Age-friendly design of health detection integrated machine: A user requirement-driven approach.

Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Sci Rep . 2026 Mar 3;16:11884. doi: 10.1038/s41598-026-42353-x Search in PMC Search in PubMed View in NLM Catalog Add to search Age-friendly design of health detection integrated machine: A user requirement-driven approach Yuanwu Shi Yuanwu Shi 1 School of Art and Design, Wuhan Textile University, Wuhan, 430073 China Find articles by Yuanwu Shi 1 , Yongxia Xie Yongxia Xie 1 School of Art and Design, Wuhan Textile University, Wuhan, 430073 China Find articles by Yongxia Xie 1, ✉ Author information Article notes Copyright and License information 1 School of Art and Design, Wuhan Textile University, Wuhan, 430073 China ✉ Corresponding author. Received 2025 Jul 18; Accepted 2026 Feb 25; Collection date 2026. © The Author(s) 2026 Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ . PMC Copyright notice PMCID: PMC13065823  PMID: 41776233 Abstract To precisely gauge and effectively prioritize the needs of elderly users, it is essential to transform these needs into key elements of product design and to explore innovative strategies for developing aging-friendly products that enhance user satisfaction. Initially, the analytic network process (ANP) is employed to construct a dual-layer correlation structure, which includes the network layer of elderly users’ requirements and the control layer of aging-friendly design specifications. This structure allows for the quantitative analysis of the relationships and feedback between elderly users’ needs and various design specifications, thereby determining a more accurate prioritization of these needs. Subsequently, the quality function deployment (QFD) method is utilized to create a House of Quality, which specifically examines the technical characteristics that require enhancement in products for the elderly. Finally, the integration of user journey mapping with the situational function-behavior-structure (FBS) model addresses the challenge of translating these needs into functional elements of the product, guiding the innovative design of aging-friendly products. Based on the needs of elderly users, demonstrated the application feasibility of this methodological framework in age-friendly product design, thereby improving the satisfaction of elderly care. The integrated application of ANP, QFD, and situational FBS effectively transforms the needs of elderly users into practical design directives and appropriate aging-friendly strategies. This study demonstrates the feasibility of this methodology through a case study involving the aging-friendly design of health detection integrated machine. Supplementary Information The online version contains supplementary material available at 10.1038/s41598-026-42353-x. Keywords: Demand analysis, Analytic Network Process, Quality Function Deployment, Function-Behavior-Structure, Elderly health detection integrated machine, Aging-friendly design Subject terms: Engineering, Health care, Mathematics and computing Introduction With the rapid advancement of modern technologies such as the Internet of Things, big data, cloud computing, and 5G, the concept of smart elderly care has garnered significant attention from both national and local governments. It has emerged as a crucial strategy for promoting active aging in the country 1 . However, the development of the smart elderly care industry in our country faces a critical challenge: the shortage of aging-friendly equipment. This shortage represents a significant bottleneck that urgently needs addressing. Currently, many industries involved in producing elderly care products neglect to consider the actual needs of the elderly or to align their designs with aging-friendly specifications. This oversight leads to resistance among the elderly, who may eventually abandon the use of these products. Consequently, these products not only fail to provide effective convenience but also result in wasted resources. Centering the needs of the elderly, fostering the high-quality development of smart elderly care, paying close attention to the preferences of elderly users, and accurately identifying their needs are essential steps needed to address current deficiencies in the supply side of smart elderly care. Currently, research on elderly product design from the perspective of requirement importance has formed a relatively systematic methodological framework, which can be primarily categorized into two research approaches. The first category focuses on the direct quantification and calculation of requirement importance. These methods typically integrate classical tools such as the Kano model, Analytic Hierarchy Process (AHP), and Quality Function Deployment (QFD) to construct requirement evaluation frameworks. This hybrid paradigm has been widely applied across various aging-friendly design scenarios. In the healthcare domain, Li et al. 2 and Ren et al. 3 combined Kansei engineering with fuzzy axiomatic design to optimize key design elements of blood glucose meters and hearing aids, respectively. In the mobility assistance domain, multiple studies have constructed combined frameworks including QFD-TRIZ 4 , KANO-AHP-FCE 5 , 6 , AHP-DEMATEL 7 , and Kano-TOPSIS 8 , successfully guiding the innovative design of walkers and smart wheelchairs. Furthermore, these methods have been extended to the optimization of smart home and elderly care facility design 9 – 12 as well as automotive human-machine interfaces 13 . However, the limitations of such methods lie in treating individual requirements as independent analytical units, failing to adequately reveal the inherent associations and interaction mechanisms among different types of elderly user requirements. The second category emphasizes the dynamic adjustment and optimization of requirement importance. To overcome the interference of subjective factors in traditional evaluation, Principal Component Analysis (PCA) 14 , game theory weighting 15 , and VIKOR multi-criteria ranking 16 have been introduced into the design process of elderly companion robots and smart home appliances. Meanwhile, for requirement elicitation under uncertain environments, Rough Set Theory (RST) combined with Grey Relational Analysis (GRA) 17 , and the Fuzzy Delphi Method 18 have proven effective in processing ambiguous information and exploring key antecedent configurations affecting satisfaction. Additionally, some studies have attempted to seek optimal solutions for customer satisfaction under cost constraints through scenario-driven bilevel requirement networks and improved TODIM methods 19 . Although these computational methods demonstrate excellent performance in correcting requirement preference biases within the elderly population, their perspectives are largely confined to subjective expectations on the user side, with limited attention to the mapping relationships between user requirements and existing design specifications and standard systems. The aforementioned computational methods for user requirement importance in elderly product design have, to a certain extent, circumvented the limitations of subjective evaluation and enhanced the accuracy of requirement weight quantification, providing relatively reasonable decision-making foundations for aging-friendly product design. Nevertheless, existing research still exhibits two deficiencies: first, the associative structures among various requirements have not been systematically modeled, making it difficult to reflect the interactions among requirement elements; second, the articulation mechanisms between user requirement analysis and design specifications warrant further exploration, as the mapping relationships between them remain unclear. To address these deficiencies, this paper proposes an integrated ANP-QFD-FBS model for elderly product requirement analysis. Compared with existing integrated methods such as AHP-QFD-FBS, the core innovations of this study are as follows: First, the Analytic Network Process (ANP) is introduced. Traditional AHP assumes that elements at each hierarchical level are mutually independent; however, elderly user requirements are often interrelated and mutually influential, rendering this assumption significantly limited when addressing complex requirements. ANP can handle dependency relationships and feedback mechanisms within network structures, enabling more precise capture and quantification of the complex relationships between the “user requirement network layer” and the “design specification control layer,” thereby obtaining more accurate requirement weights. Second, this study combines the FBS model with user journey mapping to construct a scenario-based FBS framework. Traditional FBS models focus on function-behavior-structure mapping but often remain detached from specific usage contexts. By introducing user journey mapping, this study situates the FBS model within concrete health monitoring scenarios for analysis, not only translating user requirements into product functions but also further decomposing functions into expected user behaviors under specific scenarios, thereby enhancing the practical guidance of design solutions. Based on this, the model employs QFD to transform weighted user requirements into actionable technical characteristics, and ultimately maps technical characteristics to dynamic usage scenarios through the scenario-based FBS framework, achieving a systematic transformation from requirement identification to design implementation. Research participants and ethical considerations This study has received ethical approval from the Ethics Committee of the School of Art and Design, Wuhan Textile University (Approval Numbers: WTUAD2025022801, WTUAD2025052001). All experiments were conducted in accordance with relevant guidelines and regulations. The main focus of this article is methodological research and does not involve research on human organs, animal tissues, or other organisms. All subjects were non-minors who had signed informed consent and voluntarily agreed to participate in this study. In addition, all data and information collected from participants were anonymous and no personal information was disclosed. Based on this information, all research methods and procedures outlined in this article were in compliance with ethical principles and regulations. Theoretical concepts and model integration framework Analytic network process The Analytic Network Process (ANP) 20 is a decision-making method developed from the Analytic Hierarchy Process (AHP). Its fundamental principle involves representing elements within a system as a network structure rather than a simple linear hierarchy. This method is highly practical and capable of objectively describing the connections between various entities. As a flexible tool, ANP enables decision-makers to identify potential solutions to complex problems. By decomposing a problem into a systematic hierarchical network that accounts for the interrelationships between different levels of elements and related criteria, ANP provides feedback on interdependent relationships and is, therefore, widely applied in related fields 21 . A significant strength of ANP is its ability to facilitate comparisons between any two control elements within groups, treating them as a single unit and forming an appropriate judgment matrix based on these comparisons. All factors are interdependent, maintaining comparative relationships that are integral to the network’s function. This interdependence is illustrated in Fig. 1 . Fig. 1. Open in a new tab Structure diagram of ANP network. Determination of user requirement importance and technical characteristic transformation Although elderly users desire medical-nursing products that are personalized to their needs, their lack of professional background and reliance on subjective perceptions often limit them to expressing vague and emotional expectations regarding product goals. Design specifications serve as a comprehensive framework covering product design goals, functional definitions, technical implementation, and constraints. This framework not only encompasses the emotional design intent and functional demands expressed by users but also delves into the technical details and constraints required to realize these needs. Essentially, product design is a process of translating user needs into concrete implementations while adhering to established design specifications. Since ANP 22 operates on the premise that interactions exist between pairs of factors and can accommodate correlations at the same hierarchical level, this study employs ANP to calculate various correlations and feedback relationships between design specifications and user requirement importance. This approach ensures that elderly user requirements are fully satisfied while effectively eliminating errors caused by highly divergent individual requirements and controlling the influence of unreasonable demands, thereby obtaining more authentic and precise importance ratings of elderly user requirements. To effectively guide design directions, clarifying user requirements and transforming them into specific technical characteristics is particularly critical. Quality Function Deployment (QFD) 23 [, 24 [, 25 , as a systematic analytical tool, is distinguished by its ability to establish close connections between user requirements and technical characteristics, facilitating the transformation process between them. This study constructs a relationship matrix between user requirements and technical characteristics, with the process model based on a dual-layer correlation logic that ensures both precise transmission of user requirements and effective optimization of technical characteristics, providing a solid foundation for further design refinement. The transformation process comprises two stages, as shown in Fig. 2 : the first stage employs ANP to establish a correlation model among user requirements and calculate their importance; the second stage utilizes QFD to transform user requirement importance into importance assessments of specific technical characteristics 26 [, 27 , thereby clarifying the priority ranking of design problems corresponding to technical characteristic importance, which can more precisely guide subsequent product development processes. Fig. 2. Open in a new tab Determination and conversion of user demand importance in dual-layer association. FBS mapping transformation model In 1990, Gero 28 introduced the Function-Behavior-Structure (FBS) model to analyze the interrelationships among function, behavior, and structure. The essential steps includ①Formulation; ②Synthesis; ③Analysis; ④Evaluation; ⑤Documentation; ⑥Reformulation type (new structure); ⑦Reformulation type (expected behavior); ⑧Reformulation type (new function), as illustrated in Fig. 3 . However, this model does not adequately represent user need identification and requirement definition, nor does it consider the relationship between user needs and design requirements. Fig. 3. Open in a new tab FBS model design flow chart. In 2013, Cascini et al 29 . proposed enhancements to the FBS model by incorporating user needs into the framework, leading to the development of the RE-FBS model. This revised model begins with the actual needs of users, defines critical functional indicators of the product (R→F), and thereby sets expectations for basic customer behavior (F→Be). Based on these expectations and the level of satisfaction achieved, the product’s structure is predicted (Be→S). The predicted structure is then articulated in a modeling language, enabling the formation of a potential structural-morphological scheme set (S→E), as shown in Fig. 4 . This model effectively transforms user needs and design elements, providing various structural feasibility options for specific design schemes. Fig. 4. Open in a new tab RE-FBS model design flow chart. The analysis presented in this study illustrates the primary process of constructing an integrated model based on ANP-QFD-FBS, as depicted in Fig. 5 . The first step involves using ANP to calculate the weight of user requirement. The second step consists of importing the importance of user requirement and the expanded table of product technical features into the HOQ to establish the demand-technical feature quality house model and to calculate the weight of the technical features. The third step involves using FBS mapping transformation to achieve elderly-friendly product design and to conduct a user satisfaction evaluation. Fig. 5. Open in a new tab Product design process based on ANP, QFD, and FBS integration. Methodological comparative analysis A systematic comparison of the proposed model with existing integrated design models is conducted to clearly demonstrate the methodological advantages of the ANP-QFD-FBS framework. Specific details are presented in Table 1 : Comparison of the proposed model with the AHP-QFD-FBS model; and Table 2 : Comparison of the proposed model with other integration methods. Table 1. Comparison between the Proposed Model and AHP-QFD-FBS Model. Comparison Dimension Traditional AHP-QFD-FBS Proposed ANP-QFD-FBS Handling of requirement correlations Assumes elements at each level are mutually independent, unable to handle interdependencies among requirements Capable of handling dependencies and feedback mechanisms in network structures, precisely capturing complex relationships between the “User Requirement Network Layer” and “Design Specification Control Layer” Weight calculation accuracy Ignore interactive effects among requirement elements, leading to weight bias Obtains more accurate requirement weights through supermatrix iterative calculation Applicable scenarios Suitable for simple products with relatively independent requirements More suitable for aging-friendly product design with complex, interrelated elderly user requirements Open in a new tab Table 2. Comparison between the proposed model and other integrated methods. Integrated Method Advantages Limitations Improvements in This Framework KANO-AHP-FCE Can identify requirement types Does not consider correlations among requirements, lacks contextual mapping ANP addresses correlation issues, contextualized FBS achieves scenario mapping QFD-TRIZ Strong innovative problem-solving capability Lacks systematic requirement weight calculation ANP provides systematic weight calculation AHP-DEMATEL Can analyze causal relationships Difficult to translate into specific design solutions FBS model achieves complete mapping from requirements to structure Open in a new tab A Core Methodological Advantages of This Framework. (1) Innovation of the Dual-Layer Correlation Structure. This study constructs a dual-layer correlation structure comprising the “User Requirement Network Layer” and “Design Specification Control Layer.” Unlike traditional methods that treat user requirements as independent analytical units, this framework systematically models the correlation structure among requirement elements, clarifies the linkage mechanism between user requirement analysis and design specifications, and enables dynamic adjustment of requirement weights. (2) Advantages of Contextualized FBS over Traditional FBS/RE-FBS. The traditional FBS model (Gero, 1990) focuses on abstract mapping of Function-Behavior-Structure but is detached from specific usage contexts. Even the improved RE-FBS model (Cascini et al., 2013), while introducing user requirements, still lacks systematic consideration of dynamic usage scenarios. This study combines the FBS model with user journey mapping to construct a contextualized FBS framework, with the following advantages: the detection process is divided into three consecutive scenarios—“pre-detection, during-detection, and post-detection”—with FBS mapping models established separately for each scenario; functions are no longer abstractly defined but decomposed into expected behaviors of elderly users in specific scenarios, enhancing the operability and contextual adaptability of design solutions; structural design directly responds to behavioral requirements under specific scenarios, ensuring that design decisions have clear contextual basis 30 . Transformation process of the ANP-QFD-FBS integration model Calculation of user requirement importance based on anp In this study, the ANP method is primarily employed to analyze the dual-layer correlation and to determine the importance of each user requirement through the calculation of the super matrix. The following steps are implemented in practice: Step 1: Elderly user requirements are gathered from a field survey. Subsequently, a requirement importance analytic network model is established for elderly users in a scientific and effective manner. The variable p refers to the design goal governed by the control network layer in ANP, while the control element represents the aging-friendly design specification. corresponds to the network layer associated with elderly user requirements 31 . Step 2: An unweighted relationship matrix is constructed focusing on the correlation between all elderly user requirements. The control element and the elderly user requirement associated with it are used as the criterion and standard, respectively. Two elements are grouped and compared within one unit to obtain the normalized ranking vector and to construct the judgment matrix, as indicated in Eq. ( 1 ). 1 The correlation values in the judgment matrices were obtained through ANP questionnaires designed based on preliminary user research and evaluated by invited experts. To ensure that the correlation values reflect the opinions of the majority and embody fairness and rationality, five experts from different fields were invited for the correlation value calculations, including two senior scholars in product design (with over 10 years of experience in age-friendly design research), two product managers from medical device enterprises (with experience in health monitoring equipment development), and one gerontology expert. The selection of five experts was based on the general principles of group decision-making in ANP methodology while considering the operability of the evaluation process. Expert scoring adopted the iterative feedback mechanism of the Delphi method: after the first round of scoring, results were compiled and fed back to each expert, with indicators showing high dispersion discussed in a second round, and the arithmetic mean of expert scores was ultimately adopted. Following the standard requirements of AHP/ANP methodology, we adopted CR < 0.10 as the threshold for judgment matrices to pass the consistency test. During the data processing phase, rigorous consistency tests were conducted on all judgment matrices generated from expert evaluations, ensuring that the CR values of all matrices were less than 0.10. For matrices with CR values exceeding 0.10, we employed the iterative feedback mechanism of the Delphi method to communicate with experts and re-evaluate until all judgment matrices satisfied the requirement of CR < 0.10, thereby ensuring the reliability of subsequent supermatrix calculation results 32 . Based on the questionnaire evaluations, correlation values were derived, and the maximum eigenvalue of the judgment matrix along with its corresponding eigenvector were obtained using the eigenvalue method , resulting in the judgment matrix . The correlations among elderly user requirements are then compared to produce the unweighted super matrix of elderly user requirements, as outlined in Eqs. ( 2 ) and ( 3 ). 2 3 From the calculations presented above, it is evident that the block matrix consists of . Specifically, indicates the relative importance of a particular requirement of an elderly user compared to the requirements of other elderly users, who serve as the baseline. However, alone does not represent the ultimate importance of these user requirements. Instead, it must be weighted through the weight relationship matrix, which is established by the elements of the control layer. Step 3: The weight relationship matrix for the control elements of the design criterion is constructed in accordance with the actual situation. Simultaneously, the weighted super matrix for elderly user requirements is established. ​ denotes the judgment criterion, while represents the control elements. The judgment matrix, represented by B, is constructed following the comparison of the correlation between two elements. refers to the weighted matrix, as illustrated in Eqs. ( 4 ) and ( 5 ). 4 5 The elements of the unweighted matrix are weighted by importance to derive a weighted super matrix, which delineates the importance relationships among elderly user requirements, as depicted in Eq. ( 6 ). 6 Step 4: The limit-weighted super matrix, denoted by , is created to calculate the weight of each user requirement. Although requirements vary among users, there is a degree of correlation among them. To minimize the impact of interfering factors, iterations are conducted times to ultimately derive the limit super matrix, represented by , as shown in Eq. ( 7 ). 7 The methodology behind calculating the importance of elderly user requirements, as outlined in the four steps, not only thoroughly considers the correlation among the control elements established by product design specifications but also effectively constructs the relationship between elderly user requirements. This approach significantly enhances the accuracy and reliability of the determined importance of these requirements. Technical characteristic transformation based on QFD Quality Function Deployment (QFD) refers to the process of translating user needs into product quality characteristics. It is a classic tool for analyzing the association between product attributes and user requirements, assisting engineering designers in better product design and innovation. In this study, to ensuring the final design fully satisfies elderly user needs—thereby improving user satisfaction—and to facilitate the identification of omitted needs or the timely improvement of low-impact designs during the process, an extension table of technical characteristics for aging-friendly products was created. This was achieved through an expert panel analysis of the correlation between the functions, structures, and components of aging-friendly products based on the aforementioned results. Subsequently, a QFD House of Quality (HOQ) relationship matrix was formed. To intuitively analyze user requirements, the abstract results of correlation are quantified and expressed by , as shown in Table 3 : Table 3. Correlation between elderly user requirements and technical characteristics Correlation strong correlation Medium correlation Weak correlation No correlation Value 9 3 1 0 Open in a new tab The matrix of the correlation between elderly user requirements and technical characteristics is constructed, as shown in Eq. ( 8 ): 8 Finally, the importance of technical characteristics is obtained, as shown in Eq. ( 9 ): 9 RE-FBS and scenario-based modeling The core value of designing products that are friendly to aging lies in their functional configuration and transformation. This is based on the evaluation of user needs and the importance of technical features, from which the importance ranking of key technical features has been established. The design direction is determined by this ranking, with special focus on the transformation and optimization of functions in products designed for aging. To achieve this, the innovative RE-FBS model and the situational awareness model will be utilized as analytical tools for an in-depth discussion on the design of medical and nursing products that are aging-friendly. It is important to clarify that the term “structure” in the model specifically refers to the product’s modeling structure and not to any mechanical design aspects such as structure, implementation code, etc. The transformation structure is illustrated in Fig. 6 : Fig. 6. Open in a new tab FBS mapping transformation structure diagram. Building on a profound understanding of the diverse needs of elderly users, and employing the ANP network hierarchical analysis method along with QFD, this study scientifically and rigorously analyzes and clarifies the needs of elderly users. Following the logic of the RE-FBS model (i.e., starting from user needs, defining functions, and then mapping these to specific structures), the target functions and corresponding structures of the product design are defined. This ensures that the design direction aligns with both user needs and practical applications. The transformation process guarantees that the final design solution accurately and effectively meets the personalized needs of elderly users. Case study: health detection integrated machine for older adults Research background and existing product analysis The health detection integrated machine is a product and service system designed to improve the health level of the elderly population by monitoring and warning about the user’s health status in real-time and intervening in risk factors affecting physical function. Currently, community-based intelligent health detection machine that integrate online medical consultation, appointment booking, health assessment services, and data detection are the most common health management devices. However, during preliminary development, the neglect of correlations between user needs and the constraints of design specifications and usage scenarios led to suboptimal usage frequency in real-world applications. To guide product design by correlating user need relevance, design specification constraints, and usage scenarios, the following work was conducted. Analysis of user requirements and aging-friendly design principles Through initial situational observations, interviews, and analysis of the elderly population, the pain points experienced by users of the smart health detection integrated device were identified. These were organized into four categories relevant to the health detection integrated device, as presented in Table 4 . Additionally, by combining user surveys, actual conditions, and relevant literature review, targeted elderly-friendly design principles for the smart health detection integrated device were summarized and proposed. Table 4. Summary of existing problems in the design of health detection integrated machine. Pain Points Exterior Device Functions Human-computer interaction Service Experience Description A. The human-machine size is not suitable. G. The device lacks a prompt function. N. Login processes are restrictive. Q. The operation procedures are unclear. B. The appearance is uninviting and lacks warmth. H. There is an absence of self-learning and advancement functions. O. The voice interaction system does not provide effective assistance. R. The operating procedures for different testing instruments are overly complicated. C. The arrangement of different types of testing instruments is impractical. I. The device lacks functional flexibility and has limited applicability. P. Mobile devices cannot share information. S. The feedback provided is inadequate. D. The device is not aligned with elderly users’ cognitive abilities and feels overly technical. J. There is no designated storage area for personal belongings. T. Insufficient consideration is given to elderly behaviors. E. The device’s labeling is unclear and lacks distinctive recognition. K. There is no safeguard against accidental operation. U. The device lacks targeted features. F. The layout of the main physical hardware is unreasonable. L. No sanitation or disinfection function is included. M. There is a lack of basic assistive functions during the detection process. Open in a new tab Combining the aforementioned user survey and practical conditions, we summarized and proposed targeted aging-friendly design principles for intelligent health detection integrated machines. These provide a basis for improving smart health detection integrated machine and offer guidance for future design and modification. Unlike general aging-friendly products, intelligent health detection integrated machines are subject to strict functional and technical specifications. Based on user research findings, including contextual observations and interviews with elderly users, as well as a comprehensive review of literature on age-friendly design standards and gerontological research, we derived the following design principles. These principles were further refined through consultation with designers and healthcare professionals to ensure their practical applicability. (1) Bidirectional learning principle. The elderly demographic exhibits unique characteristics. Their cognitive abilities and memory tend to decline gradually, making it challenging for them to learn complex operations quickly. Even after mastering these operations, they are likely to forget them after a short period. Additionally, they may develop a resistance or reluctance toward actively learning how to use new equipment. To address these issues effectively, it is essential to consider bidirectional learning and adaptation between the users and complex medical and nursing equipment. Generally, this involves gathering insights from users through various channels, such as analyzing their feedback and usage experiences. These insights can then be used to make adjustment and optimizations, thereby providing personalized and intelligent services tailored for elderly users. (2) Principles of professionalism and hygiene. Compared to other products for the elderly, testing equipment requires a high level of professionalism and accuracy to ensure reliable test results. From previous surveys and discussions with elderly individuals, it is evident that their judgments and evaluations of products are primarily based on past experiences and knowledge. They place significant importance on practical aspects such as product performance and value. Consequently, when designing testing equipment, it is crucial to accurately identify and meet the actual needs of the elderly. Furthermore, due to their weakened immune systems and lower resistance, the elderly are more susceptible to infectious diseases. It is therefore imperative to enhance their health protection by establishing disinfection zones and centralized areas for the disposal of medical waste, thereby effectively promoting healthy aging. (3) Principles of assistance and safety. As the elderly age, they increasingly require assistance during various physical examinations. ① Necessary Mobility Assistance: Due to changes in bodily functions, elderly users are prone to accidents such as falls or collisions when walking or standing. ② Operational Assistance: Assistance in operation is crucial for the elderly, involving voice cues and indicator light prompts. Providing such auxiliary reminders helps enhance the elderly user’s sense of security. ③ Feedback Assistance: After performing relevant operations according to instructions, elderly users are eager to know whether the action was correct and receive feedback on the next step. ④ Interaction Assistance: Some elderly people are slow to learn new things and knowledge, or even harbor rejection, leading to fear of using and refusing to understand medical-nursing products. Considering these issues, humanized design should be applied to medical-nursing products, introducing indicator lights and voice assistance to improve the level of humanized service. (4) Modularization and systematization principles. Modular design and standardized design refer to adding or removing multiple modules based on original performance and configuring modules that meet functional requirements according to different needs, satisfying diverse demands. Modular structures can make equipment maintenance and replacement more convenient, further enhancing product application flexibility. Additionally, scientific layout among modules can improve operational speed and safety performance. (5) Principles of caring and experience. In terms of product characteristics, intelligent health detection integrated machines often convey a relative coldness, causing psychological fear in users. Some researchers suggest that rounded designs should be added to the product’s form, and colors should be softer and warmer. Furthermore, the caring nature of the product experience should be considered, specifically long-term experience, covering the pre-use, during-use, and post-use stages. Construction of analytic network structure model Through the analysis described above, relevant design constraints were extracted from the elderly-friendly design specifications of the integrated health detection device. Additionally, a correlation analysis of the needs of elderly users was conducted based on these design specifications, as illustrated in Table 5 : Table 5. Analysis of the correlation between health detection integrated machine design specifications and user needs. Control element User requirement layer No. Control element User requirement layer No. Technicality A1 Professional medical service product A11 Hygiene A2 Disinfection area A21 Accurate detection report A12 Disposable medical waste Processing area A22 Aid A3 Operation reminder aid A31 Modularizing principle A4 High operating efficiency A41 Action aid A32 Easy maintenance and replacement A42 Interactive aid A33 Caring principle A5 Soft modeling A51 Service aid A34 Man-machine harmony in size A52 Open in a new tab By constructing and analyzing the network structure model, more accurate results regarding demand association can be obtained, as shown in Fig. 7 . Fig. 7. Open in a new tab Network structure model of health detection integrated machines adapted to aging user needs. Calculation of user requirement importance A judgment matrix for all elderly user requirements is established based on their interrelationships, utilizing the correlation values derived from the ANP-based Wenjuanxing survey results. The maximum eigenvalue of each judgment matrix is computed and compiled into a matrix. Subsequently, an unweighted super matrix is derived using Eqs. ( 2 ) and ( 3 ), as presented in Table 6 . Table 6. Unweighted super matrix Wp of health detection integrated machine user needs. A1 A2 A3 A4 A5 A11 A12 A21 A22 A31 A32 A33 A34 A41 A42 A51 A52 A1 A11 0 1 1 1 0.875 0.857143 0.833333 1 0.857143 0 0 0.833333 A12 1 0 0 0 0.125 0.142857 0.166667 0 0.142857 0 0 0.166667 A2 A21 0 1 0 1 0 0 0 0 0 0.833333 0 0.875 A22 1 0 1 0 0 0 1 0 1 0.166667 0 0.125 A3 A31 0.61471 0.606359 0 1 0 0 1 0 0.542261 0 0 0.875 A32 0.125778 0.120275 0 0 0 0 0 0 0.068425 0 1 0.125 A33 0.158406 0.186893 0 0 0 1 0 1 0.264047 0 0 0 A34 0.101106 0.086473 0 0 0 0 0 0 0.125266 0 0 0 A4 A41 1 0 0 0 1 0 0 0 0 0 0 1 A42 0 0 0 1 0 0 0 0 0 0 0 0 A5 A51 0 0 0 0 0 0.833333 0 0 0 0 0 1 A52 1 1 1 1 0 0.166667 1 0 1 1 1 0 Open in a new tab Following an analysis of relevant elderly-friendly design principles, a weight relationship matrix for the design criteria control elements is constructed using the eigenvalue method, which computes the eigenvector corresponding to the maximum eigenvalue of each judgment matrix. This matrix illustrates the relative importance of each design criterion control element. The weighted super matrix for the health detection integrated machine’s user requirements is then generated using Formula (6), as depicted in Table 7 . The extreme weighted super matrix is calculated using Formula (7), presented in Table 8 , which yields the importance of each elderly user requirement. The importance assigned to each elderly user is represented by the values in each row of the extreme weighted super matrix, as shown in Table 9 . Table 7. Weighted super matrix W of health detection integrated machine user needs. A1 A2 A3 A4 A5 A11 A12 A21 A22 A31 A32 A33 A34 A41 A42 A51 A52 A1 A11 0 0.606386 0.690413 0.529048 0.667054 0.553886 0.466035 0.86515 0.214807 0 0 0.398265 A12 0.557617 0 0 0 0.095293 0.092314 0.093207 0 0.035801 0 0 0.079653 A2 A21 0 0.154459 0 0.15289 0 0 0 0 0 0.071994 0 0.088588 A22 0.142037 0 0.199523 0 0 0 0.134569 0 0.058776 0.014399 0 0.012655 A3 A31 0.077387 0.083012 0 0.080831 0 0 0.087168 0 0.037449 0 0 0.075185 A32 0.015835 0.016466 0 0 0 0 0 0 0.004726 0 0.411329 0.010741 A33 0.019942 0.025586 0 0 0 0.100723 0 0.13485 0.018236 0 0 0 A34 0.012728 0.011838 0 0 0 0 0 0 0.008651 0 0 0 A4 A41 0.080426 0 0 0 0.237653 0 0 0 0 0 0 0.211942 A42 0 0 0 0.15289 0 0 0 0 0 0 0 0 A5 A51 0 0 0 0 0 0.210898 0 0 0 0 0 0.122972 A52 0.094028 0.102251 0.110065 0.08434 0 0.04218 0.219021 0 0.621555 0.913607 0.588671 0 Open in a new tab Table 8. Limit-weighted super matrix of health detection integrated machine user needs. A1 A2 A3 A4 A5 A11 A12 A21 A22 A31 A32 A33 A34 A41 A42 A51 A52 A1 A11 0.332891 0.332891 0.332891 0.332891 0.332891 0.332891 0.332891 0.332891 0.332891 0.332891 0.332891 0.332891 A12 0.207917 0.207917 0.207917 0.207917 0.207917 0.207917 0.207917 0.207917 0.207917 0.207917 0.207917 0.207917 A2 A21 0.054779 0.054779 0.054779 0.054779 0.054779 0.054779 0.054779 0.054779 0.054779 0.054779 0.054779 0.054779 A22 0.066315 0.066315 0.066315 0.066315 0.066315 0.066315 0.066315 0.066315 0.066315 0.066315 0.066315 0.066315 A3 A31 0.062411 0.062411 0.062411 0.062411 0.062411 0.062411 0.062411 0.062411 0.062411 0.062411 0.062411 0.062411 A32 0.018822 0.018822 0.018822 0.018822 0.018822 0.018822 0.018822 0.018822 0.018822 0.018822 0.018822 0.018822 A33 0.016112 0.016112 0.016112 0.016112 0.016112 0.016112 0.016112 0.016112 0.016112 0.016112 0.016112 0.016112 A34 0.007303 0.007303 0.007303 0.007303 0.007303 0.007303 0.007303 0.007303 0.007303 0.007303 0.007303 0.007303 A4 A41 0.069824 0.069824 0.069824 0.069824 0.069824 0.069824 0.069824 0.069824 0.069824 0.069824 0.069824 0.069824 A42 0.010139 0.010139 0.010139 0.010139 0.010139 0.010139 0.010139 0.010139 0.010139 0.010139 0.010139 0.010139 A5 A51 0.020342 0.020342 0.020342 0.020342 0.020342 0.020342 0.020342 0.020342 0.020342 0.020342 0.020342 0.020342 A52 0.133143 0.133143 0.133143 0.133143 0.133143 0.133143 0.133143 0.133143 0.133143 0.133143 0.133143 0.133143 Open in a new tab Table 9. Importance of user requirement. No. A11 A12 A21 A22 A31 A32 A33 A34 A41 A42 A 51 A52 Weight 0.332891 0.207917 0.054779 0.066315 0.062411 0.018822 0.016112 0.007303 0.069824 0.010139 0.020342 0.133143 Open in a new tab Sensitivity analysis (1) Control Layer Weight Perturbation Analysis. In the ANP model, the control layer weights constituted by aging-friendly design specifications directly affect the final user requirement weights. To examine the sensitivity of results to changes in control layer weights, we adopted the following method: Based on the original weights of the five design specifications in the control layer, perturbations of ± 10%, ± 20%, and ± 30% were applied to each control layer element respectively. The specific operation was: when a certain control layer element weight was increased (or decreased) by a certain proportion, the remaining four elements were proportionally scaled to maintain the total weight sum of 1. For each perturbation scenario, the weighted supermatrix and limit supermatrix were recalculated to obtain new weights and rankings for the 12 requirements, and Spearman rank correlation analysis was conducted with the original ranking. The results are shown in the following Table 10 : Table 10. Control Layer Weight Perturbation Analysis Results. Perturbation Magnitude Weight Ranking Change Spearman Rank Correlation Coefficient ±10% Top 6 core requirement rankings unchanged ρ = 0.98 ( p < 0.001) ±20% Top 4 core requirement rankings unchanged, 5th-6th items (occasional interchange) ρ = 0.95 ( p < 0.001) ±30% Top 3 core requirement rankings unchanged, local adjustments in middle levels ρ = 0.90 ( p < 0.001) Open in a new tab Analytical Explanation: We calculated the Spearman rank correlation coefficients between the perturbed requirement weight rankings and the original rankings. Under all perturbation conditions, ρ values were greater than 0.90, and all were significant at the p < 0.001 level. This means that the perturbed rankings maintained a highly consistent monotonic relationship with the original rankings. Particularly, with ± 10% perturbation, ρ = 0.98, indicating that within a reasonable range of control layer weight estimation error, the requirement priority ranking is almost unaffected. (2) Sensitivity Analysis of Expert Judgment Matrices. Considering that the correlation values in ANP judgment matrices originate from expert subjective evaluations, we conducted sensitivity testing on the pairwise comparison values in the judgment matrices: First, through partial derivative analysis, we identified the key judgment matrices that have the greatest impact on final weights. From the unweighted supermatrix structure in Table 6 , the pairwise comparisons within the Professionalism (A1) dimension where A11 and A12 are located, as well as the cross-dimensional comparisons between A52 and other requirements, have the most significant impact on final weights. Pairwise comparison values in these key judgment matrices were selected, and perturbations of ± 1 scale level were applied (e.g., if the original comparison value was 5, it was adjusted to 4 and 6 respectively). The eigenvector, supermatrix, and limit matrix were then recalculated, and final weight changes were observed. Results: After applying ± 1 scale perturbations to individual elements of key judgment matrices: the change magnitude of each user requirement weight was within 5%; the Kendall τ correlation coefficient of requirement rankings was greater than 0.85 under all perturbation scenarios; no fundamental changes occurred in the ranking of core requirements (top three items). This indicates that the ANP model in this study has good fault tolerance for deviations in individual expert judgments. (3) “Leave-One-Out” Sensitivity Analysis by Removing Individual Experts. This study invited five experts from different fields to conduct judgment matrix scoring, including two senior product design scholars, two medical device company product managers, and one gerontology expert. To verify whether the final results were excessively influenced by any single expert, a “leave-one-out” method was used for stability testing: Method: Each expert’s scoring data was sequentially removed, and all judgment matrices were reconstructed using the arithmetic mean of the remaining four experts’ scores. The complete ANP calculation process (unweighted supermatrix → weighted supermatrix → limit supermatrix) was executed to obtain five new sets of requirement weight rankings, which were then compared with the original complete five-person ranking through Spearman rank correlation analysis. Results: The Spearman rank correlation coefficients between the requirement weight rankings obtained from five leave-one-out analyses and the original results ranged from ρ = 0.96 to 0.99, indicating that no single expert’s judgment dominated the final results, and the expert panel’s judgments demonstrated good consistency and stability. (4) Comprehensive Conclusions of Sensitivity Analysis. The results of the three sensitivity analyses consistently indicate: The user requirement weight rankings obtained through ANP in this study have good robustness; Core requirements (top three rankings: A11, A12, A52) remained stable under various perturbation conditions, providing reliable input basis for subsequent QFD technical characteristic transformation and FBS design mapping; The model has low sensitivity to changes in individual parameters and deviations in individual expert judgments, enhancing the credibility of research conclusions. Technical characteristic deployment. To ensure the uniformity and comprehensiveness of technical feature identification, this study employed a combination of field investigation, literature review, and semi-structured interviews. Technical personnel and relevant stakeholders from health detection integrated machine manufacturing enterprises were consulted to collect data pertaining to product technical features. In this study, the technical features of health detection integrated machine specifically refer to the functional modules and their spatial layout configurations of the health detection integrated machine (based on the technical specifications of the “PRESON” Health Device Model PRS-X as displayed on the official website, see Fig. 8 ). Through systematic organization of the collected information, descriptive technical terminology was formulated. The technical features of health detection integrated machine (i.e., functional requirement indicators) are presented in Table 11 . Fig. 8. Open in a new tab PRESON Health Testing Machine PRS-X1. Table 11. Technical characteristics of health detection integrated machine. No. Technical characteristics No. Technical characteristics T1 Body monitoring instruments T7 Indicator light with color aid T2 Terminal software and hardware T8 Information sharing via mobile terminals T3 Sanitary and disinfection T9 Hardware function layout T4 Medical waste disposal T10 Replacement of different modules T5 Voice- and key-aided reminders T11 Use and emotional experience T6 Grasping aid T12 Human-machine interaction Open in a new tab A House of Quality (HOQ) model for health monitoring equipment design was established to calculate technical characteristic importance, with weights normalized and ranked. Characteristics ranking in the top 50% and strongly correlated with core user pain points were marked as “improve,” while characteristics with lower weights or already mature technology were marked as “defer.” Based on this, qualitative analysis of technical characteristics was conducted to clarify correlations among characteristics (“Roof” construction: to verify technical compatibility and resolve conflicts before the FBS mapping stage), as shown in Table 12 . Table 12. HOQ construction of health detection integrated machine. Open in a new tab Age-friendly design analysis of health detection integrated machine: FBS mapping and transformation The primary function of a product is its most crucial characteristic. As a highly specialized and sophisticated device, the optimization and expansion of the functions of health detection devices have become the focus of this study. This work does not delve into the functional components or intricate structures inside the health detection integrated machine. Instead, it concentrates on integrating existing functions and exploring potential auxiliary functions. Utilizing the RE-FBS model and the contextual awareness model as analytical tools, an in-depth discussion of the elderly-friendly design of the health detection integrated machine is conducted. Requirement transformation (R→F) The initial six technical features of the integrated health detection device identified in the HOQ, namely T1 various body monitoring instruments, T2 terminal hardware and software, T9 hardware function layout, T12 human-computer interaction, T5 voice-assisted reminder, and T4 medical waste treatment, are translated into specific functional indicators, as depicted in Fig. 9 . This process helps establish a functional library for health detection devices. It is important to note that the “function-behavior-structure” mapping conversion is not applied to T1 (various body monitoring instruments) and T2 (terminal hardware and software) due to their specialized involvement in medical and computer technology. Fig. 9. Open in a new tab Extraction of key functional indicators. The converted functional indicators are explained as follows: (1) Modular Function: The primary role of the modular function is to reorganize and classify existing modules so that the resulting multiple modules better adapt to user behavior. (2) Storage Function: This refers to the placement of disposable replacement medical equipment, storage of printed reports, and placement of user personal items. Essentially, the storage function must ensure that various items are effectively organized, providing both storage and protection. (3) Multi-Mode Login Function: The current health detection devices necessitate a login using a password and an ID card (number). In scenarios where users either do not remember their password or do not have their ID card, logging in becomes impossible. Hence, the introduction of multiple login methods, such as NFC, QR code scanning, fingerprint recognition, and facial recognition, is essential. (4) Voice Prompt Function: When a voice prompt is issued, users notice that the corresponding indicator light is on and receive operational reminders, ensuring they can quickly achieve their desired goals. (5) Medical Waste Recycling Function: Certain devices are equipped with disposable testing supplies, which are to be placed in a disinfection compartment before use. Post-use, these supplies must be collected and processed in designated medical waste bins. The placement of these bins is strategically determined by proximity principles and the specific sequence of operations. Construction of user expected scenarios (F→ Be) . Based on the previous analysis of health detection integrated machine usage processes combined with user task scenario analysis, numerous issues inconsistent with elderly user behaviors were identified. Building on these analytical results and leveraging context awareness, opportunity points for age-friendly functional design were identified from the perspectives of users, environment, and task scenarios, combined with extracted functional indicators, and a user expected behavior model was established. Given the high continuity during health detection integrated machine use, we divided the entire detection process into pre-detection, during-detection, and post-detection scenarios, and drew the expected behavior model according to the user journey map, as shown in Fig. 10 . FBS models are established for the situations before, during, and after the test, as shown in Figs. 11 , 12 and 13 ; specifically, the FBS models for the “pre-test situation” and the “blood glucose test situation (during the test)” are illustrated in Figs. 11 and 12 , respectively, with the FBS model for the “post-test situation” presented in Fig. 13 . Fig. 10. Open in a new tab User expected behavior journey. Fig. 11. Open in a new tab FBS model of pre-detection scenario. Fig. 12. Open in a new tab FBS Model for Blood Glucose Detection During the Test. Fig. 13. Open in a new tab FBS model for blood glucose detection. Structural design and stylish transformation (Be→ S) Following the development of the user-expected behavior model and the structural mapping of the scenarios before, during, and after the test, the aging-friendly features of health detection devices were thoroughly considered. Structural areas such as storage, multi-functional login identification, modular instrument placement, medical waste treatment, auxiliary handrails, and protection were designed. This culminated in the completion of an aging-friendly design plan for health detection devices, as shown in Fig. 14 . The product structure and details are illustrated in Figs. 15 and 16 , respectively. Fig. 14. Open in a new tab Rendering diagram of the smart health detection device solution. Fig. 15. Open in a new tab Structure diagram of smart health detection device. Fig. 16. Open in a new tab Product details ( a )Testing instrument disinfection placement module ( b ) blood pressure measurement ( c ) Auxiliary protection ss ( d ) human-machine interaction. User satisfaction evaluation To verify the actual satisfaction of the design proposal among the elderly user group, the System Usability Scale (SUS) 33 was employed to test and analyze user experience satisfaction for both the designed health detection integrated machine and existing integrated health devices currently on the market. This study utilized purposive sampling to recruit eligible elderly respondents through community health service centers. The inclusion criteria were: ①Self-sufficient elderly individuals who, despite exhibiting certain aging-related declines in physical function compared to their youth, maintain daily living behaviors indistinguishable from the general population, can independently carry out various activities to meet their needs, are open to accepting new things, and possess certain learning abilities. ②Subjects must have prior experience using intelligent integrated health monitoring devices. To satisfy these conditions, 74 self-sufficient elderly individuals were ultimately selected as questionnaire respondents, with 60 valid questionnaires collected. The key demographic characteristics of the respondents are shown in Table 13 . The specific test content is shown in Table 14 . Reliability and validity analyses were conducted on this questionnaire (due to the item settings being divided into positive and negative questions, the questionnaire was split for reliability and validity analysis), as shown in Table 15 . Table 13. Summary of respondent demographics ( N =60). Characteristics classification Number of people ( n ) Percentage (%) gender male 26 43.3% female 34 56.7% age 60–65 years old 18 30.0% 66–70 years old 22 36.7% 71–75 years old 14 23.3% 76 years and older 6 10.0% Educational attainment Primary school and below 8 13.3% junior high school 16 26.7% High school/Vocational school 21 35.0% Associate degree or higher 15 25.0% Experience using smart devices Experienced 60 100% Open in a new tab Table 14. User satisfaction test questionnaire. Serial number Question customer satisfaction 1 I would be willing to use this device frequently for health checks 0 1 2 3 4 2 There are too many interface elements, and you often don’t know where to start. 0 1 2 3 4 3 Check that each element of the device interface is consistent with my usage habits 0 1 2 3 4 4 The use of testing equipment requires guidance or help from others 0 1 2 3 4 5 The spatial layout and functional integration of the testing equipment are very good 0 1 2 3 4 6 The operations between various detection modules are not consistent 0 1 2 3 4 7 Most people can quickly learn to use health monitoring equipment 0 1 2 3 4 8 The detection process is troublesome and I am not very willing to use this device 0 1 2 3 4 9 I feel happy and confident when using health monitoring equipment 0 1 2 3 4 10 There are many things I need to know to use health monitoring equipment 0 1 2 3 4 Open in a new tab Table 15. Questionnaire reliability and validity analysis. Reliability/ Validity Analysis Project Cronbach.𝛼 coefficient and KMO value Reliability Analysis 1 Positive Items 1、3、5、7、9 0.833 Reliability Analysis 2 Negative Item 2、4、6、8、10 0.875 Validity Analysis 1 Positive Items 1、3、5、7、9 0.840 Validity Analysis 2 Negative Item 2、4、6、8、10 0.860 Open in a new tab Reliability primarily assesses the accuracy, stability, and consistency of a scale, which reflects the extent of variation in measurement results due to random errors in the measurement process. The Cronbach.α coefficient is the most commonly used measure of reliability. This coefficient measures the internal consistency of the scale by evaluating the agreement among scores for each item. The analysis shows that the Cronbach.𝛼 coefficients for this questionnaire are all above 0.8, with a value of 0.833 for positive items and 0.875 for negative items, indicating good reliability. Validity, on the other hand, assesses the accuracy, effectiveness, and correctness of the scale, specifically whether the scale accurately represents what it is intended to measure. Evaluating validity can be complex; therefore, the KMO value is employed here to gauge the validity. A KMO value above 0.8 suggests that the research data are well-suited for extracting meaningful information, indirectly indicating strong validity. The KMO value for this questionnaire exceeds 0.8, confirming its high validity. The method for calculating the SUS score is as follows: ① Determine the conversion score for each question, which ranges from 0 to 4. For positive items (odd-numbered questions), the conversion value is the original score of the scale minus 1 (Xi-1). For negative items (even-numbered questions), the conversion score is 5 minus the original score (5-Xi). ② Sum the conversion scores from all questions and multiply by 2.5 to derive the total SUS score, which ranges from 0 to 100. ③ Convert the original score to a percentile, refer to Table 16 , locate the corresponding rating, and determine the test subject’s score. Table 16. Curved grading range of SUS total score. SUS Total score Rating Percentile rank SUS Total score Rating Percentile rank SUS Total score Rating Percentile rank 84.1–100.1 A+ 96–100 74.1–77.1.1.1 B 70–79 62.7 - 64.9 C- 35–40 80.8–84.0.8.0 A 90–95 72.6–74.0.6.0 B 65–69 51.7 - 62.6 D 15–34 78.9 - 80.7 A 85–89 71.1 - 72.5 C+ 60–64 0 - 51.7 F 0–14 77.2–78.8.2.8 B+ 80–84 65 - 71 C 41–59 Open in a new tab Data processing and descriptive statistics Based on the questionnaire data, a statistical analysis was performed on the SUS scores of the 60 valid respondents, as shown in Table 17 . Table 17. Descriptive statistics results for SUS scores of the two groups. Group Sample size ( n ) Mean (M) Standard deviation (SD) minimum value Maximum value 95% CI Before improvement (commercially available product) 60 52.00 6.89 37.50 72.50 [50.22, 53.78] Improved (design plan) 60 78.50 12.03 37.50 92.50 [75.39, 81.61] Open in a new tab Statistical Test Analysis (1) Independent Samples t-test. To verify whether the improved design significantly enhanced user satisfaction, an independent samples t-test was conducted on the SUS scores of the two groups. The SUS scores for the improved health detection integrated machine (M = 78.50, SD = 12.03) were significantly higher than those for existing products on the market (M = 52.00, SD = 6.89), t (118) = 14.87, p < 0.001. According to the SUS total score curve grading range in Table 12 : the pre-improvement product mean score of 52.00 corresponds to Grade D (percentile rank 15–34), indicating that the user experience of existing health integrated machines on the market is at a relatively low level; the post-improvement product mean score of 78.50 corresponds to Grade B+ (percentile rank 80–84), indicating that the design solution achieved a good level of user experience. (2) Effect Size Analysis. Cohen’s d calculation: The effect size Cohen’s d between the two groups = 2.70, which constitutes a large effect size ( d > 0.8 indicates a large effect), demonstrating that the user satisfaction improvement brought by the improved design has strong practical significance. (3) Confidence Interval Analysis. 95% confidence interval calculation: The 95% confidence interval for the difference between the two groups’ scores was [22.97, 30.03]. This interval does not contain zero, further supporting the significance of the difference between the two groups (as shown in Table 18 ), Specifically, we are 95% statistically confident that the improved design will result in an increase in the SUS score of between 22.97 and 30.03 points compared to existing products on the market, indicating that the improved design solution is superior to existing products on the market in terms of user satisfaction. Table 18. Mean SUS scores with 95% confidence intervals. Open in a new tab Based on the comprehensive statistical analysis results: The improved health detection integrated machine scored significantly higher than existing products on the market in SUS user satisfaction scores ( p < 0.001), with a score increase of 26.50 points (an improvement of 50.96%). The large effect size of Cohen’s d = 2.70 indicates that the health detection integrated machine design solution in this study can satisfy the genuine needs of most elderly users. By simplifying operational processes, providing clear voice guidance and feedback, and ensuring equipment comfort and safety, the design reduces usage barriers and psychological burden for elderly users, thereby creating a pleasant health examination experience for them. Conclusion To establish practical correlations among the diverse needs of elderly users, enhance the guiding role of design specifications in elderly-oriented product development, and derive more precise user requirement weights for subsequent translation into key design elements of age-friendly products, this study integrates the Analytic Network Process (ANP), Quality Function Deployment (QFD), and Function-Behavior-Structure (FBS) model to achieve innovative design. At the theoretical level, the core contribution of this study lies in constructing a more systematic design transformation framework with dynamic feedback capabilities. By introducing ANP as a substitute for the traditional Analytic Hierarchy Process (AHP), this approach effectively overcomes the limitations of hierarchical analysis in addressing interdependencies among requirements, enabling more precise capture of complex correlations among elderly user needs and effectively enhancing the accuracy of user requirement identification. Furthermore, by integrating the FBS model with user journey mapping to construct a scenario-based analytical framework, the mapping process from function to behavior to structure becomes more aligned with actual usage workflows, thereby helps to enhance the efficiency and practical guidance of design transformation. At the practical level, this innovative design paradigm substantially enhances the granularity and accuracy of user requirement identification, effectively addressing the prevalent challenges in traditional age-friendly product design phases, including ambiguous user requirements and insufficiently scientific solution formulation. It facilitates optimal allocation of design resources from a holistic perspective, ensuring that each design decision is grounded in profound understanding and precise response to user needs. Through the empirical validation using a health detection integrated machine, this study demonstrates the effectiveness of the integrated application of ANP, QFD, and FBS. This integrated model not only clearly reveals the dynamic interaction between design specifications and user feedback but also elucidates how user requirements are refined into specific product functional characteristics and ultimately translated into key elements in design practice. This process strengthens the scientific rigor and specificity of design, providing a clear and traceable pathway and guidance for innovative age-friendly product design. Limitations and future work This study has several limitations. First, user satisfaction evaluation was conducted with only 60 elderly participants in Wuhan, China, and this geographical concentration of the sample may limit the generalizability of research findings to elderly populations with different cultural backgrounds or healthcare systems. Second, the ANP judgment matrices and QFD correlation assessments rely on expert subjective judgment; although multiple experts were engaged to mitigate bias, subjective factors cannot be entirely eliminated. Additionally, this study employed only a health monitoring integrated device as a case for methodological validation, and the applicability to diverse product categories such as mobility aids and home care devices remains to be further examined. To address these limitations, future research may proceed in the following directions: first, expanding the geographical coverage and cultural diversity of samples to enhance the external validity of research conclusions; second, introducing data-driven approaches such as machine learning to reduce subjective dependence in expert evaluation processes; third, validating the generalizability of this integrated framework across a broader range of age-friendly product categories. Furthermore, subsequent research will focus on deeper exploration of latent needs among elderly users, continuously enriching and refining the theoretical framework for age-friendly product design, providing more humanized and practical product solutions for the elderly population, and contributing to the construction of a more age-friendly social environment. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 (1.4MB, pdf) Acknowledgements The authors would like to express appreciation to the following contributor: Wuhan Textile University Key Project：Development and design of smart wearable products based on the Internet of Things (No. 2024441). Author contributions Yuan Wu Shi: Conceptualization, Methodology, PI of the research Funding and Editing, Revising. Yong Xia Xie: Data curation, conceptual design formulation, Editing. All authors have reviewed the manuscript. Funding The authors would like to express appreciation to the following contributor: Wuhan Textile University Fund Special Project: Development and Design of Smart Wearable Products Based on the Internet of Things (No. 2024441). Data availability The datasets used and/or analysed during the current study available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References 1. Dou, J. & Qin, J. Age-friendly design of smart health care products and research methods for elderly users. Packaging Eng. 42 (6), 62–68. 10.19554/j.cnki.1001-3563.2021.06.009 (2021). [ Google Scholar ] 2. Xiaojie & Li Hong Li. Age-appropriate design of domestic intelligent medical products: An example of smart blood glucose detector for the elderly with AHP-QFD Joint KE. Heliyon 10 (5), e27387. 10.1016/J.HELIYON.2024.E27387 (2024). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 3. Yan sheng Ren & Kangheui Cha. Optimization strategies for elderly hearing aid user satisfaction based on the Kano QFD integrated model and FAD theory. Sci. Rep. 15 (1), 34441–34441. 10.1038/S41598-025-17546-5 (2025). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 4. Zeng, X. & Yi, M. D. Design of wheelchair for the elderly based on QFD/TRIZ integrated model. Mech. Des. 38 (4), 134–138. 10.13841/j.cnki.jxsj.2021.04.022 (2021). [ Google Scholar ] 5. Tao Wang, Y., Zhao, X. & Zhang, Y. X. Leah Ling Li Pang. Design of walking aids for the elderly based on the Kano-AHP-FEC method. Sci. Rep. 15 (1), 2663–2663. 10.1038/S41598-025-85540-Y (2025). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 6. Liu, C. & Zhang, H. Research on the design of walking aids for elderly people with Parkinson’s disease based on AHP-FCE evaluation. Packaging Eng. 46 (8), 98–105. 10.19554/j.cnki.1001-3563.2025.08.009 (2025). [ Google Scholar ] 7. Sun, M. & Zhao, Z. Research on the design of elderly walking aid products based on AHP-DEMATEL/QFD theory. Packaging Eng. 46 (10), 128–138. 10.19554/j.cnki.1001-3563.2025.10.013 (2025). [ Google Scholar ] 8. Ye, D., Tao, C. & Duan, D. Research on the design path of elderly smart wheelchair products integrating Kano/AHP/TOPSIS. Packaging Eng. 46 (2), 311–322. 10.19554/j.cnki.1001-3563.2025.02.028 (2025). [ Google Scholar ] 9. Li, Yongchuan & Ghazilla Raja Ariffin Raja & Abdul Rashid Salwa Hanim. QFD-Based Research on Sustainable User Experience Optimization Design of Smart Home Products for the Elderly: A Case Study of Smart Refrigerators. Int. J. Environ. Res. Public Health . 19 (21), 13742–13742. 10.3390/IJERPH192113742 (2022). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. Zhigang, H. & Jia, D. Nan Zhang. Construction and application of product optimisation design model driven by user requirements. Sci. Rep. 14 (1), 16381. 10.1038/S41598-024-67406-X (2024). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Wang, T. & Wang, R. Multi-level fuzzy comprehensive evaluation of satisfaction with elderly-friendly products in elderly care facilities. Packaging Eng. 43 (12), 192–198. 10.19554/j.cnki.1001-3563.2022.12.022 (2022). [ Google Scholar ] 12. Yang, H. Research on the design of toilet for the elderly based on barrier-free concept. Packaging Eng. 45 (4), 326–335. 10.19554/j.cnki.1001-3563.2024.04.035 (2024). [ Google Scholar ] 13. Wei Liu, Y., Li & Jindan Cai. Research on Aging Design of Passenger Car Center Control Interface Based on Kano/AHP/QFD Models. Electronics , 13 (24), 5004–5004 10.3390/ELECTRONICS13245004. (2024). [ Google Scholar ] 14. Hu, S. et al. Design of intelligent nursing care products based on structural equation model. Mech. Des. 38 (7), 110–117. 10.13841/j.cnki.jxsj.2021.07.019 (2021). [ Google Scholar ] 15. Shan, H., Qi, J., Linlin, D. & Jialin, H. Guo Weiqi. An evaluation method for product design solutions for healthy aging companionship. Front. Public. Health . 10 , 919300–919300. 10.3389/FPUBH.2022.919300 (2022). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 16. Xie, W. & Xu, C. Research on the design of smart kitchen appliances for the elderly based on KJ-AHP-VIKOR integration method. Packaging Eng. 45 (22), 411–419. 10.19554/j.cnki.1001-3563.2024.22.040 (2024). [ Google Scholar ] 17. Zhang, W., Huang, L. & Wang, Z. Shijian Cang. A user-centered framework for designing smart mobility vehicles for older adults: An integrated RST-QFD-GRA approach. Expert Syst. Appl. 300 , 130430–130430. 10.1016/J.ESWA.2025.130430 (2026). [ Google Scholar ] 18. Rui Wang, W. Y., Ding, Z. Y. & Zhou Xiao Cheng Li & Yi Fan Zhang. Optimizing equipment requirements and configuration rules for elderly home treatment environments: a rough set analysis framework. Front. Med. 12 , 1700646–1700646. 10.3389/FMED.2025.1700646 (2025). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 19. Xiuli Geng, Y. & Li, D. W. Qingchao Zhou. A scenario-driven sustainable product and service system design for elderly nursing based on QFD. Adv. Eng. Inform. 60 , 102368. 10.1016/J.AEI.2024.102368 (2024). [ Google Scholar ] 20. Wang, L. Theory and algorithm of Analytical Network Process (ANP). Syst. Eng. Theory Pract. 3 , 44–50 (2001). [ Google Scholar ] 21. He, C. & Wang, Y. Review of the research on analytic network hierarchy process. Sci. Technol. Manage. Res. 34 (3), 204–208. 10.3969/j.issn.1000-7695.2014.03.044 (2014). [ Google Scholar ] 22. Fargnoli, M. & Haber, N. A practical ANP-QFD methodology for dealing with requirements’ inner dependency in PSS development. Comput. Ind. Eng. 127 , 76–90. 10.1016/j.cie.2018.10.007 (2019). [ Google Scholar ] 23. Sun, X., Guo, X., Zhang, Y. & Du, D. An automotive human-machine interface design method integrating Fuzzy Kano-QFD and physiological data. Ergonomics 66 (1–2), 21–22. 10.1080/00140139.2025.1233456 (2025). [ DOI ] [ PubMed ] [ Google Scholar ] 24. Guo, P., Ning, S. & Zhong, X. Research in design of ENT treatment unit based on AHP-QFD-FBS. Int. J. Biology Life Sci. 10 (2), 114–122. 10.5281/zenodo.1234567 (2025). [ Google Scholar ] 25. Li, N., Wang, J. & Wei, H. The application of flow–clutch states and AHP-QFD-FBS in the design of parent–child interaction exercise bikes. Appl. Sci. 15 (6), 3270. 10.3390/app15063270 (2025). [ Google Scholar ] 26. Miao, Y., Xie, X., Wang, H. & Xu, W. A study on the design of literacy toy for children with parent-child interactions. Sci. Rep. 15 (1), 6793. 10.1038/s41598-025-98765-2 (2025). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 27. Hu, R., Tao, X., Gong, J. & Wang, F. Quality function deployment approach to urban ecological public art design centred on resident needs. Sci. Rep. 14 (1), 22814. 10.1038/s41598-024-53465-7 (2024). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Gero, J. S. Design prototypes: A knowledge representation schema for design. AI Magazine . 11 (4), 26–36 (1990). [ Google Scholar ] 29. Cascini, G., Fantoni, G. & Montagna, F. Situating needs and requirements in the FBS framework. Des. Stud. 34 (5), 636–662. 10.1016/j.destud.2013.04.003 (2013). [ Google Scholar ] 30. Rasoulivalajoozi, M. & Farhoudi, M. Integrating user perceptions of socio-emotional aspects in wheelchair design: A pilot study using Kansei Engineering. J. Transp. Health . 42 , 102002. 10.1016/j.jth.2025.102002 (2025). [ Google Scholar ] 31. Wang, Z., Yang, K., Wang, Y., Zhu, Z. & Liang, X. Embrace the era of drones: A new practical design approach to emergency rescue drones. Appl. Sci. 15 (1), 135. 10.3390/app15010135 (2024). [ Google Scholar ] 32. Lee, C. K. M., Tsang, Y. P., Chong, W. W., Au, Y. S. & Liang, J. Y. Achieving eco-innovative smart glass design with the integration of opinion mining, QFD and TRIZ. Sci. Rep. 14 (1), 9822. 10.1038/s41598-024-60483-y (2024). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 33. Bangor, A., Kortum, P. T. & Miller, J. T. An empirical evaluation of the System Usability Scale. Int. J. Hum Comput Interact. 24 (6), 574–594. 10.1080/10447310802205776 (2008). [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplementary Material 1 (1.4MB, pdf) Data Availability Statement The datasets used and/or analysed during the current study available from the corresponding author on reasonable request. Articles from Scientific Reports are provided here courtesy of Nature Publishing Group ACTIONS View on publisher site PDF (2.7 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top