill & 0.75\hfill & 0.60\hfill & 0.60\hfill \\ 0.25\hfill & 0.50\hfill & 0.35\hfill & 0.35\hfill \\ 0.40\hfill & 0.65\hfill & 0.50\hfill & 0.50\hfill \\ 0.40\hfill & 0.65\hfill & 0.50\hfill & 0.50\hfill \end{array}\right]w_1 = \left[\begin{array}{l}0.287\hfill \\ 0.204\hfill \\ 0.254\hfill \\ 0.254\hfill \end{array}\right]$$

The scoring and calculation results of auxiliary function $A_2$ of the fuzzy judgment matrix are as follows:

$$A_2 = \left[\begin{array}{lll}0.50\hfill & 0.45\hfill & 0.50\hfill \\ 0.55\hfill & 0.50\hfill & 0.55\hfill \\ 0.50\hfill & 0.45\hfill & 0.50\hfill \end{array}\right] w_2 = \left[\begin{array}{l}0.325\hfill \\ 0.350\hfill \\ 0.325\hfill \end{array}\right]$$

The scoring and calculation results of man–machine performance $A_3$ in the fuzzy judgment are as follows:

$$A_3 = \left[\begin{array}{lll}0.50\hfill & 0.60\hfill & 0.60\hfill \\ 0.40\hfill & 0.50\hfill & 0.50\hfill \\ 0.40\hfill & 0.50\hfill & 0.50\hfill \end{array}\right]w_3 = \left[\begin{array}{l}0.367\hfill \\ 0.317\hfill \\ 0.317\hfill \end{array}\right]$$

The scoring and calculation results of safety $A_4$ in the fuzzy judgment matrix are as follows:

$$A_4 = \left[\begin{array}{llll}0.50\hfill & 0.80\hfill & 0.60\hfill & 0.65\hfill \\ 0.20\hfill & 0.50\hfill & 0.30\hfill & 0.35\hfill \\ 0.40\hfill & 0.70\hfill & 0.50\hfill & 0.55\hfill \\ 0.35\hfill & 0.65\hfill & 0.45\hfill & 0.50\hfill \end{array}\right]w_4 = \left[\begin{array}{l}0.296\hfill \\ 0.196\hfill \\ 0.263\hfill \\ 0.246\hfill \end{array}\right]$$

The scoring and calculation results of appearance $A_5$ in the fuzzy judgment matrix are as follows:

$$A_5 = \left[\begin{array}{lll}0.50\hfill & 0.65\hfill & 0.45\hfill \\ 0.35\hfill & 0.50\hfill & 0.30\hfill \\ 0.55\hfill & 0.70\hfill & 0.50\hfill \end{array}\right] w_5 = \left[\begin{array}{l}0.350\hfill \\ 0.275\hfill \\ 0.375\hfill \end{array}\right]$$

Based on Formulas (1) and (2), the weight of each demand of the disinfection and epidemic prevention robot and each level’s demand are determined, and the fuzzy consistency test is carried out. According to the calculation, all the items accord with the fuzzy consistency test. Finally, the comprehensive demand weight of each index is calculated, which is shown in

Table 2. Comprehensive demand weight.

Objective Level	Criterion Level	Weight	Index Level		Weight (w1–w5)	Comprehensive Weight	Order
Disinfection and anti-epidemic robot	Main function (A1)	0.255	Disinfection	A11	0.287	0.073	1
			Fast forward movement	A12	0.204	0.052	15
			Upstairs movement	A13	0.254	0.065	3
			Material distribution	A14	0.254	0.065	4
	Auxiliary function (A2)	0.167	Temperature measurement	A21	0.325	0.054	12
			Real-time monitoring	A22	0.350	0.059	9
			Voice communication	A23	0.325	0.054	13
	Man–machine performance (A3)	0.192	High automation	A31	0.367	0.071	2
			Good operation	A32	0.317	0.061	7
			Smooth running	A33	0.317	0.061	8
	Safety (A4)	0.217	Stable center of gravity	A41	0.296	0.064	5
			Safe disinfection	A42	0.196	0.043	17
			Low failure rate	A43	0.263	0.057	11
			Sound construction	A44	0.246	0.053	14
	Appearance (A5)	0.167	Texture	A51	0.350	0.059	10
			Color	A52	0.275	0.046	16
			Modeling	A53	0.375	0.063	6

.

3.3. Demand Transformation of the Disinfection and Epidemic Prevention Robot

By using the fuzzy QFD tool, the weight of robot demand obtained by fuzzy AHP is transformed into the corresponding robot’s technical characteristics. A direct mapping relationship between the present demand for the robot and the technical demand for the robot is established, ensuring that the design and development process is highly consistent with the needs of the current context. The core of fuzzy QFD is the House of Quality (HOQ) model [23]. HOQ establishes a mapping model between each independent index through an intuitive presentation, and, in the fuzzy QFD framework of this paper, the same HOQ model is established as the core; its construction process is shown in Figure 3.

Based on the demand level constructed using fuzzy AHP, the needs of the present context for the disinfection and epidemic prevention robot and the corresponding technical characteristics are developed in a one-to-one correspondence, as shown in

Table 3. The disinfection and epidemic prevention robot’s present needs and corresponding technical indicators.

First-Grade Demand	Second-Grade Demand	Corresponding Technical Specifications	First-Grade Demand	Second-Grade Demand	Corresponding Technical Specifications
Main function	Disinfection	Disinfection device	Man–machine performance	High automation	Remote control
		Automatic switchover of tools			Automatic switchover of tools
		Movement up and down stairs		Good operation	Simple operation
					Clear feedback
	Fast forward movement	Steady moving		Smooth running	Simple operation
					Stable construction
		Operation speed	Safety	Stable center of gravity	Stable construction
					Steady moving
	Upstairs movement	Steady moving		Safe disinfection	Disinfection device
					Steady moving
		Stable construction		Low failure rate	Stable construction
		Movement up and down stairs		Sound construction	Stable construction
Auxiliary function	Temperature measurement	Automatic switchover of tools	Appearance	Textural quality	Reasonable materials
		Infrared thermometry		Color	Coordinative colors
					Marked night operation
	Real-time monitoring	Automatic switchover of tools		Beautiful appearance	Size of the whole machine
					Beautiful appearance

.

Based on the summary of Table 3, the disinfection and epidemic prevention robot is divided into seven parts: the power part, the body part, the disinfection part, the artificial intelligence part, the remote sensing part, the whole machine, and the human–machine interface. Among them, the motion stability, running speed, and up- and downstairs movement comprise the robot’s power part; all-around disinfection comprises the robot’s disinfection part; the sensor recognition, on-site video monitor, voice synthesis, and recognition technology comprise the artificial intelligence part; the remote control and the tools’ automatic switching represent the remote sensing part; color coordination, reasonable materials, beautiful shape, marked night operation, stable structure, and the reasonable size of the whole machine represent the whole robot; and simple operation and clear feedback represent the man–machine interface of the robot, as shown in Figure 4.

The weight of robot demand is introduced to the left side of the House of Quality, and each technical feature of the robot is introduced to the ceiling of the House of Quality. The correlation degree between robot users’ demand and technical characteristics is determined by the relevant symbols. According to the strength of the correlation, the corresponding triangular fuzzy number is converted to construct the fuzzy judgment matrix of the House of Quality [24]. The triangular fuzzy number is a type of fuzzy number, and its basic form is $A = \left(a_1,b_1,c_1\right)$. Here, $a \le b \le c$. $a$ and $c$ are the left and right boundaries, respectively [25]. $\left(c-a\right)$ reflects its fuzzy degree; if $\left(c-a\right)$ is larger, its fuzzy degree will be stronger. The symbols of correlation degree judgment and the triangular fuzzy numbers after conversion are shown in

Table 4. The symbol of correlation degree in fuzzy QFD and its corresponding triangular fuzzy number.

Relevant Symbol	Express Meaning	Fuzzy Number After Conversion
Blank	No correlation	(0, 0, 0)
△	Weak correlation	(1, 1, 3)
○	Relatively weak correlation	(1, 3, 5)
◎	Medium correlation	(3, 5, 7)
●	Relatively strong correlation	(5, 7, 9)
▲	Strong correlation	(7, 9, 9)

.

Suppose that there are two arbitrary triangular fuzzy numbers $A = \left(a_1,b_1,c_1\right)$ and $B = \left(a_2,b_2,c_2\right)$, the operational rules are follows:

$$A\oplus B = \left(a_1+a_2,b_1+b_2,c_1+c_2\right)$$

$$A\otimes B = \left(a_1\cdot a_2,b_1\cdot b_2,c_1\cdot c_2\right)$$

(3)

$${\left(A\right)}^{-1} = \left(\frac{1}{c},\frac{1}{b},\frac{1}{a}\right)$$

In the weight calculation of the technical characteristics of the disinfection and epidemic prevention robot, according to the correlation degree between users’ demand weight and the index of technical characteristics, a symbol is inserted to calculate the correlation degree scoring, thus constructing the correlation degree matrix—namely, the house body of HOQ is constructed. According to Formula 4 below, the absolute weight and relative weight of the disinfection and epidemic prevention robot’s technical characteristics are calculated, which represent the absolute weight of the technical characteristics. $H_j$ is the absolute weight of the technical characteristics; $W_i$ is the i social demand weight (calculation result from fuzzy AHP); $P_{ij}$ is the correlation coefficient; and $h_j$ is the relative weight of technical demand.

$$\begin{gathered} H_j = \sum _{i=1}^q{W}_i\otimes P_{ij} \\ {h}_j = \frac{H_j}{{\sum }_{i=1}^q{H}_j} \end{gathered}$$

(4)

Then, the fuzzy weight value is processed with defuzzification by adopting the center of gravity method to solve fuzzing. Suppose that $H_j = \left(a,b,c\right)$, its defuzzification is as follows:

$$H_j = \frac{a+b+c}{3}$$

(5)

The demand weight of the disinfection and epidemic prevention robot is translated into the weight of technical characteristics, which constitutes the basement of the House of Quality (HOQ). Moreover, the positive and negative correlation degrees of each technical index of the disinfection and epidemic prevention robot between pairs are analyzed and marked on the House of Quality (HOQ)’s roof. A positive correlation is denoted by “+”, while a negative correlation is denoted by “$-$”. Therefore, the conflicts among the technical characteristics can be obtained [26,27]. The disinfection and epidemic prevention robot’s House of Quality is shown in Figure 5.

3.4. Conflict Resolution of the Disinfection and Epidemic Prevention Robot

The ARIZ algorithm, as an important method used to solve engineering problems, can be perfectly combined with QFD. By observing the roof of the House of Quality, the positive and negative correlations between the technical indicators of the disinfection and epidemic prevention robot can be determined intuitively. “+” indicates a positive correlation; an improvement in one technology will lead to the optimization of another technology. “$-$” indicates a negative correlation; an improvement in one technology will lead to the deterioration of another technology. The blank sections indicate that there is no correlation among technologies, as shown in Figure 6.

It can be seen from Figure 6 that there are five pairs of major contradictions in the various technical indicators of the disinfection and epidemic prevention robot. These are as follows: ① Steady movement and movement up and down stairs. It is required that the disinfection and epidemic prevention robot should move horizontally when it is in the horizontal plane. This requires movement with the combination of traveling up and down and moving horizontally when the robot enters the stairs. Importing this problem into the ARIZ algorithm process, the technical contradiction cannot resolve this conflict. This is a pair of physical contradictions under different conditions, and we need to adopt the principle of conditional analysis to solve the robot’s running state under different conditions. ② Running speed and movement up and down stairs. It is required that the robot’s chassis tire maintains the speed when moving up and down, as well as when moving forward and backward; at the same time, it is necessary to keep the robot’s overall center of gravity balanced. Importing this problem into the ARIZ algorithm process, neither technical nor physical contradictions can resolve this conflict. This is a complex invention problem that requires Su-Field Models to find the optimal solution. ③ Moving forward steadily and all-around disinfection. Importing this problem into the ARIZ algorithm process, this is a pair of technical contradictions. The disinfection and epidemic prevention robot should both move steadily and carry out the disinfection work in each corner of the space, without any missed spaces. ④ All-around disinfection and beautiful shape. The complexity of the disinfection structure will inevitably reduce the formal aesthetic feeling of the robot. Importing this problem into the ARIZ algorithm process, the technical contradiction cannot resolve this conflict. This is a pair of physical contradictions. The principle of space separation is adopted so as to allow the disinfection device to perform well in a space; this can not only ensure the flexibility of the disinfection structure but can also take the aesthetic feeling of form into account. ⑤ Infrared temperature measurement and on-site video monitoring. Importing this problem into the ARIZ algorithm process, this is a pair of technical contradictions. As there are two important sensor components of the robot, it is necessary to ensure that the two components are independent of each other and adapt to each other, but also to ensure the operational efficiency of the two components, as shown in

Table 5. The symbols of the correlation degree in fuzzy QFD and its corresponding triangular fuzzy numbers.

Conflict	Conflict Type	Improvement Parameters	Worsening Parameters	Recommended Invention Principle
① Steady movement—movement up and down stairs	Physical conflict	Principle of conditional separation		3, 17, 19, 31, 32, 40
② Running speed—movement up and down stairs	Complex invention problem	Construct Su-Field models		76 standard solutions
③ Steady moving—all-around disinfection	Technical conflict	Stability of power device	Adaptability and versatility of disinfection devices	35, 30, 34, 2
④ All-around disinfection—beautiful shape	Physical conflict	Principle of spatial isolation		1, 2, 3, 7, 4, 17
⑤ Infrared thermometry—on-site video monitoring	Technical conflict	Adaptability and versatility of sensing devices	Operating efficiency of sensing devices	35, 28, 6, 37

.

According to the contradiction analysis presented in Table 5 and the corresponding invention principle, and in accordance with the technical knowledge and actual situation, the suitable invention principle is selected to solve the contradictions and conflicts among the technical indicators of the disinfection and epidemic prevention robot.

In No. ①, the contradiction of movement forward steadily and moving up and down stairs, principle No. 17, i.e., the multidimensional operation principle, is adopted to design the robot with the combination of horizontal two-dimensional operation and vertical three-dimensional operation, as shown in Table 6.

Table 6. Resolution of No. ① conflict.

Conflict	Solution Principle	Solution Plan
Steady movement—movement up and down the stairs	No. 17 multidimensional operation principle	See Figure 7

Table 6. Resolution of No. ① conflict.

Conflict	Solution Principle	Solution Plan
Steady movement—movement up and down the stairs	No. 17 multidimensional operation principle	See Figure 7

In the running speed and movement up and down stairs, which constitute conflict No. ②, the material-field model is constructed, as shown in Table 7. In the system of the disinfection and epidemic prevention robot, when climbing stairs, two pairs of different substances, S1 and S2, are composed of the robot’s two pairs of tire support and the stairs. The motor that controls the rise and fall of the tire support is the mechanical field F1. In the process of the disinfection and epidemic prevention robot climbing stairs, the machine’s body will lose its center of gravity when the mechanical field F1 drives the tire support S1, resulting in the failure of S1 to act on the stairs S2. The standard solution 1.2.1 should be adopted, where, in the presence of both useful and harmful effects in a system, S3 is introduced to counteract the harmful effects. Two new sets of tire support need to be introduced to offset the effect of weight instability. However, during the process of the disinfection and epidemic prevention robot’s forward movement, S3 lacks the mechanical field to drive the rise and fall, meaning that the robot cannot climb over the steps. Therefore, standard solution 1.1.7 needs to be adopted. In a system with poor controllability, there are some situations that cannot be changed, and a second field can be connected. The second mechanical field F2 (rise and fall motor) is introduced to carry out the rise and fall of S3 (tire support), which carries out the alternative rise and fall of S1 (tire support) and acts on S2 (stairs), thus achieving the effect of the robot climbing stairs.

Table 7. Resolution principle of No. ② conflict.

Conflict	Resolution Principle	Resolution Plan
Running speed—movement up and down stairs	Standard solution1.2.1 Standard solution 2.1.1	See Figure 8, Figure 9 and Figure 10

Table 7. Resolution principle of No. ② conflict.

Conflict	Resolution Principle	Resolution Plan
Running speed—movement up and down stairs	Standard solution1.2.1 Standard solution 2.1.1	See Figure 8, Figure 9 and Figure 10

In the steady forward movement and all-around disinfection of conflict No. ③, principle No. 35, i.e., the performance conversion principle, is adopted, designing the traditional disinfection spraying device in the mechanical arm with more flexible operation and more variable directions. Due to the stability of the disinfection and epidemic prevention robot’s movement, many narrow spaces with obstacles cannot be touched. By changing the flexibility degree of traditional disinfection and epidemic prevention devices, performance optimization and change are realized, thus ensuring both the disinfection and epidemic prevention robot’s stable movement and the completion of indoor and outdoor disinfection tasks without any missed spaces, spanning 360 degrees, as shown in Table 8.

Table 8. Resolution principle of No. ③ conflict.

Conflict	Recommended Invention Principle	Resolution
Steady movement—all-around disinfection	No. 35 performance conversion principle	See Figure 11

Table 8. Resolution principle of No. ③ conflict.

Conflict	Recommended Invention Principle	Resolution
Steady movement—all-around disinfection	No. 35 performance conversion principle	See Figure 11

In the all-around disinfection and aesthetic appearance of conflict No. ④, principle No. 2, i.e., the embedding principle, is adopted, as shown in Table 9. The complex structure of the mechanical arm will inevitably bring about a reduction in the shape’s aesthetic feeling, which requires the disinfection and epidemic prevention robot to not only ensure all-around disinfection, but also to strengthen the shape’s sense of simplicity and aesthetic feeling. By using the embedding principle, the complex mechanical arm is embedded inside the protection cover with simple modeling, which can not only protect the complex mechanical structure but can also echo the overall shape and increase the aesthetic feeling of the form.

Table 9. Resolution principle of No. ④ conflict.

Conflict	Recommended Invention Principle	Resolution
All-around disinfection—beautiful shape	No. 2 embedded principle	See Figure 12

Table 9. Resolution principle of No. ④ conflict.

Conflict	Recommended Invention Principle	Resolution
All-around disinfection—beautiful shape	No. 2 embedded principle	See Figure 12

In the infrared temperature measurement and on-site video monitoring of conflict No. ⑤ conflict, principle No. 6, i.e., the versatility principle, is adopted, as shown in Table 10. Two sensors for infrared temperature measurement and on-site video monitoring are integrated into the same part in order to increase the versatility of the sensor and improve its working efficiency.

Table 10. Resolution principle of No. ⑤ conflict.

Conflict	Recommended Invention Principle	Resolution
Infrared thermometry—on-site video monitoring	No. 6 versatility principle	See Figure 13

Table 10. Resolution principle of No. ⑤ conflict.

Conflict	Recommended Invention Principle	Resolution
Infrared thermometry—on-site video monitoring	No. 6 versatility principle	See Figure 13

3.5. Transformation of Technological Characteristics of the Disinfection and Epidemic Prevention Robot

In order to carry out the design successfully, it is necessary to define the component characteristics and the component characteristics’ weights in the disinfection and epidemic prevention robot. The fuzzy QFD tool needs to be used again to transform the technical feature weights of the disinfection and epidemic prevention robot into the component feature weights. Firstly, based on the solving principle determined by the ARIZ algorithm, the parts of the disinfection and epidemic prevention robot are further classified. The disinfection and epidemic prevention robot is divided into seven parts: the power part, the body part, the disinfection part, the artificial intelligence part, the remote sensing part, the whole machine, and the man–machine interface, as shown in

Table 11. Expansion of the disinfection and epidemic prevention robot’s part characteristics.

Category of Components	Index of Components
Power part	Chassis and motor construction
	Tire
	Obstacle recognizer
Body part	Machine casing
	Bearing compartment
	Protective cover for disinfection
Disinfection part	Mechanical arm for disinfection
	Storage tank for disinfectant
Artificial intelligence part	Infrared thermometer
	Thermal detector
	Voice recognition and player device
	Feedback screen
Remote sensing part	Receiver and transmitter
Whole machine	Size of the whole machine
	Floodlight
	Decorative lamp
Man–machine interface	Display screen

. The power part includes the chassis and motor structure, tires, and obstacle recognition. The machine part includes the outer shell of the machine, bearing cabin, bumper, and protective cover for disinfection. The disinfection part includes the mechanical arm for disinfection and the disinfectant storage tank. The artificial intelligence part includes the infrared thermometer, heat detector, voice recognition and player device, and feedback screen. The remote sensing part includes the receiver and transmitter. The whole machine includes the machine size, floodlight, and decorative lamp. The man–machine interface is the display screen.

According to the established component indicators, they are imported into the roof of the House of Quality. The weight values of the technical characteristics shown in Figure 6 are imported into the left wall of the House of Quality. The association matrix between the robot’s technical characteristics and the component characteristics is established by using the fuzzy trigonometric number again, and the mapping relationship between the technical characteristics and the component characteristics is constructed. The weight value of the technical characteristics is converted into the weight value of the component characteristics of the disinfection and epidemic prevention robot, as shown in Figure 14.

According to the relative weights of parts and the components’ House of Quality, it can be seen that the chassis motor structure (11.71), the disinfection mechanical arm (9.84), and the outer shell of the machine (8.74) have the highest weight values, indicating that they are the most important in the design and development of a disinfection and epidemic prevention robot. The heat detector (8.18), body temperature and warning feedback screen (7.94), remote receiver and transmitter (7.91), and tires (7.35) are relatively important in the design and development of a disinfection and epidemic prevention robot. The obstacle recognition device (7.34), size of the whole machine (5.18), display screen (4.93), and voice recognition and play device (4.77) are generally important in the design and development of a disinfection and epidemic prevention robot. The infrared thermometer (3.66), disinfection protective cover (3.43), and load-bearing cabin (3.19) are the second most important in the design and development of a disinfection and epidemic prevention robot. The lighting lamp (2.74), disinfectant storage compartment (2.32), and decorative lamp (0.76) are the least important, and they do not need to be paid much attention in the design and development of a disinfection and epidemic prevention robot.

4. Design Practice and Discussion

4.1. Design Practice

According to the abovementioned order of importance of components and the ARIZ problem-solving principle, the function, structure, and modeling of the disinfection and epidemic prevention robot are innovatively obtained. Firstly, the dynamic structure of the disinfection and epidemic prevention robot is designed so as to realize the functional innovation of the disinfection and epidemic prevention robot’s ability to independently climb stairs, which is shown in Figure 15. Here, the icons ①–④ denote the U-shaped supports from No. 1 to No 4. The crossover rise and fall of the four U-shaped supports are used to achieve the stair-climbing movement. Moreover, ⑤ is the machine’s frame, which is used to fix the U-shaped support and the motor and also to connect the outer shell of the robot; ⑥ is the power supply device that is used to provide electric energy for the motor; ⑦ is the motor device used to drive the U-shaped support up and down; ⑧ is the tire; and ⑨ is the obstacle recognition device. In order to realize the function of automatic stair climbing, it is used to detect the steps of the stairs and determine whether the feet of the supporting device are close to the front steps (ascending stairs) or the edge of the steps (descending stairs). Among them, U-shaped support No. 2 has no motor drive and always maintains a fixed state with the machine’s frame. The rise and fall of U-shaped support No. 2 will always lead to the overall rise and fall of the whole machine.

A climbing simulation is performed to determine the power of the disinfection and epidemic prevention robot. Firstly, when the obstacle identification device of U-shaped support No. 2 identifies the obstacle of stairs, a rising and falling process will act on support No. 1 to adapt to the height of the stairs; when U-shaped support No. 1 stays at the steps on the first floor, the chassis will rise to the height of U-shaped support No. 1 together with U-shaped support No. 2 simultaneously and move forward; when the obstacle identification device of U-shaped support No. 1 identifies the steps of the second floor, it will rise together with U-shaped support No. 3 simultaneously and move forward; when the obstacle identification device of U-shaped support No. 2 identifies the steps of the second floor, it will rise together with the chassis and U-shaped support No. 4 simultaneously; the horizontal height of U-shaped support No. 2 will remain the same as the horizontal height of U-shaped support No. 1; the horizontal height of U-shaped support No. 4 will remain the same as the horizontal height of U-shaped support No. 3 and move forward; and this process will be repeated successively. U-shaped supports No. 3 and No. 4 will rise simultaneously and move forward. U-shaped support No. 2, the chassis, and U-shaped support No. 4 will rise simultaneously and move forward. Instead, movement downstairs requires the opposite process and a similar operation will be performed as well, as shown in Figure 16.

The disinfection and epidemic prevention robot is designed and rationally arranged according to the feature weights of the parts, as shown in Figure 17. Here, 1 is an infrared thermometer, which measures the temperature of passersby using infrared light; 2 is a thermal energy detector, which is used to capture thermal life activity so as to monitor and restrict people in the closed area from moving around and prevent cross-infection; 3 is the decorative lamp; and 4 is the temperature and warning feedback screen, which is used to feed back the detected temperatures of passersby and provide warning feedback to the activity personnel present in non-permitted activity areas. Furthermore, 5 is the control screen, which is used to control the robot and set tasks; 6 is the disinfection mechanical arm, which is used for the all-around disinfection of indoor and outdoor spaces; and 7 is the lighting lamp used to remind passersby who are driving at night and prevent accidents so as to improve safety. Meanwhile, 8 is the connector of the control screen; 9 is a voice recognition and play device, which is used to recognize human language and translate the data into speech to realize interaction with people; and 10 is the remote receiver and transmitter, used to send real-time data to the backstage and receive data instructions from the backstage. Moreover, 11 is the disinfectant storage tank used to store disinfectant; 12 is the disinfection protection cover that is composed of flexible folding material. It is used to protect the structure of the disinfection arm and beautify the appearance when the machine stops the disinfection work; 13 is the load-bearing cabin used for loading materials; and 14 is the outer shell.

Finally, the overall shape of the disinfection and epidemic prevention robot is designed. The software programs Rhino and Keyshot are used to build and render the disinfection and epidemic prevention robot, thus forming an image of the final effect, as shown in Figure 18. The design style combines squares and circles and is mainly blue and white, in line with the design style of medical and environmental protection equipment. This disinfection and epidemic prevention robot solves the drawbacks of traditional disinfection robots, which cannot sterilize public spaces such as stairs. Moreover, traditional disinfection arms cannot sterilize places with many obstacles and narrow barriers. However, this disinfection and epidemic prevention robot can realize the integration of disinfection processing, temperature detection, real-time monitoring, material transportation, and other epidemic prevention functions.

4.2. Discussion

This paper proposes to combine QFD theory with the AHP method and fuzzy mathematics to improve objectivity and accuracy in design decision making. The method of importing the weight values derived from AHP, as important parameters for design decisions, into the left side of QFD has been used by many scholars in design decisions. The traditional QFD theory and AHP method do not take into account the human as the subject of decision making, which has insurmountable subjectivity and is difficult to quantify accurately in reality. The correlation analysis in QFD and the House of Quality needs to fully take into account the fuzziness in design decisions, and it is necessary to use fuzzy set theory to make correlation degree judgments, in order to analyze the relationships between social needs, technical features, and component features in robot design by qualitative and quantitative means, and to realize their mutual transformation. In order to solve the problem of ambiguity between user requirements and technical characteristics and component characteristics, many scholars have combined computer algorithms, such as genetic algorithms [28], artificial neural networks [29], and grey system theory [30]. Genetic algorithms can only compute a small number of optimal solutions. Artificial neural network models are considered to be black-box operations with many complex control parameters that require human control and manipulation, and their performance also depends on the data size, quality, and model structure. Grey system theory usually applies first-order linear differential equations to establish mapping relationships, which are suitable for dealing with input data, but it is difficult to precisely establish the correlation analysis between human needs and design features [31]. In contrast, the fuzzy theory is an effective mathematical tool for recounting real-world fuzzy phenomena, which can transform binary values that are not 0 or 1 into metrics in the natural environment, with less information missing in the arithmetic process of transformation [32]. Therefore, in this study, based on the fuzzy QFD method, the helper fuzzy AHP is imported into the left side to improve the objectivity of HQO decisions more naturally and accurately.

The ARIZ algorithm, as a derivative of the TRIZ innovation theory, is an important method in analyzing and solving problems to address engineering difficulties. The ARIZ algorithm combines all the methods of the TRIZ innovation theory to effectively solve practical problems. This algorithm is a continuous, incremental process that helps designers to establish a process for identifying problems, analyzing them, and solving them. In contrast to the use of traditional TRIZ methods, the process of solving complex problems with the ARIZ algorithm involves the continuous dissection, filtering, and narrowing of the initial problem. In the process of analyzing ARIZ, the problem is found and precisely analyzed and defined; in addition to the function of finding conflicting objects, the complete process of solving the problem is applied to the design in the form of an algorithm, and the complete logical process is extracted to program the design problem and solve the problem in a targeted, efficient, and fast manner to achieve the optimal goal of the design. Fuzzy QFD and the ARIZ algorithm represent a natural combination, and combining them in the design and development of disinfection and epidemic prevention robots can accurately capture user needs and social demands while effectively solving each complex engineering problem associated with robot design.

5. Conclusions

The method of fuzzy AHP was used to obtain the weight values of demands for disinfection and epidemic prevention robots in current society. In combination with the fuzzy QFD tool, the mapping relationship between the social demand for disinfection and epidemic prevention robots and the technical characteristics and component characteristics was established; thus, the contradictions and conflicts were obtained. Then, the ARIZ algorithm was applied to determine the solution principles of the conflicts. Finally, according to the solving principle and the importance of parts and components, a disinfection and epidemic prevention robot was designed in an all-around manner. Thus, the robot was innovated to include the functions of all-around disinfection and epidemic prevention, as well as climbing stairs, temperature measurement, real-time monitoring, and material transportation, so that a disinfection and epidemic prevention robot that can meet the needs and the demands of the present era and society was designed. The study also verified its scientificity and rationality.

This paper contributes in the following ways:

(1)

To reduce the bias and subjectivity in decision making, this paper combines AHP and QFD theories with fuzzy mathematics as a means to improve the objectivity of the design strategy.

(2)

A multi-level transformation of the fuzzy QFD tool was performed to ensure that the social demand for disinfection robots, the technical characteristics of disinfection robots, and the component characteristics of disinfection robots remain highly consistent.

(3)

The designed disinfection and epidemic prevention robot can replace people in all aspects and fields to complete the disinfection and epidemic prevention work.

This paper has several implications:

(1)

In the context of the globalization of epidemics, it is very important and necessary to design a robot that can completely replace humans in epidemic prevention tasks.

(2)

The research methods in this paper can also provide research ideas and references that can be used so solve other related industrial product design problems.

(3)

Combining statistical science, decision science, mathematical derivation, and other disciplines into design is a future research field that is very important.

However, this work neglected to perform a finite element analysis of the disinfection and epidemic prevention robot’s dynamic structure when it climbs stairs. Thus, we cannot confirm the stability and safety of the dynamic structure, and the robot needs to be further perfected in future research so that it can achieve higher efficiency, a wider application range, and more intelligent and safer operation.