Abstract
This study explores how culturally situated technologies can support person-centered dementia care in long-term facilities. As pharmacological options remain limited and standardized care models often misalign with local routines, there is a growing interest in relational, design-led alternatives. Grounded in Kitwood’s theory of personhood and the DementiAbility Method, we adopted a three-year practice-led co-design process informed by participatory and human-centered design principles. Caregivers were involved throughout, from problem framing to prototype refinement, ensuring that cultural norms and relational expectations shaped development. Three interactive systems (Dream Coin Bank, Memory Bracelet, and Hydration Butler) were co-created and deployed in a Taiwanese veterans’ home. Each system is embedded in simple, emotionally meaningful interactions within daily rituals, supporting memory recall, emotional resonance, and relational bridging between residents and caregivers. The findings show that even simple, culturally meaningful technologies can scaffold ritualized engagement, evoke autobiographical memory, and foster emotional resonance through familiar symbols and shared rituals. Rather than emphasizing technological novelty, our findings highlight the value of co-created systems rooted in everyday rituals, sensory cues, and cultural meanings.
Keywords
Introduction
Dementia presents one of the most complex challenges in contemporary health and social care. With the global population aging, the number of individuals living with dementia is expected to rise from 55 million in 2019 to 139 million by 2050, exerting increasing pressure on long-term care systems (World Health Organization, 2021). While pharmacological treatments remain limited in efficacy, a growing body of research has called for person-centered, non-pharmacological approaches emphasizing dignity, meaning, and relational engagement (Kitwood, 1997; Livingston et al., 2020). In this context, design-led strategies have emerged as promising complements to clinical care.
Person-centered care (PCC) offers a philosophical and practical framework that recognizes the enduring agency and relational identity of people with dementia. Kitwood’s theory of personhood (1997) reframes dementia as both cognitive decline and a disruption of social recognition, connection, and meaning. Building on this foundation approaches such as the DementiAbility Method activate procedural memory through meaningful routines, sensory prompts, and environmental cues (Elliot, 2019). These frameworks underscore the importance of engaging the embodied self, where interactional rhythms, gestures, and rituals remain accessible even when verbal capacity fades (Kontos, 2005; Kontos et al., 2017).
Despite increasing endorsement of PCC, its implementation within institutional care settings remains limited. Staffing shortages, rigid workflows, and standardized interventions often constrain opportunities for meaningful interaction (Meiland et al., 2017). Moreover, many assistive technologies in dementia care focus on tracking behavior and ensuring safety rather than supporting meaningful relationships and cultural practices (Greenhalgh et al., 2013; Niemeijer et al., 2010), often developed without sufficient input from caregivers or residents. Such technologies may inadvertently produce alienation, infantilization, or emotional disengagement, particularly when introduced without regard to residents’ lived contexts.
Recent research in dementia care design has begun to challenge this paradigm, exploring how assistive technologies might support human connection rather than simply complete tasks (Bharucha et al., 2009; Brankaert & den Ouden, 2017). These efforts emphasize the co-construction of meaning through familiar gestures, symbolic cues, and emotionally resonant interactions. In particular, the use of participatory design (PD) and human-centered design (HCD) methods allows designers to collaborate with caregivers and communities to ground systems in cultural norms, relational practices, and everyday routines (Norman, 2013; Sanders & Stappers, 2008).
This cultural dimension is especially critical in East Asian dementia care, where values such as collectivism, filial piety, emotional restraint, and indirect communication shape the caregiving experience (Falzarano et al., 2022; Yeh & Bedford, 2003). In such contexts, interventions that rely on direct instruction or generic, one-size-fits-all designs may fail to connect with residents emotionally. By contrast, symbolic rituals such as offering tea or playing a nostalgic melody are often more effective in sustaining engagement and relational continuity. However, few care technologies have been explicitly designed with these cultural values in mind.
This study explores how co-designed, culturally grounded technologies can support person-centered dementia care in Taiwanese long-term care. Accordingly, this study asks: How can participatory, culturally embedded technologies guided by principles of person-centered care and relational personhood enable continuity, recognition, and embodied engagement among people with dementia and their caregivers? Through a three-year practice-led design process, we collaborated with caregivers and residents to develop and deploy three interactive systems, Dream Coin Bank, Memory Bracelet, and Hydration Butler, within a veterans’ home dementia unit. Each system integrates sensory prompts, symbolic routines, and relational cues to facilitate emotional resonance, memory recall, and shared participation in daily care routines.
By aligning assistive technology development with local caregiving practices, embodied memory, and relational engagement, this work contributes to a growing conversation on ethically attuned, culturally meaningful innovation in dementia care. Rather than seeking universal models, we argue that ritual, symbolism, and co-creation must serve as foundational design anchors, particularly in culturally specific care settings, if technologies are to affirm identity, foster agency, and sustain relational care.
Methods
Study Context and Participants
This study was conducted in a dementia care unit of a veteran’s home in North Taiwan, a long-term care facility serving older male residents with moderate to advanced dementia. The unit accommodates up to 40 residents in a secure environment. All residents in this unit were male and had received formal diagnoses of dementia from institutional physicians before the study. However, specific dementia subtypes (e.g., Alzheimer’s disease, vascular dementia) were not documented in the institutional records available to the research team.
Participant selection was carried out in collaboration with the interdisciplinary care team. Based on daily observations and professional evaluations, seven residents were recommended for their responsiveness, engagement potential, and safety. While these participants shared relatively preserved cognitive abilities and willingness to engage, they differed in their personal histories and personalities, which shaped their individual ways of interacting with others. However, due to the COVID-19 pandemic, two residents passed away near the conclusion of the Dream Coin Bank study, resulting in five active participants for subsequent phases.
During the Memory Bracelet study, signal interference from multiple concurrent users led the research and care teams to focus on observing one resident at a time. This strategy allowed for more precise behavioral observation and iterative design refinement. Instead of requiring all participants to follow the same procedures, we adapted our methods to fit the real-world conditions of the care setting. These adjustments reflect the practical realities of dementia care research, where flexibility and ethical responsiveness are essential.
The study was supported by ten frontline caregivers (nurses and care aides) and an interdisciplinary design research team. Residents contributed through lived interaction and behavioral feedback, while caregivers were co-designers and co-observers throughout the study. The design team included researchers with expertise in interaction design, ethnographic methods, and IoT technologies, ensuring methodological rigor across design and deployment stages.
Ethical Considerations
This study received ethical approval from National Taiwan University Research Ethics Committee (NTU-REC No. 202105ES156). The approval covered the study period from August 1, 2021 to July 31, 2022 and was subsequently extended to July 31, 2024 through a continuing review. All procedures complied with the ethical standards outlined by the committee and relevant governmental regulations. Legal guardians provided written consent, and verbal assent was obtained from residents before all interactions. Assent was reaffirmed at each session, given the potential for cognitive fluctuation. All procedures were aligned with the ethical principles of autonomy, dignity, and respect in dementia care.
Design Process
The project followed a three-phase iterative process, corresponding to key care domains identified collaboratively with caregivers: emotional expression, routine reinforcement, and hydration support. Each phase followed a modified Plan–Do–Study–Act (PDSA) cycle, a structured, iterative model often used for continuous improvement in healthcare and human-centered design, and included: • Contextual inquiry (including site observations and caregiver interviews) • Co-design workshops • Prototyping and personalized content development • In-situ deployment and iteration • Feedback sessions and refinement
The three resulting systems: Dream Coin Bank, Memory Bracelet, and Hydration Butler, were co-designed to align with procedural memory, routines that carry personal or cultural meaning (such as saving coins or sharing tea), and cultural expectations. These systems were designed to foster engagement and emotional resonance by incorporating sensory prompts and personalized content such as familiar objects, nostalgic music, and caregiver-recorded videos. These co-design processes were further attuned to East Asian care dynamics, emphasizing indirect communication (e.g., using music or objects to express care instead of verbal instructions), sensory rituals (e.g., the sounds of coins dropping or tea being poured), and familial symbolism (e.g., a toy airplane representing a resident’s wish to visit family).
Briefly, each system addressed a specific care need: (1) the Dream Coin Bank is a box-shaped device with a rotating display on top; when residents insert coins, the device plays personalized nostalgic music while a symbolic object, chosen based on the resident’s life story, rotates on the display, designed to evoke autobiographical memories and provide emotional comfort; (2) the Memory Bracelet is a wearable device paired with a Bluetooth beacon that triggers personalized video messages on a nearby screen when the wearer approaches a designated location, designed to reinforce daily routines; and (3) the Hydration Butler is a mobile beverage cart equipped with NFC (Near Field Communication)-enabled cups and sensory prompts (lights and music), designed to encourage fluid intake through a culturally familiar tea-sharing ritual. Detailed descriptions of each system are provided in the Findings section.
Data Collection
A multi-method qualitative strategy was employed, emphasizing naturalistic and participatory observation. Frontline caregivers played an active role in documenting resident engagement within daily routines: • Caregiver-facilitated field observation: Over 150 hours of field data were collected through caregiver logs and reflections, capturing behavioral cues, affective responses, and system use. • Semi-structured interviews: Caregiver interviews were conducted before and after deployment phases, examining usability, emotional engagement, and care integration. • System interaction logs: Device sensors recorded engagement events, including coin insertions, proximity triggers, and hydration usage. • Researcher reflection journals: Weekly design meetings generated reflexive memos documenting the evolving design logic and relational insights.
This integrated observation strategy acknowledged caregivers’ tacit knowledge and allowed for sustained embedded data collection. Triangulation across data sources enhanced interpretive reliability.
Data Analysis
Data were analyzed using reflexive thematic analysis (Braun & Clarke, 2006, 2019). Initial codes were generated inductively and refined through collaborative discussion among researchers. NVivo software supported coding and theme development. Observations of residents’ behaviors, recorded primarily by caregivers during their daily routines rather than through prolonged researcher presence, were combined with interview data and device usage records (e.g., timestamps of coin insertions or cup placements) to strengthen the reliability of our interpretations.
Themes were interpreted relationally, focusing on how behaviors and interactions signaled emotional engagement, moments when residents expressed or reconnected with their sense of self, and caregiver–resident dynamics shifts. The analysis yielded three central themes: ritualized engagement (interactions following familiar, repeatable patterns with symbolic significance), emotional resonance, and relational bridging, corresponding to the design goals and emergent care patterns observed. Coding discussions were conducted iteratively among team members to enhance interpretive alignment and ensure coherence across field notes, interview narratives, and behavioral observations.
Findings
This section presents the findings from deploying three co-designed interactive systems, Dream Coin Bank, Memory Bracelet, and Hydration Butler, within a Taiwanese long-term dementia care facility. These systems were developed through a participatory, practice-led approach and are examined here not as discrete technologies, but as relational and cultural interventions embedded in daily caregiving routines.
In alignment with the study’s aim to support person-centered dementia care through ritualized engagement, symbolic cues, and culturally grounded design, this section highlights how each artifact operated as a medium for memory, recognition, and relational connection. The analysis is organized around three core themes: ritualized engagement, emotional resonance, and relational bridging, emerging across all three systems.
Dream Coin Bank: Evoking Memory through Ritual
Design Motivation and Cultural Foundations
The Dream Coin Bank was the first interactive prototype developed in the study. Drawing on the cultural familiarity of piggy banks and coin-based rituals in Taiwanese society, the system was designed to stimulate reminiscence and emotional expression through symbolic, low-effort interaction. A slot-based form factor embedded with an Arduino microcontroller, sound module, and LED lighting translated this ritual into an emotionally resonant experience.
Upon coin insertion, the device played a personalized song associated with the resident’s youth, triggering autobiographical memory and emotional engagement. A rotating acrylic disc atop the device displayed individualized ‘dream objects,’ small symbolic tokens representing personal aspirations, former occupations, or cherished memories. It evoked the gentle motion of a music box (see Figure 1). The Dream Coin Bank features a personalized symbolic object, a miniature airplane mounted on the rotating display. The design invites residents to insert coins as a ritualized, memory-evoking interaction.
Iterative Co-Design and Deployment
Feedback from co-design sessions with caregivers led to refinements, including smoothing the casing edges, adjusting audio volume, and curating personalized song selections based on each resident’s life history. Early prototypes had generic songs and stark white casings, which caregivers flagged as emotionally distant and physically uninviting. These were replaced with warmer color schemes and culturally familiar musical selections, resulting in a more embodied and emotionally accessible interaction. The final prototype emphasized sensory consistency, cultural relevance, and predictable feedback qualities shown to support relational engagement and mood regulation.
The system was deployed from September to November 2022. Residents were invited to deposit symbolic coins into the bank, which triggered audio and visual feedback. The rotating dream object served as a visual anchor for memory recall. Data logs indicated increased interaction during late afternoons. While this pattern aligned with caregivers’ observations of sundowning behaviors, it may also reflect the timing of scheduled activities or the availability of unstructured time during this period.
Narrative Interaction and Memory Activation
Residents frequently engaged with the system as a prompt for storytelling and emotional reflection (see Figure 2). One participant, whose dream object was a miniature airplane, remarked, “I am saving up for a plane ticket to Brazil to see my child,” linking the symbolic object to a cherished personal aspiration. This also demonstrated preserved recall of meaningful biographical details, a noteworthy observation for someone living with dementia. Another participant, upon hearing a beloved song, shared with visible excitement, “This is one of my favorites,” illustrating the system’s ability to evoke autobiographical memory through familiar sensory cues. A resident and caregiver engage with the Dream Coin Bank on a quiet afternoon. The system served as a calming focal point, evoking autobiographical memory and prompting personal storytelling.
Caregivers reported that the device helped to de-escalate agitation, such as restless pacing. “When residents become anxious, we bring them to the coin bank. It calms them, they get distracted in a good way,” one caregiver noted. While caregivers typically attempted verbal reassurance first, they found that offering the tangible, familiar activity often proved more effective than direct questioning about the source of distress. Similarly, another caregiver observed, “Even if they forget the interaction afterward, they are still excited to return and do it again.”
Even without active coin insertion, the symbolic presence of the Dream Coin Bank often invited spontaneous conversation. One resident stood quietly near the display and began recounting a childhood story after noticing the airplane figure. Another routinely paused at the train-themed object to initiate a conversation with others. These unstructured interactions offered caregivers deeper insight into residents’ identities and life narratives. As one nurse reflected, “Just having the object there gives us something to talk about... it helps us see who they were and who they still are.” The Dream Coin Bank units were placed at a shared care station, allowing residents to encounter them during daily routines and fostering informal peer interaction (see Figure 3). Multiple Dream Coin Bank units are placed at a shared care station, each featuring personalized symbolic objects selected through participatory design. These visual artifacts served as passive cues that invited reflection and fostered informal peer interaction among residents.
Thematic Insights
These observational findings converged into three central themes illustrating how the Dream Coin Bank scaffolded memory, modulated emotional states, and fostered relational engagement.
Theme 1: Ritualized Engagement as Memory Scaffold
The coin ritual provided a simple but meaningful structure that supported memory expression. Even residents with limited verbal ability were drawn to the act of coin insertion, often linking the gesture to deeply held narratives.
Theme 2: Ambient Cues and Emotional Resonance
Audio prompts and dream objects created sensory anchors that evoked emotional responses. Music served as a mood regulator, while symbolic figures invited gentle reflection.
Theme 3: Co-Designed Systems as Relational Bridges
Iterative refinements enhanced the system’s cultural alignment and emotional tone. For caregivers, the Dream Coin Bank became a shared emotional resource and a conversational gateway.
Memory Bracelet: Reinforcing Routines through Contextual Cues
Design Purpose and Contextual Rationale
The second interactive system, the Memory Bracelet, was designed to support daily routines through context-sensitive, wearable prompts. Developed in close consultation with caregivers, the system employed a Bluetooth Low Energy (BLE) beacon installed at a key care environment location, such as a hallway, to trigger personalized video messages displayed on a nearby wall-mounted screen. Each video featured a familiar caregiver prompting the resident to complete a specific task, such as drinking water or eating a meal.
Routine mapping sessions and contextual interviews with frontline caregivers shaped the design process. Co-design workshops emphasized experiential needs such as unobtrusiveness, aesthetic acceptability, and comfort. Iterative testing led to refinements, including screen brightness adjustments, strap resizing to accommodate diverse wrist sizes, and interface simplification to reduce cognitive load.
Figure 4 presents the Memory Bracelet prototype alongside a BLE tag, highlighting how the wrist-worn device detects proximity and triggers audiovisual prompts based on each resident’s personalized routine. Figure 5 illustrates the interaction scenario: As a resident wearing the Memory Bracelet approaches a beacon zone, a nearby screen displays a familiar video message featuring a caregiver. This interaction reinforces memory cues and emotional continuity through personalized, relational content. Bluetooth-enabled Memory Bracelet with integrated beacon triggers personalized video prompts on an adjacent display, supporting routine engagement through ambient cues. Illustration of a resident interacting with a video display triggered by the Memory Bracelet system. Videos featured familiar caregivers or personally meaningful content to support recognition and engagement.
Deployment, Adaptation, and Behavioral Impact
The system was initially deployed in June 2023. Early testing involving multiple residents revealed a key challenge: when more than one resident approached a screen simultaneously, the system struggled to determine which video to play, resulting in interference and confusion. Caregivers also observed that residents were less engaged unless the video featured emotionally salient, familiar caregivers.
In response, the care and research teams temporarily revised the deployment strategy. Rather than continuing with a multi-user configuration, the team selected a single participant for focused observation: an older adult with a pre-dementia routine of practicing Tai Chi. This strategy allowed for clearer system performance and more accurate behavioral interpretation.
Personalized videos featuring slow, familiar Tai Chi movements were introduced in Week 2. By Week 3, the resident began lingering near the screen; by Week 4, he mimicked the movements. When the device was temporarily removed in Week 7, the behavior ceased. Upon reintroduction, the resident resumed practicing and even began teaching gestures to staff. One caregiver remarked: “At first, he barely responded. Later, he was focused and even taught us Tai Chi.” Another added, “This changed how we saw him. The prompts brought back parts of him we thought were gone.”
This single-user deployment provided a clearer view of how ambient, personalized content could function not only as a memory aid but also as a behavioral cueing tool grounded in recognition and embodied familiarity. By embedding emotionally resonant media into a familiar space, the Memory Bracelet supported agency, enriched relational dynamics, and reaffirmed personhood through sensorimotor interaction. This case also laid the groundwork for understanding how cueing technologies may evolve in complexity as they scale beyond single-user deployment.
Design Learning from Varied Engagement Patterns
While the Tai Chi case offered a compelling example, the effects should not be overgeneralized. The decision to focus on one participant provided operational clarity, but it also limited exposure to a broader range of responses. Even in earlier stages, more passive forms of engagement were observed, including pausing near a screen, watching silently, or smiling upon recognizing a caregiver’s face. Caregivers emphasized that even brief moments of attention or acknowledgment fostered interpersonal connection. One caregiver noted: “Even if they just stop for a few seconds, that moment is a connection.”
From a design standpoint, these subtler responses revealed the value of emotionally resonant, ambient content, even when overt behavior was limited. The Tai Chi case thus serves not as a normative outcome but as a proof of possibility, illustrating how well-timed, personalized media can unlock latent capacities when matched with context, memory, and relationship.
Thematic Insights
These findings converged into three central themes that highlight how personalized, ambient interactions scaffolded memory, emotional recognition, and relational continuity through the Memory Bracelet system.
Theme 1: Ritualized Engagement as a Memory Scaffold
Personalized video routines anchored behavioral repetition. For the Tai Chi participant, the repetition of context-specific prompts reactivated long-dormant procedural memory and created a structure for meaningful engagement.
Theme 2: Ambient Cues and Emotional Resonance
Residents responded more readily when videos featured emotionally familiar caregivers, reinforcing how relational recognition enhances attention and affective engagement.
Theme 3: Co-Designed Systems as Relational Bridges
Weekly care team discussions led to real-time adjustments in video sequencing and tone. This iterative process positioned the Memory Bracelet as a shared, evolving emotional medium. As one staff member reflected: “We began talking about what videos to show… not just what works, but what they might love.”
Hydration Butler: Supporting Autonomy through Ambient Prompting
Design Motivation and System Overview
The third interactive system, the Hydration Butler, was developed to promote fluid intake among residents through personalized reminders and unobtrusive tracking. The system featured a mobile beverage cart equipped with NFC (Near Field Communication) readers, which allowed caregivers to record when and to whom water was offered by scanning each resident’s personalized cup, an integrated speaker, and a modular hydration-tracking platform to transform hydration into a culturally meaningful and routine activity. Each resident received a cut-out cup (a cup with a curved cutout that allows drinking without tilting the head back, helping prevent choking) labeled with their name and embedded with an NFC tag (a small chip that allows the cup to be identified when scanned). When the water-filled cup was placed on the light-up base, a weight sensor detected the water and the base illuminated, both indicating that water was available and attracting residents’ attention (see Figure 6). Simultaneously, the cart recorded timestamped drinking events via the NFC system, enabling caregivers to estimate whether residents approached the recommended daily fluid intake of 1500 cc (see Figure 7). Cut-out cup with embedded NFC tag, paired with a light-up base. The Hydration Butler cart includes personalized cups and a modular tracking system. Cups with NFC tags rest on light-up bases that illuminate when filled with water. Timestamped interactions are logged, enabling hygienic distribution and behavioral tracking.
Although the full cart system was made available to 20 residents, the study focused on five residents who formally participated in the research. This narrowed scope ensured methodological alignment while enabling deeper observation within contextualized care routines.
Deployment and Behavioral Evolution
Deployed between May and early July 2024, the Hydration Butler was initially met with two challenges: residents were not attracted to the cart as expected, and caregivers were unsure where to scan cups because the NFC reader location was poorly marked. The system was updated to include a clearer visual indicator on the NFC reader, brief musical cues, and simplified lighting to attract residents and make the system easier to use.
Following these refinements, resident engagement improved. One nurse observed, “When they hear the music, they come over. It’s like a gentle lure.” Caregivers reported that the system helped them estimate hydration levels, even if volumes were approximate. “It’s hard to get them to drink. This way helps us gauge how much they’re taking in,” one aide noted.
By week six, some caregivers noted a perceived decrease in UTI symptoms. A nurse shared, “Maybe it’s a coincidence, but since using this cart, I haven’t seen any new UTI cases.” The cart evolved from a functional hydration device into a symbolic and relational object. Residents began to anticipate its arrival and engage with it more autonomously. One resident smiled and tapped their cup to the rhythm of the music before drinking, an interaction that caregivers described as previously rare.
Social and Sensory Engagement
Beyond individual behavior change, the cart began facilitating shared rituals. Residents were observed gesturing toward the cart or speaking about it before it entered the room, suggesting emotional anticipation and temporal orientation. A caregiver remarked, “It’s no longer just a tool; it feels like part of the day. They notice it before we even roll it in.”
Team reflections emphasized the cart’s value in promoting group-based hydration. “Group rounds with the cart help us see who’s not drinking. It’s harder to track when we bring water individually,” explained one nurse. The shared visibility of the cart created subtle social reinforcement, encouraging residents to mirror their peers’ actions (see Figure 8). A caregiver assists a resident in approaching the Hydration Butler cart. The interaction reframes hydration as a moment of social and sensory connection.
Design Adaptation and Future Potential
Caregivers observed that residents were more receptive when hydration was paired with personally appealing options such as flavored drinks or music. “They don’t like plain water. But when we change the flavor or play music, it helps,” said one care aide. These observations reaffirmed the value of multisensory, affect-driven interaction in dementia care settings, as well as the importance of treating residents as persons first who happen to be living with dementia.
At the same time, operational concerns arose around the cart’s stability and unattended use. Care team members discussed the need for improved wheelbase design and parking protocols to prevent obstruction or accidental tipping. While the revised cart version proved more user-friendly, staff emphasized the importance of continued iteration and environmental adaptation.
Notably, the cart drew the attention of visiting staff from other facilities. “People from other centers have asked where to get it. There is a real future here, but we still need to fine-tune cups and routines,” noted one nurse. This interest reinforced the potential of the cart for broader application, particularly if future iterations remain attentive to sensory cues, cultural familiarity, and relational dynamics.
Thematic Insights
These findings converged into three central themes highlighting how personalized, multisensory interactions scaffolded memory, emotional resonance, and relational engagement through the Hydration Butler system.
Theme 1: Ritualized Engagement as a Memory Scaffold
Hydration became more than a clinical task, becoming a predictable, shared ritual. Music and personalized cups created a sensory routine that residents could anticipate and participate in. “They know the cart means it’s time. Some even wait near it,” a caregiver shared.
Theme 2: Ambient Cues and Emotional Resonance
The cart’s presence, audio cues, and visual aesthetics offered ambient emotional prompts. As one nurse put it, “The music draws them in. It’s like the cart has a personality.”
Theme 3: Co-Designed Systems as Relational Bridges
Feedback-driven refinements shaped the cart into a tool of shared engagement. “We used to push water on them. Now they come to us,” one nurse reflected. The cart became functional and relational, reshaping hydration as a social interaction.
Taken together, the Dream Coin Bank, Memory Bracelet, and Hydration Butler illustrate how relationally attuned technologies, although operating through different modalities, can support common care values of identity, presence, and emotional resonance.
Cross-System Insights and Integrative Reflections
Building on the system-level findings, this section integrates cross-system insights to show how the Dream Coin Bank, Memory Bracelet, and Hydration Butler collectively illustrate how relationally attuned technologies, though operating through different modalities, can support common care values of identity, presence, and emotional resonance.
The three systems demonstrate how ritualized, ambient, and culturally situated technologies can reconfigure dementia care from a task-oriented model into one that affirms identity, fosters emotional resonance, and sustains relational continuity. Despite operating through different modalities, symbolic engagement, wearable cueing, and environmental prompting, they converge around a shared ethos: enabling residents to act, feel, and be recognized within the rhythms of everyday care.
While the Memory Bracelet was ultimately deployed with a single participant, this focused use surfaced nuanced relational patterns that might have been obscured in broader group trials, showing that in-depth observation of individual users can reveal insights that larger-scale deployments might miss.
These systems did not displace caregiver roles but rather expanded the relational space available for engagement. They embedded design within routines already familiar to residents, activating procedural memory, emotional recall, and embodied habit. This fostered new entry points for micro-interactions such as smiles, gestures, and shared stories that strengthened caregiver–resident relationships, as illustrated by residents sharing life stories when seeing their dream objects or caregivers discovering residents’ hidden abilities through video prompts.
These systems also directly respond to the study’s central research question: How can participatory, culturally embedded technologies guided by principles of person-centered care and relational personhood enable continuity, recognition, and embodied engagement among people with dementia and their caregivers? The findings demonstrate that when assistive systems are co-created with caregivers, grounded in familiar rituals, and tuned to cultural values, they can foster engagement, emotional presence, narrative activation, and sustained relational connection.
Comparative Overview of Three Co-Designed Interactive Systems for Dementia Care
Together, these systems demonstrated complementary strengths, showing how co-designed, culturally situated technologies can be layered to support diverse modes of engagement. These systems act not merely as behavioral cues but as relational prosthetic material interventions that extend and reinforce the relational fabric between residents and caregivers.
Discussion
In this discussion, we interpret the study’s findings by returning to the theoretical foundations outlined in the introduction, particularly Kitwood’s (1997) concept of relational personhood, the DementiAbility Method’s (Elliot, 2019) emphasis on procedural memory and environmental cues, and participatory and human-centered design frameworks (Norman, 2013; Sanders & Stappers, 2008). Rather than treating technologies as neutral tools, the findings foreground their relational, cultural, and ethical significance within long-term dementia care.
Rethinking Dementia Practice: Technologies as Social Companions
Across all three systems, caregivers observed that residents responded not to technological novelty but to emotionally salient, simple interactions that required minimal effort. As Kitwood (1997) asserted, dementia does not erase personhood but necessitates alternative modes of recognition, such as non-verbal, sensory, and ritualistic ways of acknowledging a person’s identity when verbal communication becomes limited. Through coin rituals, ambient videos, and shared hydration cues, residents engaged not as passive recipients but as active participants who could connect with and respond to others. These observations resonate with Livingston et al. (2020), who advocate for interventions that nurture social presence and emotional well-being rather than focus solely on symptom management.
As one nurse described, the coin bank ritual became “a way to calm residents without medication.” Another noted that “even without full understanding, residents smiled when they heard a song or saw something familiar.” Such affective responses, though modest, reveal modes of engagement often overlooked in clinically defined metrics of efficacy. These responses also suggest that understanding may take forms beyond what caregivers can readily observe or measure. By prioritizing relational interaction, these technologies reposition care as co-presence rather than compliance.
Material and Cultural Contexts: Situated Rituals and Embodied Memories
The effectiveness of these interventions stemmed not primarily from explicit cognitive stimulation but from engagement through familiar routines embedded within the physical and symbolic textures of daily life. This finding echoes the DementiAbility approach, which emphasizes procedural memory and environmental prompts to support autonomy (Elliot, 2019) and aligns with Clare and Woods’ (2004) call to leverage retained habits in cognitive rehabilitation.
These observations also suggest that implicit memory, which refers to memory expressed through actions, emotions, and preferences without conscious recall, remained accessible to residents. For example, residents who ‘forgot’ previous interactions with the Dream Coin Bank nonetheless returned to it with evident enthusiasm. Similarly, a participant in the Memory Bracelet study responded positively to familiar faces in video prompts despite being unable to verbally identify caregivers. Residents also began to anticipate the Hydration Butler’s arrival, suggesting learned associations between sensory cues and the hydration routine. Such patterns align with research showing that implicit and emotional memory systems are often preserved in dementia, even when explicit recall is impaired (Kitwood, 1997). This finding reinforces the importance of designing interventions that engage these retained capacities rather than focusing solely on cognitive deficits.
In the Taiwanese care context, where indirect communication, sensory ritual, and shared routine shape interaction (Yeh & Bedford, 2003), systems such as the Dream Coin Bank and Hydration Butler succeeded precisely because they invoked symbolic gestures rather than requiring verbal comprehension. One caregiver reflected, “The residents may not speak, but they understand rhythm. The routine helps them feel grounded.” This observation reminds us that verbal expression is not the sole indicator of comprehension. Residents may understand far more than they can articulate, and their non-verbal engagement should be recognized as meaningful communication.
These insights reinforce the value of culturally attuned design. Falzarano et al. (2022) emphasize that culturally congruent interventions elicit stronger emotional responses and reduce resistance. Rather than imposing standardized solutions, this study’s systems aligned with daily values such as filial piety, emotional subtlety, and communal rhythm, contributing to their acceptance and efficacy.
Theoretical and Methodological Reflections
This study adopted a practice-led co-design method grounded in in-situ deployment and contextual adaptation. As Brankaert and den Ouden (2017) argue, co-design with stakeholders in dementia care is not only ethically sound but epistemologically necessary, as knowledge is often embedded in tacit gestures, routines, and relational cues. The iterative refinement of all three systems embodies this principle.
A key methodological insight emerged during the deployment of the Memory Bracelet. Initial efforts to support multiple users simultaneously encountered technical limitations. Specifically, screen interference is caused by overlapping signals from nearby bracelets. This disrupted interaction flow and obscured emotional and behavioral patterns in group settings. In response, the team adopted a single-user deployment strategy, which allowed for more precise tracking of cause and effect. The Tai Chi case became a focal point not by design but by necessity. This pivot also enabled more nuanced caregiver feedback and offered a higher-resolution view of how relational, context-sensitive technologies might scaffold routine participation. While the limited scope reduced generalizability, the depth of engagement captured through this focused lens underscores the value of ‘deep slice’ approaches in participatory design for dementia care.
Nevertheless, several tensions emerged. While participatory design and human-centered design enable highly localized and resonant interventions, they may not address structural limitations, such as staff shortages or infrastructural constraints. Furthermore, while the DementiAbility framework informed much of the design logic, it remains under-theorized in terms of cultural variation, suggesting the need for future research to localize its principles more rigorously.
This study contributes to a broader conceptual shift from evaluating assistive technologies based solely on task performance to assessing their contribution to lived experience, symbolic participation, and relational continuity. This perspective aligns with Kontos et al. (2010), who call for ethical attentiveness to the ‘embodied selfhood’ of persons with dementia. Technologies must be judged not merely by what they do but by what they make possible.
Cultural Reflections: Dementia Care in the Taiwanese Context
Taiwanese dementia care reflects a hybrid of Confucian values, intergenerational obligations, and institutional characteristics. Caregivers emphasize environmental familiarity, emotional subtlety, and non-confrontational prompting practices shaped by cultural ideals of harmony and respect.
Indirect language and contextual cues often guide communication to elicit participation without confrontation (Yen, 2021). Filial piety shapes caregiving responsibilities as a collective duty rather than an individual choice. Li et al. (2024) note that such norms can complicate discussions around autonomy, risk, and end-of-life care.
The study’s findings echo these cultural dynamics. Residents responded more readily to multisensory, indirect cues than explicit commands, and technologies grounded in personal memory, familiar songs, symbolic objects, and known caregiver voices proved more effective than generic prompts.
This affirms the need for dementia care design in Taiwan to integrate symbolic familiarity, non-verbal affect, and culturally resonant interaction patterns. Taiwan’s Dementia Action Plan 2.0 (2018–2025) has supported the development of integrated care centers and community services (Ministry of Health and Welfare, 2018), laying a policy foundation for incorporating culturally sensitive technologies.
The DementiAbility Method, initially developed in Western contexts, showed cultural adaptability. Its emphasis on structured activity, sensory engagement, and retained capacities aligned with East Asian caregiving values. When localized through co-design, the approach proved both emotionally and culturally meaningful.
Implications for Practice
The study yields several implications for practitioners and designers: • Technologies need not be complex to be effective. Simple, ritualized, and symbolically meaningful tools like music, familiar objects, or caregiver-voiced videos can foster engagement when embedded in daily routines. • Co-design should remain iterative and context-specific. Institutional support is needed to sustain adaptive design, encouraging caregivers to refine tools based on relational knowledge continually. • Cultural responsiveness is essential. Designers should integrate local communication patterns, affective norms, and rhythms of care into interactional logic and aesthetic form. • Reframe technologies as tools that support human connection. Success should be measured not only by task completion but also by how technologies enable shared meaning, emotional resonance, and identity expression.
An additional consideration for practice relates to deployment strategy. The Memory Bracelet deployment revealed that even well-designed systems can encounter challenges when used in shared environments with multiple residents. In this case, multiple users’ proximity to BLE beacon zones led to confusion about which personalized video to play, disrupting engagement. The care team responded by shifting to a single-user configuration for observation, resolving the technical interference, and providing richer behavioral insights and caregiver feedback. This adjustment underscores the need for flexible deployment planning to respond to dynamic care environments, spatial constraints, and staff capacity. Designers and implementers should remain attentive to operational realities and be prepared to scale up from individualized use based on early-stage learnings.
Summary of Practice-Oriented Design Principles for Developing Culturally Grounded and Relational Dementia Care Technologies
Limitations and Future Research
Several limitations should be acknowledged. First, although the deployment periods were substantial, the findings are based primarily on qualitative data from a single institutional setting. Future research could strengthen validity by employing mixed methods approaches that combine measurable behavioral data (e.g., frequency of device use, duration of engagement) with lived experience.
Second, the sample was small and limited to residents with mild-to-moderate dementia, which constrains the generalizability of the findings. Although caregivers acted as interpretive proxies, future studies should explore more inclusive methods to directly access the perspectives of individuals living with more advanced dementia.
Third, specific dementia subtypes were not differentiated in the available diagnostic records. Given that different types of dementia (e.g., Alzheimer’s disease, vascular dementia, Lewy body dementia) may affect responsiveness to sensory and social cues differently, future research should examine whether the observed engagement patterns vary across dementia subtypes.
Fourth, clinical outcomes such as fluid intake levels or agitation reduction were not systematically measured. Future research could integrate ethnographic insights with standardized clinical metrics to bridge subjective experience and measurable outcomes.
Finally, the study’s institutional and cultural specificity, conducted within a Taiwanese veterans’ care home, limits broader applicability. Replication across diverse cultural and organizational contexts is necessary to test the transferability and adaptability of the proposed design principles.
Summary
This study reframes assistive technologies in dementia care not as instruments of efficiency or automation, but as tools that support relationships and cultural practices, scaffolding emotional presence, symbolic expression, and everyday participation. Through a practice-led co-design method, three interactive systems including Dream Coin Bank, Memory Bracelet, and Hydration Butler were developed to align with residents’ lived experiences, sensory preferences, and caregiving traditions within Taiwan’s long-term care context.
Rather than emphasizing technological novelty, the study highlights the value of familiar rituals, personalized cues, and co-created routines in fostering recognition, affirming identity, and deepening caregiver–resident relationships. Technologies grounded in symbolic familiarity, indirect communication, and relational rhythms proved particularly resonant, reflecting Taiwanese cultural values such as filial piety, collective ritual, and emotional subtlety.
This work contributes to the growing movement to humanize dementia care by demonstrating how culturally grounded, person-centered technologies can affirm dignity, foster emotional resonance, and support relational continuity within daily care routines. The success of these systems lay not in their technical complexity but in their ability to mirror the aesthetic, ethical, and relational contours of the care environment.
Future research and policy should prioritize longitudinal and cross-cultural investigations and advocate for the thoughtful integration of relational technologies into national dementia care strategies. Such integration must remain grounded in local caregiving practices, embodied memory, and cultural knowledge to ensure that innovation not only improves clinical outcomes but also enriches the relational textures, emotional affordances, and symbolic vitality of care.
By demonstrating how everyday technologies, when co-designed with caregivers and attuned to embodied memory, can elicit storytelling, evoke recognition, and scaffold relational connection, this study answers the call for contextually responsive, emotionally meaningful innovation in dementia care. It underscores the transformative potential of interdisciplinary collaboration and culturally sensitive design in reshaping not only care tasks but also the lived experience of caregiving itself.
Footnotes
Acknowledgments
This study was supported by the National Science and Technology Council (NSTC), Taiwan, under a three-year research project titled “Integrating the Concept of ‘DementiAbility’ with the Notion of Human-Centered Design to Construct the Cognitive Rehabilitation Devices for Assisting Long-term Care Organizations.” The research was funded through grant numbers 110-2410-H-036-008-, 111-2410-H-036-006-, and 112-2410-H-036-003-. We also extend our sincere appreciation to the Taipei Veterans Home, under the administration of the Veterans Affairs Council of Taiwan, for their institutional support and for providing a valuable care environment for field deployment. We are especially grateful to Mr. Yi-Ting Lu for his technical consultation and guidance on IoT integration throughout the project. Most importantly, we are deeply grateful to the dementia care team and the residents who participated in this study. Their engagement, insights, and collaboration were instrumental in shaping the co-design process and bringing the artifacts to life.
Ethical Considerations
The study was approved by National Taiwan University Research Ethics Committee (NTU-REC No. 202105ES156). The approval covered the study period from August 1, 2021 to July 31, 2022 and was subsequently extended to July 31, 2024 through a continuing review. All procedures complied with the ethical standards outlined by the committee and relevant governmental regulations. Legal guardians provided written consent, and verbal assent was obtained from residents before all interactions. Assent was reaffirmed at each session, given the potential for cognitive fluctuation. All procedures were aligned with the ethical principles of autonomy, dignity, and respect in dementia care.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the National Science and Technology Council, Taiwan (Grant numbers 110-2410-H-036-008-, 111-2410-H-036-006-, and 112-2410-H-036-003-).
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.