The attitudes and acceptance of functional assistive robots among older adults with disabilities: a mixed-methods study

Abstract

Background and Objectives

The growth in older adults living with disabilities and the decline in the workforce call for technology-driven nursing services to alleviate the burden. Functional assistive robots have emerged as a promising solution; however, their acceptance and attitude toward them remain underexplored. Therefore, this study aimed to investigate the attitudes and requirements of older adults living with disabilities regarding functional assistive robots.

Research Design and Methods

A mixed-methods study was conducted between November 2023 and January 2024, comprising a cross-sectional survey of 85 older adults with varying levels of disability and semi-structured interviews with 10 participants. Quantitative data were analyzed descriptively, and qualitative data were thematically analyzed to enrich and contextualize the findings.

Results

Fifty-three percent of participants expressed willingness to use a functional assistive robot, with safety, affordability, ease of use, and multifunctionality identified as the most influential factors. Walking assistance and toilet transfer were the most desired functions across all disability levels. Qualitative findings reinforced the quantitative findings, revealing dissatisfaction with current aids due to their limited stability and maneuverability, alongside a strong desire for increased independence and reduced strain on family caregivers. Attitudes toward robots ranged from enthusiasm to skepticism, influenced by perceived usefulness, disability level, and care context. Participants emphasized the need for compact, stable designs with simple interfaces, and some expressed interest in light social features, provided they did not compromise core functionality.

Discussion and Implications

This study highlights the demand for functional assistive robots that address critical mobility-related tasks. Functional assistive robots are viewed as promising, but adoption hinges on improving safety, usability, and affordability. Multifunctional, user-friendly designs are essential for practical use. The findings offer guidance for nurses, caregivers, and developers by clarifying core functional needs and concerns, supporting the creation of acceptable technologies that enhance autonomy and quality of life.

Innovation and Translational Significance.

The growth in older adults living with disabilities and the decline in the workforce in China call for novel assistive solutions. This mixed-methods study revealed that while older adults express cautious optimism about functional assistive robots, acceptance hinges on safety, affordability, and ease of operation. Key desired functions include walking assistance and toilet transfers. Translating these insights into user-centered design can enhance autonomy, reduce caregiver strain, and improve overall quality of life. By prioritizing practical features, such as compact design and simple interfaces, developers and caregivers could collaboratively develop effective and cost-efficient assistive robots for aging societies.

Background and objectives

Aging populations are a global trend due to rising life expectancy and declining fertility rates (World Health Organization, 2024). An increasing number of older adults become disabled due to aging, illness, accidents, and other factors, necessitating reliance on others for daily activities (Ramadhani & Rogers, 2022). Data from the World Health Organization (WHO) show that approximately 15.6% of the world’s population has varying degrees of disability, and the disability rate increases significantly with age (World Health Organization, 2018). In China, there are currently about 44 million people living with disability (Du & Wang, 2023); along with this comes substantial care needs. However, as the working-age population declines and family structures tend to become smaller, the shortage of geriatric care workers is widening, which cannot meet the nursing needs of older adults and requires technology-driven nursing services (Sato et al., 2020).

With the healthcare industry gradually entering the Fourth Industrial Age, assistive robots for older adults with disabilities are now widely regarded as an emerging, viable, effective, and promising strategy to increase independence and enhance their well-being and also save the care workforce and reduce caregivers’ workload (Abdi et al., 2018; Castro E Melo & Faria Araújo, 2020; Kastner et al., 2024; Kodate et al., 2022; Tan & Taeihagh, 2020; van Wynsberghe, 2015). The assistive robots developed for older adults are generally categorized into three types (Sharkey & Sharkey, 2012): (1) robots for older adults with physical or functional deficiencies to assist with performing daily activities, such as walking, transferring, and feeding (Jeong et al., 2021; Li et al., 2019; Liu et al., 2022; Papageorgiou et al., 2014); (2) social robots provide companionship for lonely or cognitively impaired older adults, such as Paro, Pepper, and Kompai (Caleb-Solly et al., 2018; Klein et al., 2013; Kolstad et al., 2020; Zsiga et al., 2018); and (3) robots used to monitor and supervise the health and safety status of older adults, reminding them of routine activities and monitoring emergency events or situations (Sandhu et al., 2024). The first type of assistive robot should be applied to help older adults living with disabilities perform daily activities as independently as possible. This not only improves their quality of life and preserves their dignity in aging but also significantly alleviates the burden on caregivers (Güttler et al., 2015).

Despite the growing interest in developing this type of assistive robot, most remain at the development or laboratory testing stages, with limited implementation in real-world settings (Ohneberg et al., 2023). Recent studies have already examined older adults’ preferences for domestic‑task or “service” robots (Lee et al., 2024; Sawik et al., 2023) and have reported generally positive—but function‑specific—acceptance of robots that support activities of daily living (ADL), such as meal preparation or light housekeeping (Hall et al., 2019). Qualitative investigations (Gasteiger et al., 2022) further showed that experiential factors (e.g., safety, reliability, appearance) and the interaction between older adults and their human caregivers jointly shaped acceptance. Nevertheless, most prior work has focused on cognitively impaired older adults or on social companion robots (Koh et al., 2021; Pino et al., 2015; Wu et al., 2014), whereas substantially fewer in‑depth studies have explored physically or functionally disabled older adults whose primary needs revolve around bodily assistance with mobility, transfers, toileting, and other hands‑on ADL tasks. Moreover, very little is known about how the design of such functional assistive robots might align with China’s emerging long-term care insurance framework, in which both older adults and caregivers may have opportunity to interact with these devices. Understanding the attitudes and requirements of this specific user group is therefore critical; mismatches between robot functions and user needs remain a key barrier to the adoption of robots in real-world settings (Koh et al., 2021).

This study formed part of a larger engineering program aimed at developing an assistive robot capable of enhancing the independence of older adults with physical or functional disabilities, whether at home or in long-term care institutions (LTCIs). The primary objective of this study was to elicit the perceptions, functional priorities, and design preferences of older adults regarding functional assistive robots, thereby ensuring that subsequent prototypes align with user needs through a combination of quantitative and qualitative research methods. The secondary objective was to document participants’ current utilization of traditional assistive tools so that the strengths and limitations of existing solutions can directly inform robot specification. Together, the findings would guide the development of a practical robotic aid aimed at improving everyday autonomy for this population.

Research design and methods

Study design

This study employed a mixed-methods design, combining a cross-sectional survey and semi-structured interviews. The cross-sectional survey was conducted first to collect quantitative data on the current usage of assistive devices and the acceptance and preferences of older adults living with disabilities for assistive robots, followed by semi-structured interviews to gather qualitative insights on these topics. This study was approved by the Institutional Review Boards of Sichuan University West China Hospital (Approval number: 2022-852).

Participants and sample size calculation

Participants for the cross-sectional survey were recruited using a convenience sampling method. The inclusion criteria for participants were (1) aged 60 years or above; (2) had a Barthel Index (BI) of less than 100 points; and (3) participated voluntarily in this study. Participants who were unable to communicate were excluded. Given the exploratory nature of this study, a formal sample size calculation was not conducted before data collection. The sample size was estimated based on similar studies in the field, which typically used sample sizes ranging from 15 to 55 participants. As such, a target of 100 participants was set, resulting in an initial total of 109 participants with 85 valid responses.

Participants for the semi-structured interviews were selected using the purposive sampling method, taking into account maximum variability in factors, such as age, gender, education level, living setting, and disability level. Individuals aged 60 and older who could communicate with interviewers were included. The sample size for the semi-structured interviews was determined based on the principle of theoretical saturation. After conducting ten interviews, no new themes or views emerged from the data. All participants in this study signed the informed consent form.

Definition and classification of disability

Disability is commonly defined as a condition in which individuals, due to aging, illness, or injury, experience partial or complete loss of certain body functions, leading to limitations or an inability to perform daily activities (Granger et al., 1979). The severity of disability was most commonly assessed using the BI. A total BI score of 100 indicated that the individual was independent of assistance from others, while a score of less than 100 meant that the individual needed assistance from others to complete daily activities (Shah et al., 1989). In this study, we adopted the disability classification system issued by the National Healthcare Security Administration of China, which is used in the implementation of the national long-term care insurance system. Based on this system, a total BI score of 60–95 suggests mild disability, 40–60 moderate disability, and ≤40 severe disability (National Healthcare Security Administration, 2021).

Data collection

Cross-sectional survey

The survey was conducted in settings of LTCI and communities between November 1, 2023 and January 31, 2024. A structured questionnaire was developed in four sequential steps. First, we reviewed the relevant studies on care assistive robots, assistive‑device utilization, and relevant theories on technology acceptance (e.g., the Technology Acceptance Model [TAM], the Unified Theory of Acceptance and Use of Technology [UTAUT], and the Matching Person and Technology [MPT], all of which highlight perceived usefulness, ease of use, facilitating conditions, and person–technology fit as key determinants of assistive‑technology adoption), to construct the survey item pool. Subsequently, we invited two geriatric nurses, one rehabilitation physician, one health service researcher, and two assistive robot engineers to evaluate the initial item pool for relevance, clarity, and practicality. Then, a pilot survey was conducted with ten older adults having varying BI scores to investigate whether the question was ambiguous and whether the older adults could understand it. Finally, the questionnaire comprised four sections, including (1) socio-demographics (age, gender, education level, living condition, living setting, care expenses per month, and caregiver type), (2) disability level (evaluated with the BI), (3) current use of assistive tools (type, number, cost, and purchase-decision factors), and (4) attitude and requirements for functional assistive robots, which were organized along the TAM and the UTAUT dimensions, that was, overall attitude, willingness to pay, perceived usefulness (task priorities), perceived ease of use, facilitating conditions, and design‑feature requirements. Prior to the block of questions on assistive robots, a standardized one‑sentence description was stated: Imagine an assistive robot that can assist you with transfers (bed‑to‑chair, chair‑to‑toilet), support walking, help with bathing, toileting, going up and down stairs, monitor activities, navigate, and so on. The data were collected face to face by trained investigators to ensure the accuracy and completeness of the information.

Semi-structured interview

The semi-structured interview guide (Table 1) was constructed based on three complementary frameworks: the International Classification of Functioning, Disability and Health (ICF), the MPT, and the TAM, which emphasizes perceived usefulness, ease of use, and behavioral intention. Accordingly, the guide was organized along three axes: (1) Daily‑life challenges (ICF: activity limitations): questions probing concrete difficulties in ADL tasks (Q1). (2) Current assistive solutions (MPT: person–technology fit): questions on use, advantages, and shortcomings of traditional assistive devices (Q2). (3) Attitudes toward and requirements for functional assistive robots (TAM + MPT): questions exploring perceived usefulness, concerns, willingness to use/pay, and desired functions or design features (Q3–Q5). The draft guide was reviewed by a multidisciplinary panel mentioned above and adjusted after two pilot interviews to improve clarity and flow.

Table 1.

Questions used to guide the semi-structured interview.

No.	Questions
Q1	What difficulties do you have in performing activities of daily living (e.g., eating, walking, bathing, washing up, using the toilet, up and down stairs, etc.)? How do you cope with these difficulties?
Q2	Do you use any assistive tools to cope with these difficulties? How do you feel about the assistive tools (e.g., walker, handrails, cane) you used? Have you encountered any difficulties or inconveniences when using these tools? What changes would you make if you could improve any of these tools?
Q3	Studies indicated that nursing robots are useful in assisting individuals with disabilities in performing daily activities, such as walking, transferring from bed to chair or toilet, rehabilitation, etc. What do you think about this idea?
Q4	If given the opportunity to use a robot, would you be willing to try it? Why or why not? Do you have any concerns or worries about using a robot? (e.g., safety, privacy, complexity of the technology)
Q5	Do you have any preferences about the design or function of the robot? What do you think about an assistive robot with the following functions: feeding assistance, walking assistance, stair assistance, bed-chair transferring, toilet transferring, activity monitoring, intelligent navigation, and voice control? Which functions do you think are most important and useful in your daily life?

Prior to the interview, each participant was shown a single-page color schematic of our laboratory prototype (Supplementary Figure 1), while the interviewer verbally explained its functions. The prototype was described as a mobile assistive robot for ADL support, incorporating five modules: (1) Powered transfer assistance, including bed-to-chair, chair-to-toilet, and chair-to-bath transfers. (2) Multi‑posture transformation, including seated, reclining, and supported‑standing positions. (3) Walking aid with intelligent following and motion monitoring. (4) Short‑flight stair‑climbing mechanism. (5) Voice‑control and navigation for safe indoor and outdoor movement. The device was explicitly framed as an ADL aid, not a medical treatment machine. Two trained researchers conducted the interviews in participants’ residences or facility rooms; each session lasted at least 30 min and was audio-recorded with the participant’s consent. Recordings were transcribed verbatim for analysis.

Data analysis

The data analysis was performed using SPSS 26.0 and NVivo 12 software. Participants’ characteristics were described using means, standard deviations (SD), frequencies, or percentages. The qualitative data from the semi-structured interviews were analyzed inductively, guided by Braun and Clarke’s six-step thematic analysis framework (Braun & Clarke, 2006). In step 1 (Familiarization), both coders (author Y. Wang: an engineering PhD with 5 years of experience in developing rehabilitation robots for individuals with disabilities; author H. Chen: a PhD in geriatric nursing with experience in long‑term care for older adults with disabilities) read each transcript twice and recorded reflexive notes. In step 2 (Initial coding), each coder performed line-by-line, data-driven coding in NVivo 12, maintaining an audit trail of their decisions. In step 3 (Comparing codes and reconciliation), the initial code lists were compared, and minor discrepancies were discussed and resolved by consensus. The shared codebook was then refined. In step 4 (Theme development), the coders jointly grouped related codes into candidate themes, iteratively reviewing these against the entire data set for coherence and distinctness. In step 5 (Peer debriefing), a senior qualitative researcher audited the evolving theme map at two checkpoints, providing critical feedback that was incorporated into subsequent refinements. In step 6 (Credibility checks), a succinct thematic summary was returned to three participants (with mild, moderate, and severe disabilities) for member checking; their feedback led to minor wording adjustments. Theme salience was reported descriptively as the number of participants endorsing each theme. Before analysis, they conducted a reflexive discussion to articulate potential biases (e.g., Coder A’s positive expectations regarding technology and Coder B’s emphasis on clinical feasibility) and agreed to bracket these views during coding.

Results

Participants characteristics

In the cross-sectional survey, a total of 85 valid questionnaires were collected. Most participants were 75 years or older (68.3%). Most had a secondary school education (48.2%) or higher, with 29.4% holding a college or bachelor’s degree. Approximately half of the participants (52.9%) lived at home, while 47.1% resided in LTCI. Primary caregivers were predominantly children (49.4%), followed by spouses (37.6%) and care workers or nannies (12.9%). For ADL, 37.7%, 41.2%, and 21.1% of participants were mild, moderate, and severely disabled. Participants most commonly required assistance with stairs (67.1%), bathing (55.3%), and toilet use (47.1%). More details are presented in Table 2.

Table 2.

The basic characteristics of participants in the cross-sectional survey.

Characteristics	n (%)
Age (years)
60–64	8 (9.4)
65–74	19 (22.4)
75–84	32 (37.7)
≥85	26 (30.6)
Gender
Female	41 (48.2)
Male	44 (51.8)
Education background
Primary school or illiteracy	11 (12.9)
Secondary school	41 (48.2)
College or bachelor's degree	25 (29.4)
Postgraduate degree	8 (9.4)
Living environment
The non-ground floor of walk-up buildings	30 (35.3)
The ground floor of walk-up buildings	29 (34.1)
Building with elevator	26 (30.6)
Living setting
Long-term care institutions	40 (47.1)
Home	45 (52.9)
Care expenses per month (¥)
<1,000	12 (14.1)
1,000–2,000	9 (10.6)
2,000–3,000	21 (24.7)
3,000–4,000	14 (16.5)
4,000–5,000	3 (3.5)
>5,000	5 (5.9)
Unclear	21 (24.7)
Primary caregiver
Care worker/Nanny	11 (12.9)
Spouse	32 (37.6)
Children	42 (49.4)
ADL requiring assistance
Stairs	57 (67.1)
Bathing	47 (55.3)
Toilet use	40 (47.1)
Mobility (on level ground)	30 (35.3)
Transfers (bed to chair)	25 (29.4)
Grooming	13 (15.3)
Dressing	12 (14.1)
Bladder control	11 (12.9)
Bowel control	10 (11.8)
Feeding	9 (10.6)

Note. ADL = activities of daily living.

Ten participants were involved in the semi-structured interviews. Five resided in LTCI, with three cared for by a care worker and two cared for by family members (e.g., spouse, children). The other five lived at home, with four cared for by family members and one cared for by a care worker. The average age of the participants was 82.5 years, with six males and four females. Four participants had mild disabilities, four had moderate disabilities, and two had severe disabilities.

Results of the cross-sectional survey

Usage of assistive tools

The most commonly used assistive tools were walking sticks (69.4%) and wheelchairs (60.0%). About half of the participants (49.4%) spent more than 5,000 yuan on buying assistive tools. When choosing assistive tools, the top four most important factors that more than half of the participants considered were safety (81.2%), ease of operation (61.2%), ease of carrying (52.9%), and a reasonable price (51.8%, Figure 1).

Figure 1.

Current use, cost, and purchase considerations of assistive tools among participants.

Note. All 85 participants answered every item; numbers on the bars show the absolute count, and percentages are calculated as count / 85. Cost ranges are in thousand Chinese yuan (CNY).

Attitude and requirements for functional assistive robots

The responses showed a mix of openness and reluctance. Approximately half of the participants expressed a positive attitude toward using assistive robots to assist with daily tasks. As for the money they would like to spend on assistive robots, the most common spending range was between 5,000 and 10,000 yuan, and nearly 40% of them would like to spend more than 10,000 yuan on it. The top four most important factors that affected participants’ choice of the assistive robot were safety, reasonable price, ease of operation, and multifunction (Figure 2), similar to the results for assistive tools.

Figure 2.

Participants’ attitudes, willingness‑to‑pay, and decision factors regarding assistive robots.

As for the requirements, their primary expectations for functional assistive robots included assistance with walking, transitioning between different scenarios, assistance with climbing stairs, and activity monitoring and protection (Figure 3). Other required functions included voice control, smart navigation, and bath transfer. When analyzing the needs of older adults based on their levels of disability, we found that while functional requirements varied, there was a common demand for toilet transfer and walking assistance (Figure 3). This suggests that in designing assistive robots, these two functions should serve as core modules to address fundamental needs across all disability levels. Additional features could then be layered according to older adults’ needs and disability level.

Figure 3.

Overview of the importance of participants’ functional requirements for assistive robots.

Results of semi-structured interviews

The thematic analysis revealed four themes concerning participants’ perceptions and attitudes toward traditional assistive tools and functional assistive robots: mobility challenges and limitations of assistive tools, the desire for increased independence and reduced caregiver burden, positive and negative attitudes toward functional assistive robots, and preferences and requirements for functional assistive robots.

Mobility challenges and assistive tools limitations

All 10 respondents reported mobility problems, necessitating assistive tools, such as wheelchairs, canes, and walkers. Detailed cross‑tabulation of disability level and primary mobility aid is presented in Supplementary Table 1, showing a stepwise shift from canes to wheelchairs as BI scores decrease. However, respondents reported significant limitations with current mobility aids, particularly mentioning the bulkiness, heaviness, and maneuverability problems of wheelchairs. These issues make them inconvenient in confined spaces and complicate transfers. Canes and walkers were frequently described as insufficiently stable for long distances and overly cumbersome for daily use, especially among individuals with weakened limbs or arthritis. They were preferred for lightweight and portable assistive tools that provide enhanced stability and support. Additionally, respondents mentioned the need for designs that reduce caregiver strain, offer improved body support, and seamlessly integrate into daily routines without attracting undue attention.

Walking is hard for me because of my left leg…it’s so weak…I use a cane, but it’s still a struggle…I don’t use the walker, they are too clunky. (Respondent 4, moderate disability)

I rely on a wheelchair to sit, but the chair isn’t well-designed…it’s hard to keep me upright, and I always slide down… My caregivers do their best, but it’s a lot of work. (Respondent 7, severe disability)

Desire for increased independence and reduced caregiver burden

Participants’ attitudes toward independence varied by disability level. Those with milder disabilities more often expressed a desire to regain autonomy and reduce caregiver burden, while participants with more severe disabilities tended to be more skeptical of such possibilities.

Five of the ten interviewees (three of the four mild‑disability and two of the four moderate‑disability participants; four of them cared for by family members) stated that “living more independently” was a major goal; two of them said they would consider paying up to 50,000 CNY for a robot that reduced caregiver workload. In contrast, the other five interviewees (two respondents with severe disabilities, two with moderate disabilities, and one with mild disabilities; three of whom were cared for by the care worker) felt that their physical condition “could hardly improve” and were therefore skeptical about any technology promising independence. Desired independence was closely linked to attitudes toward robotic aids: four of the former five optimistic respondents also expressed “willing” to use the assistive robot in the future and prioritized toilet transfer and walking assistance, whereas the latter five resigned respondents stated “neutral,” “unwilling” to use the assistive robot in the future and placed little value on robotic solutions. This pattern suggests that greater functional capacity (higher BI scores) is associated with higher openness to functional assistive robots. In contrast, the most dependent participants remain cautious due to concerns about reliability and their existing reliance on human care.

She did so much for me already, and I felt guilty about how hard it was for her…I think it could be a good way to share the burden…help with the more difficult tasks. (Respondent 1, moderate disability)

My kids and grandkids help me when they have spare time, but I know they have their own lives and responsibilities…I wouldn’t want to be a burden on my family. (Respondent 2, mild disability)

My caregivers already take care of everything for me. They help me with transfers, bathing, and all the things I can’t do on my own. So, I don’t need a robot, this will increase my expenses. (Respondent 7, severe disability)

If the robot that you invented could help me stand up or walk safely, and reduced the burden on my wife because her back was not very good, then we should be willing to spend money to buy one, if the price of this robot was less than 50,000. (Respondent 9, mild disability)

Positive and negative attitudes toward functional assistive robots

The responses from respondents revealed a range of perspectives regarding the potential use of functional assistive robots. Four participants (two with mild disability, two with moderate disability) viewed robotic assistance positively, believing that rapid technological advancements could significantly improve their daily lives. They particularly valued the prospect of the robot assisting with tasks, such as transfers and bathing, improving their independence and reducing their reliance on caregivers. Three respondents (two with moderate disability, one with severe disability) took a neutral stance, stating that they would need to see a live demonstration of the assistive robot before making a decision. They emphasized the importance of seeing how the robot functions in real life, expressing a willingness to try it if the demonstration was convincing but cautioning that they would not use it if the robot did not meet their expectations. The remaining three respondents (two with mild disability, one with severe disability) were pessimistic, primarily due to concerns about the robot’s cost, safety, and social isolation. Additionally, we found that attitudes differed by care context. Among the five home-dwelling participants (four of whom were cared for by family members), four were willing to adopt an assistive robot, whereas none of the five LTCI residents expressed outright willingness to do so. Respondents living at home emphasized that a robot could “share the burden on my spouse/children and improve my independence,” while LTCI residents, already receiving full care from the care worker, felt a robot might be redundant and add extra cost. Notably, the most positive attitudes occurred among mildly disabled participants cared for by family members, whereas severe-disability residents under professional care expressed the greatest reluctance, with moderate-disability respondents falling in between. This suggests a possible interaction between disability level and caregiver type in shaping robot acceptance.

Positive attitudes:

I would like to try a robot, especially if it could take some of the strain off my spouse. (Respondent 1, moderate disability)

A robot sounds like it could be very useful, especially for walking and rehabilitation exercises. (Respondent 6, moderate disability)

Negative attitudes:

Robots feel impersonal to me, and I think I’d feel even lonelier using one. (Respondent 5, mild disability)

I don’t think a robot would work for me. I already have caregivers who know how to handle my needs. (Respondent 8, severe disability)

I like to stay active and independent, and I worry that using a robot might make me too reliant on it. (Respondent 10, mild disability)

Neutral attitudes:

I don’t know if I really need a robot, as I do not see a real robot and I do not know how it works…I’d only consider it if it could do something my caregiver can’t. (Respondent 3, moderate disability)

I could see myself trying a robot for certain things, but I’d have to be cautious about its safety…I’m worried it may make me fall. (Respondent 4, moderate disability)

I suppose a robot might be helpful in some ways, but I don’t see why I’d need one, as my caregivers already take care of everything for me…. I am wondering whether the functions you mentioned will really come true. The ideal is good, but you know, it is always hard to come true…. (Respondent 7, severe disability)

Preferences and requirements for functional assistive robots

Most respondents (n = 5) prioritized bed‑to‑chair and bed/chair‑to‑toilet transfer functions as these were the most difficult daily tasks, followed by indoor walking support (mentioned by four). The outdoor-oriented modules, that is, stair climbing, intelligent navigation, auto-follow, and motion monitoring, were generally viewed as advanced and practical. However, they might be risky in actual application due to factors, such as uneven pavements, weather, and the fear that “no one would be around if it malfunctioned.” Accordingly, outdoor capability was rated as “nice to have” rather than essential.  Across all contexts, respondents stressed a compact, stable, and user-friendly design with simple voice or push-button controls.

Although the interview guide centered on physical assistance, three respondents (all with mild or moderate disabilities) spontaneously mentioned that they would like the robot to “talk,” “greet,” or “remind” them, provided it did not slow down functional tasks. Conversely, four respondents (particularly those living in LTCI) worried that a functional robot might reduce human contact: “If the robot does everything, the staff may come in less often.” Two respondents suggested a compromise: a “task‑first” operation combined with simple voice feedback so the device felt “less cold.” Overall, emotional features were viewed as desirable only if they did not undermine the primary goal of physical assistance. It was noteworthy that only participants with mild or moderate disabilities mentioned optional social features, whereas those with moderate to severe disabilities, particularly in institutional care, focused exclusively on core mobility functions.

I think the most important functions for me would be bed-chair transferring and toilet transferring…Voice control could be nice. (Respondent 1, moderate disability)

I would prioritise walking assistance and bed-chair transferring…As for the design, I’d like something stable and gentle. (Respondent 3, moderate disability)

I’d want the robot to be simple and not too bulky. It should work quietly and not be intimidating. (Respondent 7, severe disability)

If the robot could keep me steady while I walk and let me know if I’m about to fall, it would make me feel safer. (Respondent 10, mild disability)

Discussion and implications

This study utilized a mixed-methods approach to comprehensively examine the functional limitations faced by older adults living with disabilities, their perceptions of traditional assistive tools, and their attitudes and functional requirements regarding functional assistive robots. The findings identified key daily challenges faced by older adults living with disabilities, including difficulties with stairs, bathing, toilet use, and walking. Walking assistance (71.8%) and toilet transfer (65.9%) were the most frequently demanded functions across all disability levels. This quantitative pattern was echoed in the qualitative interviews, where participants repeatedly emphasized the physical burden of walking and toileting, often linking these to their desire for robotic assistance. These findings underscore the importance of assistive robots prioritizing core mobility-related tasks. Among traditional assistive tools, walking aids were the most commonly used; however, these tools exhibited several limitations, failing to meet the needs of older adults living with disabilities. Approximately half of the participants expressed a positive attitude about the potential of functional assistive robots, although the majority remained hesitant to adopt such technologies. The primary concerns influencing their reluctance included safety, cost, and ease of use.

Our finding on daily challenges faced by older adults living with disabilities was consistent with previous studies, which identified that mobility, bathing, transferring, dressing, and toileting were frequently and firstly impaired among older adults, with feeding being the last to decline (Cohen-Mansfield & Jensen, 2005; Jagger et al., 2001; Startzell et al., 2000). This alignment indicates a persistent pattern in the progression of functional decline among older adults, reinforcing the critical need for interventions targeting mobility and personal hygiene tasks. In terms of traditional assistive tool usage, prior surveys have consistently reported that walking aids are the most prevalent. However, users frequently cite issues related to bulkiness, limited maneuverability, and insufficient support (Godilano et al., 2018), which echo the concerns expressed by our participants. This consistency emphasizes the persistent gaps in current assistive technologies and reinforces the need for more effective solutions tailored to the specific needs of older adults living with disabilities.

Compared with earlier work on social robots, the present sample exhibited a higher overall acceptance of functional assistive robots. For example, Wu et al. reported low intention to use after one month of daily interaction with a social robot among older adults with mild cognitive impairment (Wu et al., 2014), whereas 4 of 10 interviewees in the semi-structured interview and 53.0% of respondents in the cross-sectional survey were willing to adopt a functional robot in the current study. Three factors may explain this discrepancy. First, our participants had a relatively high educational level, a variable that has been shown to increase technology acceptance. Second, perceived usefulness, the strongest predictor of intention (Wu et al., 2016), was foregrounded in our prototype. Transfer assistance, walking support, and toileting aid address immediate ADL needs, whereas social robots often target higher‑order emotional needs.  Third, unlike prior studies that have older adults interact with a working robot, we presented only a picture and functional description; actual hands-on experience, especially with early-stage hardware, can lower acceptance by exposing usability frustrations. Future studies should re‑evaluate attitudes once a fully operational prototype is available. Social considerations nevertheless surfaced. Three participants with mild or moderate disabilities requested simple social features, such as voice greetings and reminders, “as long as it doesn’t slow the real work.” At the same time, four LTCI residents expressed concern that a robot might reduce human contact, echoing Eftring and Frennert (2016). These findings suggest that a “task‑first, light‑social” design could deliver tangible utility without undermining desired human interaction.

Meanwhile, approximately half of the participants expressed negative attitudes toward adopting functional assistive robots, mainly due to concerns about safety, cost, and the complexity of operating the robots, while concerns about the brand, appearance, and social acceptance were less prevalent. Participants in the semi-structured interview expressed concern that using robots would reduce interaction with people. These results are partially consistent with findings from other studies (Eftring & Frennert, 2016; Esmaeilzadeh & Maddah, 2024; Koh et al., 2021; Y. H. Wu et al., 2014). In addition, Koh et al. (Koh et al., 2021) performed a systematic review to summarize the factors that hindered older adults from using assistive robots and found that in addition to the factors mentioned above, factors such as poor user interface or interaction, less audibility, physical inaccessibility, and intrusiveness or privacy would also affect older adults’ intention to use assistive robots. However, these findings were drawn from studies on social robots and older adults with cognitive impairment. There is a lack of studies exploring the reasons or concerns of negative attitudes toward functional assistance robots. To deepen this understanding in our own data, we noted that negative or neutral attitudes were particularly prevalent among participants with severer disabilities living in institutional care (always cared for by care workers or nursing assistants), whereas those with milder impairments under family care (always cared for by family members) tended to express greater openness to robotic assistance. This pattern suggests a potential interaction between disability severity and the care context (or the role of the caregiver) in influencing older adults’ acceptance of functional assistive robots. However, future research with larger, stratified samples is needed to further investigate how these two factors jointly shape technology adoption decisions.

When deciding to choose a functional assistive robot, we found that the following four factors were considered most by respondents: safety, price, ease of use, and multifunctionality. Regarding safety, participants repeatedly cited fear of falls or device malfunction as the primary barrier to using the device. Accordingly, any mobility module would include redundant braking, low‑speed collision detection, and a “safe‑stop” fail‑safe. Regarding affordability, most commercially available assistive robots with advanced features, such as exoskeletons or companion robots like Paro, are priced above 50,000 CNY, which significantly exceeds the price range acceptable to most of our respondents. In our survey and interviews, the majority of older adults indicated an affordable threshold of around 10,000 CNY, with only a few willing to consider prices above 30,000 CNY. This discrepancy underscores a critical affordability barrier to adoption among older adults. To mitigate this constraint, future implementation strategies may include advocating for public reimbursement mechanisms (e.g., inclusion in long-term care insurance schemes), promoting rental-based access models, and facilitating the deployment of shared robots within community or institutional settings. These approaches may help reduce individual financial burdens and improve equitable access to robotic assistance.

To ensure accessibility and ease of operation, respondents emphasized the importance of intuitive control interfaces, such as a large, clearly labeled physical button and simple voice commands. However, when implementing voice control functionalities, it is essential to account for common age-related speech characteristics, such as reduced articulation, slower speech rate, and diminished vocal strength. Furthermore, the diversity of regional dialects among older adults poses an additional challenge to speech recognition accuracy. As such, voice interfaces should be designed with a high tolerance for speech variability and, where feasible, include adaptive features or customizable language settings to accommodate dialectal differences and enhance user experience. These considerations are crucial for minimizing operational barriers and fostering sustained engagement with the device.

Regarding multifunction, most assistive robots currently available or under development are designed for a single purpose, such as transfer assistance (e.g., RIBA-II), walking support (e.g., Gait Trainer), or companionship (e.g., Paro). Although task-specific robots are well-suited for addressing individual functional needs with technical precision, their use in scenarios involving multiple simultaneous care requirements, common among older adults, may introduce challenges related to cost, space utilization, and system integration (Bedaf et al., 2015). In this context, a modular multifunctional robot built on a core mobility platform (e.g., a robotic wheelchair) may offer a more scalable and personalized solution. By integrating key modules identified in the current study, such as bed-to-chair transfers, chair-to-toilet transfers, and walking assistance, the robot can be tailored to meet specific care needs without excessive redundancy. Modular architecture also facilitates ease of upgrading and cost control, as only the necessary functions are installed according to user demand (Graf et al., 2004).

Regarding robot design, our findings suggest that for functional assistive robots, appearance is not a primary concern among older adults. Participants prioritized features, such as compactness, stability, and a non-intimidating form, emphasizing functional reliability and safety over aesthetic attributes. This is consistent with a prior study (Eftring & Frennert, 2016), which also found that older adults prioritized functionality over appearance in assistive robot design. While previous studies on social robots have highlighted user preferences for anthropomorphic or animal-like appearance to foster companionship and emotional engagement, such considerations were not echoed in our data (Dino et al., 2022). Instead, participants in the current study valued soft edges, smooth movement, and simplicity in design, which they believed promoted comfort and ease of use. This distinction highlights the importance of aligning design priorities with the robot’s intended function, specifically whether it provides emotional support or physical assistance.

Finally, it is necessary to acknowledge that while the qualitative component enriched the interpretation of survey findings, its smaller sample size (n = 10) relative to the quantitative cohort (n = 85) may limit the breadth of perspectives captured. The interviews were intended to provide contextual depth rather than direct representativeness, aligning with the explanatory aims of the mixed-methods design. Future studies could extend the qualitative sampling to additional regions and care contexts to capture a broader range of views and experiences beyond the saturation achieved in the present setting.

Limitations

This study had limitations that need to be acknowledged. First, the overall sample size of this study was small, and participants were recruited using a convenience sampling method within a city in China, which may limit the generalizability and representativeness of the findings to the broader population of older adults with disabilities. The specific characteristics of the sample, such as their living environments, local culture, and levels of disability, may not fully capture the diversity within the larger population. To address this limitation, future research will involve a larger and more diverse sample once the prototype of the functional assistive robot is developed, enabling a more comprehensive assessment of the perspectives of older adults living with disabilities. Second, because participants viewed only a schematic and did not interact with a working device, their acceptance ratings may differ from reactions to a live, hands‑on demonstration; previous TAM research shows that physical interaction can increase perceived usefulness and ease of use. Finally, although this study identified older adults’ positive attitudes and specific functional preferences for assistive robots, it remains uncertain whether these stated intentions would translate into actual adoption behavior in real-life settings. The discrepancy between expressed willingness and actual usage is a well-documented phenomenon in technology acceptance research, particularly among older adults. Therefore, future studies should incorporate real-world pilot implementation to assess behavioral responses and sustained engagement over time.

Implications for practice and future research

By integrating quantitative and qualitative data across Barthel-defined disability levels and care settings, this study provided several novel insights into older adults’ perceptions of functional assistive robots. Healthcare providers and caregivers should recognize that older adults’ acceptance of functional assistive robots may vary according to their disability severity and care contexts. Thus, careful assessments should be made to identify those individuals who would most likely benefit from robotic support, particularly focusing on mildly disabled older adults in family-care contexts who exhibit higher receptiveness. Robot designers and engineers are advised to prioritize safety, affordability, and intuitive usability, such as accommodating age-related speech variability and dialect differences, and to adopt a modular approach that begins with core mobility functions while allowing user-specific customization. Furthermore, given that affordability emerged as a key concern, policy makers might explore the inclusion of assistive robots in long-term care insurance schemes, support rental or leasing models, or facilitate community-based sharing initiatives to improve accessibility. Lastly, future studies could further investigate the interplay between disability severity and care contexts using larger, stratified samples and conduct longitudinal real-world studies to examine actual usage patterns and sustained engagement with functional assistive robots in both home and institutional settings.

Conclusion

This study identified mobility-related activities, particularly walking and toilet transfers, as the most prevalent and burdensome challenges among older adults with disabilities. While traditional mobility aids were widely used, many participants reported limitations related to safety, maneuverability, and usability, indicating a gap between existing tools and user needs. Approximately half of the respondents expressed positive attitudes toward functional assistive robots, particularly when they were perceived as tools to reduce caregiver burden and promote autonomy. Key determinants of acceptance included safety, affordability, ease of use, and multifunctionality. Notably, acceptance varied by disability level and care context, with older adults cared for by family members generally showing greater openness to robotic support than those receiving institutional care. These findings demonstrate the need for modular, user-centered robots that prioritize essential mobility functions while accommodating individual needs and economic constraints. Future development and implementation should emphasize customizable function packages, intuitive control interfaces (e.g., accommodating speech variability in older adults), and deployment models such as insurance coverage, rental, or community-based shared use to improve accessibility and adoption.

Supplementary material

Supplementary data are available at Innovation in Aging online.

Funding

This study was supported by the Youth Science Fund Project of the National Natural Science Foundation of China [grant numbers 62306195; 62303092; and 62103084]; the China Postdoctoral Science Foundation [grant numbers 2024M762215 and 2025M771952]; and the 2023 Central Government Guiding Local Science and Technology Development Fund of Mianyang Science and Technology Bureau, Sichuan Province [grant number 2023ZYDF093].

Conflict of interest

None declared.

Data availability

This study was not preregistered. The raw data used to support the findings of this study are available from the corresponding author upon request.

Author contributions

Yilin Wang and Hongxiu Chen contributed equally to this work. All authors approved the final version of the manuscript. Conceptualization: Xiuying Hu, Hong Cheng, Jing Qiu; Supervision: Xiuying Hu, Hong Cheng, Menghong Liu; Methodology: Rui Huang, Yilin Wang, Hongxiu Chen; Formal analysis: Yilin Wang, Hongxiu Chen, Chaobin Zou, Guangkui Song; Investigation: Yilin Wang, Hongxiu Chen, Qian Liu, Jiali Zhang; Writing—Original Draft: Hongxiu Chen, Yilin Wang; Writing—Review & Editing: Xiuying Hu, Hong Cheng; Funding acquisition: Yilin Wang, Chaobin Zou, Guangkui Song, Hongxiu Chen.