“I am afraid of dying alone at home” [1]; this fear, expressed by an older woman in Singapore, underscores a growing global phenomenon. As health care advances have prolonged life expectancy [2], demographic projections suggest that by 2030, 1 in 6 people worldwide will be aged ≥60 years, thus entering older adulthood. Unfortunately, many factors, including declining birth rates, changing norms, urbanization, and migration, have led to the evolution of family structures, causing a considerable proportion of these older people to live alone. In the United States, about 28% of the people aged ≥65 years live by themselves, including slightly <6 million men and slightly >10 million women [3]. Similar trends are seen across the globe, such as in Canada [4], Singapore [5], and France [6]. While living alone as an older adult is not inherently problematic and can be an indicator of robust health [7], it remains challenging for some individuals, particularly impoverished older adults who rely on government or not-for-profit organizations’ support.

In response, developments in the Internet of Things (IoT) technologies and computational data science have introduced novel digital health interventions, such as ambient assisted living (AAL) technologies. AAL technologies are a type of smart home technology that enables individuals to remain socially connected while living independently [8]. Smart homes denote residences equipped with a high-technology network, linking sensors and other IoT-based domestic devices, appliances, and features that can be remotely monitored, accessed, or controlled and provide services that respond to the needs of the inhabitants [9]. The use of AAL technologies among older adults started to gain increasing research attention around the early 2000s [10,11]. These technologies hold an innovative promise in providing support for older adults, offering tools that enhance safety, independence, and overall quality of life (QOL) for those facing challenges associated with living alone [12,13].

However, end-user adoption and continued use of AAL technologies are still very low [8]. We believe that a significant obstacle to the sustainable use of AAL technologies lies in the absence of a robust, theory-driven approach to understanding how AAL technologies become routinely integrated and normalized in their social contexts. Intervention development and evaluation require a strong theoretical foundation [14]. Because social distancing protocols implemented during the COVID-19 pandemic have given impetus to an already growing trend in the adoption of AAL technologies [15,16], we believe that work theorizing the processes by which AAL technologies become routinely embedded in the care of older adults is both timely and relevant. Thus far, 3 distinct bodies of literature on AAL technologies have emerged: technical, human, and integrative perspectives [17]. The technical perspective primarily addresses design challenges and the implementation of AAL technologies. In contrast, the human perspective concerns the sociopsychological factors influencing adoption and shaping the use of AAL technologies.

While a great deal of good work has been done, recent research still highlights a lack of richer contextual and long-term approaches to understanding technology use among older adults requiring assisted living [18]. Indeed, most theoretical tools researchers use in the humanistic perspective of AAL technologies imply a psychological approach focusing on older adults as a unit of analysis. This is exemplified by the use of rational choice theories and theories of individual motivation, such as the theory of planned behavior, the technology acceptance model, and self-determination theory [19-21]. Such psychologizing of AAL technologies’ use greatly curtails an understanding of how sociopsychological factors shape the routinization of these technologies in the care of older adults. It neglects that technology acceptance in this context is culturally dependent and will vary from one society to another [18,22]. Indeed, older adults cannot be studied in isolation, especially those requiring assisted living, because they live in interdependence with one another, their caregivers, and technological interventions [17,18].

Against this background, the aim of this paper is to promote the sustainable use of AAL technologies in the care of older adults by sharing a qualitative narrative and lessons learned from a longitudinal mixed methods study involving 19 older adults in Southeast Asia, a hitherto under-studied context. In doing so, we contribute to an emerging agenda of a rich sociological understanding of how older adults adopt and continue using AAL technologies by investigating an assemblage of sociotechnical linkages between different actors, including older adults, their caregivers, and the AAL technologies [18,23]. Such a focus is vital because AAL technologies can significantly impact the QOL for both older adults and caregivers if and only if they can be effectively integrated into routine practice [14].

To this end, we draw on 2 longitudinal AAL technology deployment studies to address the following research questions:

In this paper, we first review related literature and introduce the theoretical framework guiding our narrative and proposed recommendations. Next, we detail the study’s context and the specific AAL technology implemented. Finally, we share key lessons from the technology deployment process.

As mentioned earlier, a lot of great work has been done, and some researchers have also shared personal narratives and lessons learned. While these lessons learned by other researchers provide valuable insights, they lag in 3 important ways that motivate our work.

First, a few of these studies have used quantitative methods and found that several barriers, including usability concerns, a lack of perceived usefulness, and low technological efficacy, impede the routine integration of AAL technologies in older adults’ care [24-26], and indeed, these are helpful. However, richer contextual insights are difficult to capture through quantitative data [27]. Beyond their contributions to specific areas of empirical research, qualitative narratives can also serve to reorient theories. They may influence theories not only relating to individual AAL technology users but also regarding the society within which the use occurs. This is done by bringing attention to the social and cultural dynamics through which the routinization of these technologies occurs [28]. There are surely exceptions. Drawing insights from an ethnographic case study, Doyle et al [29] presented recommendations on the 4 interrelated and dynamic life cycles of independent living technology development: ethnographic inquiry, design, deployment, and ongoing evaluation. However, this differs from our objective to investigate sociopsychological and cultural dynamics that facilitate or impede the integration of AAL technologies into older adults’ care.

Second, some personal narratives lack an overarching theoretical framework to organize the key recommendations. For example, Consolvo et al [30] highlighted design insights supported by practical experiences, such as keeping the ambient sensors out of older adults’ view and providing a “human touch.” “Human touch” here refers to capturing qualitative details, such as the reasons behind specific actions or feelings, to create a more empathetic and personalized understanding of older adults’ experiences instead of focusing solely on sensor data. However, there is a lack of an overarching theory to create a complete and meaningful picture by linking their work to existing theories and previous research [31], which is understandable, as their work is design oriented and not psychosocial in nature. Still, to build a cumulative body of knowledge, personal narrative studies should be theoretically driven [32], as this will provide valuable support to the identification of patterns, themes, and connections in the narratives.

Third, only a handful of these studies were conducted in Southeast Asia. Most of those conducted in Southeast Asia used rational choice theories and quantitative approaches [9,33,34]. As indicated earlier, we believe qualitative research is more suited to understanding how AAL use evolves over time and why it evolves this way. The use of rational choice theories, with older adults being the primary data source, assumes that the decision to adopt and use a technology lies with the individual older user. However, the reality is far different because AAL technologies are better understood as complex adaptive systems, which are systems composed of interacting agents described in terms of rules [35]. Agents include the IoT components, older adults, and caregivers [35]. Such a view of AAL technologies’ use focuses attention on the fundamental mechanisms of emergence [36]; that is, the interactions between users and AAL technologies, along with their caregivers, are intricately linked and shape the overall context of AAL use [37]. Therefore, adoption, acceptance, and use are culture dependent and will vary from one society to another [18,22]. From such a perspective, the use of AAL technologies is viewed as (1) socially and culturally situated and (2) determined by the interactions of multiple actors, including older adults, their caregivers, and the AAL technologies themselves, who engage in nonlinear and, at times, chaotic ways [38].

It is recommended that personal narratives be shared within the confines of extant theoretical frameworks to better organize the lessons presented [32,38]. Therefore, we anchor our narrative and present our lessons based on normalization process theory (NPT) [14,39] because a fundamental theoretical and empirical puzzle in the complex health care interventions literature relates to how the use of these technologies can transcend conscious behavior and become part of normal routine in care activities [14,36]. NPT is a sociological theory explaining key mechanisms that facilitate or impede the implementation, embedding, and integration of new health techniques, technologies, and other complex interventions [39]. It is based on the premise that health care interventions are complex social interventions that require multiple people to work together to achieve a desired outcome [14]. It identifies 4 key mechanisms of the normalization process.

First, “coherence” pertains to target users’ sensemaking or understanding of the nature and purpose of the intervention. It is related to the notion of perceived usefulness in the information systems and human-computer interaction literature [40]. The intervention must make sense to the target users and be compatible with their values and expectations to be perceived as useful [41]. Second, “cognitive participation” means that the intervention must engage and include everyone at every health care system level, from patients and their families to health care professionals and administrators. Third, “collective action” means that the intervention needs to be integrated into routine practice and be supported by the necessary resources, infrastructure, and training. Finally, “reflexive monitoring” means that the intervention needs to be evaluated and monitored regularly to ensure that it remains effective and sustainable over time.

This study received ethics approval from the institutional review board of the National University of Singapore (A-16-349; protocol title: “City for All Ages: Elderly friendly City Services for Active and Healthy Aging”) and is registered with ClinicalTrials.gov (NCT06486935). All participants were referred to the study by their caregivers. No reimbursement was offered for participation. Informed consent was obtained from each participant prior to enrollment, including consent for the future use of their data in research. All data used for analysis were deidentified to preserve participant confidentiality.

We situated our research within the context of older adults living alone in Singapore’s subsidized public housing, relying on professional caregiving services provided by government-supported care centers, specifically senior activity centers (SACs). Currently, SACs are expanding their reach and services under the newly adopted designation of active aging centers. The SACs provide various services, including social activities, monitoring homebound older adults, befriending them, and responding to emergency alerts. As with any other care facility, a daunting challenge for SACs is a shortage of workforce. Therefore, AAL technologies were found to be an innovative promise to help older adults age in place, which can be defined as remaining in the community, with some level of independence, rather than in residential care [42].

Having described the Ubismart system, we turn to how it deviates from systems in previous literature. As mentioned, the number of studies on AAL technologies started to soar around 2015 [53], meaning that our deployment was one of the first studies to use AAL technologies in the care of older adults, with our first deployment in August 2016 (Table 2). Previous and current AAL technology deployments made use of audio or visual input. This is an increasingly pressing issue that our deployments avoided by excluding these sources of information. Privacy and fear of recordings are important aspects during deployments. From a technical perspective, we realized the concept of a “smart home in a box” by developing a system that can be deployed in any environment within 30 minutes. This was achieved through efforts to streamline sensor registration and create an easily extensible system built on semantic web principles [54,55]. In addition, our proprietary algorithms enabled noninvasive sleep tracking and could detect potential respiratory issues [56].

As outlined earlier, we had initially planned to assess QOL using the validated OPQOL instrument. However, pilot testing revealed significant difficulties. Many participants, especially those with no formal education (7/19, 37%) or only primary schooling (6/19, 32%), struggled to meaningfully differentiate between points on the Likert scale, even when the survey was translated into local dialects. These challenges posed a risk of introducing measurement error and participant disengagement.

To ensure comprehension and more meaningful self-expression, we developed a gamified version of the QOL assessment (Figure 9). Participants were given 35 coins and asked to allocate those across 6 core QOL dimensions. The root question was simplified to “How do you feel you are doing in each of these areas?” The number of coins assigned to each domain served as an intuitive proxy for participants’ self-assessed well-being in that area.

We selected 35 coins to strike a balance between simplicity and sensitivity. The number was large enough to allow participants to express meaningful variation across the 6 domains while still being cognitively manageable for older adults with limited numeracy.

Using data collected through the gamified QOL assessment tool, we conducted a paired sample, 2-tailed t tests to examine whether the installation of AAL technology led to significant changes in the distribution of perceived well-being across QOL domains. Because the total number of coins (n=35) was fixed at all time points, our focus was not on changes in overall QOL scores but rather on how participants reallocated their perceived well-being across individual dimensions. Because the assessment was simplified to the domain level, we interpreted any meaningful shifts in specific domains as indicative of potential changes in participants’ overall QOL.

When we analyzed QOL across its individual dimensions, 2 areas showed statistically significant gains. For leisure and social activities, participants reported a significant increase in scores following the intervention (before: mean 2.67, SD 2.29; after: mean 4.67, SD 1.22; P=.03), indicating enhanced social engagement and participation. Similarly, psychological and emotional well-being improved significantly (before: mean 5.44, SD 1.51; after: mean 6.78, SD 1.48; P=.01), suggesting a positive impact on emotional stability and mental wellness.

No statistically significant changes were observed in the dimensions of health, independence and freedom, social relationships, or home and neighborhood safety (all P>.05). The absence of change in “independence and freedom” was expected, given that the system was designed to be unobtrusive and did not alter participants’ physical mobility or routines. Similarly, “health” was largely outside the scope of the AAL system, which did not provide direct medical interventions. However, the lack of significant changes in “social relationships” and “home and neighborhood safety” was unexpected. These findings are further explored in the A Qualitative Narrative of the Impact of Ubismart on Older Adults’ QOL section, which offers deeper insight into how participants perceived and experienced these domains beyond what was captured through quantitative measures.

Although our primary focus was on QOL, we also explored participants’ broader experiences with the Ubismart system, including any concerns that arose during its use. As with many technological interventions, older adults expressed several reservations. Through open coding of the interview data, we identified 5 key concerns. The first related to electricity consumption, which was raised both during interviews with the research team and in conversations with SAC caregivers:

Second, privacy violation was a major concern in the beginning. The older adults suspected that the sensors might have video recording capabilities. Some took afternoon naps on their chairs instead of the bed out of fear that the bed sensor would capture it, and they might be judged for unproductivity. However, we explained to the older adults that the goal is to keep them safe and not judge their sleep patterns. In addition, privacy concerns waned over time as the older adults understood that the sensors did not have video recording capabilities. This is reflected in the following quote:

Third, physical harm was a major concern for one of the frail participants. She suspected that the sensors were the cause of her body pain:

As a result, she wanted to uninstall the sensors. After reassurance from her nurse that the pain was not attributable to the sensors and that the sensors would be beneficial in case of an emergency, the participant chose to continue using the technology:

Fourth, two participants had concerns about the inconveniences caused by technology in some ADLs, such as cleaning:

After a closer look at the sensors, we found that the bed sensor required an extension cable, which we informed the older adults could be elevated to not obstruct the cleaning. This resolved the issue. Once the aforementioned concerns were addressed, most of the participants were satisfied with the system. They seemed happy to have something new in their home. We believe that they felt special to have been chosen to participate in this study because the older adults consistently asked if the sensors could be left at their homes after the study had ended.

Fifth, fears about radiation exposure unexpectedly emerged as a key barrier to participation. Although we had addressed more common concerns, such as electricity consumption, privacy, and cleaning inconveniences, our initial planning did not anticipate that some participants might associate ambient sensors with radiation risk. During the initial briefing, one older adult who appeared to be an informal leader among peers voiced this concern and declined to proceed with installation. His decision significantly influenced others, some of whom had previously agreed to participate, leading to a noticeable drop in participation.

Motivated by the shortage of longitudinal studies on the integration of AAL technologies in older adults’ care, especially in Southeast Asia, we used an in-depth case study to explore the deployment and impact of an AAL technology, which we termed Ubismart, on older adults’ QOL. Our study revealed that AAL technologies positively impacted certain QOL dimensions for older adults, particularly enhancing psychological well-being and a sense of security. Many participants reported feeling safer with the technology, describing it as a reassuring presence in their homes. However, the impact on social relationships was mixed; while some older adults felt more cared for due to the technology, others experienced fewer in-person visits and a sense of isolation, as caregivers increasingly relied on AAL systems for monitoring.

Key lessons emerged regarding the integration and normalization of AAL technology. First, to establish coherence, it is essential to prioritize culturally grounded and relationally sensitive strategies that help older adults understand the purpose and relevance of AAL technology. Involving caregivers (ie, trusted figures in participants’ lives) in joint briefings would be especially effective in building trust and reinforcing the system’s value within existing care relationships. Second, to support cognitive engagement, it is crucial to recognize that adoption is shaped not only by formal caregivers but also by influential peer networks within the ecosystem. In our study, a single participant’s concern about radiation led others to withdraw, demonstrating how quickly skepticism can spread in tightly connected communities. Third, collective action requires sensitivity to cultural norms such as saving face. Some participants altered their behavior, such as napping on a chair, to avoid perceived judgment. Reinforcing that the system monitors for anomalies, not habits, is key to preserving data accuracy. Finally, reflexive monitoring brought attention to an important unintended consequence; while the technology promoted safety, it also inadvertently reduced social interactions for some older adults, highlighting the need for vigilance in balancing AAL technology’s benefits with ongoing human connection.

As mentioned earlier, much good work has been done on the use of AAL in the care of older adults. Similar to other projects, including HomeAssist [58] and NESTORE [59], both conducted in the European Union, our intervention supported aging in place by offering unobtrusive monitoring. Overall, our findings align with previous AAL studies that reported improved QOL among older adults [60,61]. However, our work contributes novel insights into several key areas.

First, as Taramasco et al [60] mentioned, data pertaining to QOL are scarce. They highlighted the methodological challenge of having to adjust for baseline imbalances in QOL measures before detecting a significant effect, underscoring how sensitive and complex this construct can be to capture. Our study contributes to this conversation not only by examining QOL outcomes but also by reflecting critically on the measurement process itself. We document the challenges of using standard Likert-based tools with older adults who have limited formal education and propose a culturally adapted, gamified alternative that offers a more accessible way for participants to express their perceived well-being. This approach enables us to surface meaningful shifts in specific QOL domains while also contributing to a broader methodological discussion on how best to measure QOL in underserved populations.

Second, recent works have shown that about 5 countries (ie, the United States, the United Kingdom, France, Canada, and Japan) have been leading research in AAL technologies thus far, and of these 5, only Japan has a collectivistic culture [62]. Answering the call for more research from diverse cultures [62], our study offers rare empirical insights from a subpopulation within a collectivistic culture. Specifically, we focused on older adults in Singapore who, despite living in a society that values familial interdependence, often experience social isolation due to impoverishment, limited education, and estrangement from close family. These older adults reside in heavily subsidized housing and depend primarily on caregiving staff at community-based SACs for emotional and social support. Unlike the typical AAL studies conducted in individualistic, high-resource settings, our work surfaces how the adoption and routinization of AAL technologies are mediated by unique cultural and relational dynamics, including peer influence, saving face [57], and the central role of institutional caregivers. These findings highlight the importance of designing AAL systems not only for technical robustness but also for cultural fit and relational integration within the lived caregiving ecologies of susceptible older adults.

Third, the OPQOL is one of the most widely used and validated instruments for assessing QOL in older adult populations, offering a multidimensional perspective that includes domains such as health, independence, social relationships, emotional well-being, and home and neighborhood safety [43,63]. However, our findings suggest that this latter domain, particularly in both the full [43] and brief versions [63] of OPQOL, may mask intervention-specific effects when home and neighborhood safety are treated as a single construct. Although our quantitative analysis using a gamified proxy of the OPQOL did not show a statistically significant change in the aggregated safety domain, qualitative narratives consistently revealed an increased sense of home-based safety. Participants described feeling reassured knowing help would be available during emergencies, especially at night or when living alone. In contrast, neighborhood safety was rarely mentioned and remained largely unaffected by the intervention. While the OPQOL-brief includes the item “I feel safe where I live,” this phrasing still conflates 2 distinct contexts. On the basis of our findings, we suggest that future studies consider explicitly separating this domain into “I feel safe at home” and “I feel safe in my neighborhood” to more accurately detect the localized effects of AAL interventions and avoid measurement dilution.

Taken together, these contributions reinforce the need for culturally sensitive, relationally grounded, and methodologically flexible approaches to designing and evaluating AAL technologies, particularly in diverse, underrepresented populations.

As with all research, this study is not without its limitations. We highlight 3 limitations that we believe are particularly important. First, a potential limitation of this study is that we focused on one context, Singapore. While the findings of our study may be generalizable to similar cultures (ie, collectivistic), some of our findings, especially pertaining to the relationship between older adults and caregivers, may not be applicable in individualistic cultures. However, our goal was to explore a context-specific understanding of the routinization of AAL technologies in this hitherto under-studied context. In doing so, we answer calls for research that redirects focus from general theories of technology use to context-specific theorizing [64]. Second, even though we explored our study context comprehensively, by having 19 older adults in this study, we are again unable to claim broad generalizability of the proposed recommendations. More research is needed before such claims can be made. However, we note that our sample size is comparable to similar studies with sample sizes ranging from 9 to 20 [24,29], owing to the cost of deployment. Third, we used a validated measure of QOL that is largely used in Western contexts [43]; therefore, our participants struggled to understand it, and we had to improvise and create the “gamified QOL measurement tool” as a conversation starter. Indeed, QOL may be experienced differently based on demographic differences [65].

Future research should address the aforementioned limitations to enhance the study’s robustness and applicability. First, cross-cultural studies that examine AAL technology use in both collectivistic and individualistic contexts could provide insights into the role of cultural values in shaping older adults’ and caregivers’ dynamics. This would help determine the extent to which our findings apply beyond Singapore. Second, while our sample size aligns with similar studies, larger and more diverse samples would strengthen the generalizability of our recommendations. Future studies could partner with health care organizations to increase sample size and diversity. Finally, developing culturally responsive QOL measurement tools would mitigate the challenges we encountered with Western-centric QOL assessments, allowing a more accurate understanding of well-being across diverse populations.

In conclusion, this study highlights the promise of AAL technologies in addressing the challenges faced by older adults living independently; however, it underscores that no technology is a panacea. While AAL tools can enhance safety, social connection, and QOL, they cannot fully substitute for the support systems that older adults need. Future research should continue to explore these technologies’ potential across different cultural contexts and emphasize the importance of complementary social and familial support in promoting older adults’ well-being.