Reassembling the elderly consumption ensemble: retaining independence through smart assisted living technologies

ABSTRACT

Smart assisted living technologies are often touted as the solution to the challenges associated with an ageing population. Viewing elderly consumers, their relatives, and technologies as comprising an assemblage, this article aims to understand how smart objects actively reshape the everyday practices in families with elderly consumers. Interviews with and observations of users of smart alarm systems indicate a stratification of the paired experiences of users and systems and identify a tension between enabling experiences of the elderly and constraining experiences of the relatives. This article contributes to views of families with elderly as assemblages by providing insights into joint and disjoint consumer experiences in multiple consumer-object assemblages, identity negotiations of the elderly and their relatives, and the hidden costs of smart assisted living technologies.

KEYWORDS:

• Elderly consumption ensemble
• consumer-object assemblage
• smart objects
• Internet of Things
• assisted living technology
• assemblage theory

Introduction

As a direct consequence of an ageing population, elderly consumers represent an increasingly large and important group of consumers (Guido et al., 2018). Elderly consumers are exposed to vulnerabilities due to waning physical and mental abilities (Roy & Sanyal, 2017), progressively limiting consumption as an individual phenomenon as age progresses and abilities decline. In the context of family care, prior research has posited that the consumption of elderly consumers is a group phenomenon that is divided among the members of an elderly consumption ensemble (Barnhart & Peñaloza, 2013) comprising the elderly consumer, their relatives (Dean et al., 2020), and formal caregivers (Barnhart et al., 2014).

The desire and preference of most elderly consumers to live as independently as possible even when they regularly require either or both health and social care is presenting a major societal challenge (Offermann-van Heek et al., 2019) as there is a disparity between the growing number of such elderly consumers and the stagnating capacities of welfare and caregivers. Smart objects such as assistive robots (Bemelmans et al., 2012) and ambient assisted living technologies (Campos et al., 2016) have been presented and touted as solutions to this challenge. They have the potential to enable and facilitate both the physical care for (Bedaf et al., 2015) and the social assistance to (Čaić et al., 2018) elderly consumers, counterbalancing their physical and social vulnerabilities.

Consequently, elderly consumers, who find it challenging to live completely independently but still strive to do so, are increasingly adopting and consuming smart assisted living technologies and services to compensate for their increasing vulnerability. This development warrants revisiting and rethinking the elderly consumption ensemble from a new materialistic perspective. To this end, in this article, we embrace a ‘flat’ ontology, where human and non-human materials exist in the same category and human materials are no more fundamental than non-human ones (Feely, 2020). Concretely, we apply assemblage theory to be able to consider objects such as assisted living technologies as agentic materials in addition to the members of the elderly consumption ensemble. This article aims to understand how smart objects actively affect the capacities of the members of the elderly consumption ensemble and their relations.

Drawing upon new materialist perspectives on consumer-object assemblages (Hoffman et al., 2018), we investigate the integration of smart objects into the elderly consumption ensemble through a qualitative study of how an Internet of Things-based smart alarm system affects the everyday practices and relations in the elderly consumption ensemble. We analyse the emerging consumer experience assemblages, identifying a tension between the enabling experiences of the elderly consumers and the constraining experiences of their relatives with their liminal identity between informal and formal caregivers.

The significance of these findings extends beyond the particular context of smart alarm systems studied in this project and paves the way for several contributions. First, we provide empirical evidence to Hoffman et al. (2018) theoretically-grounded concept of consumer-object assemblages. In addition, we extend it to the case of multiple heterogeneous consumers possibly filling a multiplicity of roles. As a consequence, we are able to show a variety in consumer experiences, for example, in the degree to which they are shared in the assemblage. We also demonstrate that this degree may depend on the type of experiences considered.

Second, we show that identity negotiations of elderly and relatives are connected through enabling and constraining consumer experiences. Our discussions and conclusions about identity formation pertain to the type of assemblages we studied and are, thus, not necessarily valid for social constructions of ‘old age identity’, in general. Our focus is not on ‘identities’ but on ‘identity negotiations’, i.e. ongoing processes as part of the studied socio-technical assemblages. In line with Rosenthal et al. (2020), we find that the elderly that are part of this study strive to increase their capacities and destigmatise their old age identities through consumption while the relatives experience diminished capacities and liminal identities between formal and informal caregivers. Particularly female relatives struggle to shed such liminal identities due to identities of being a ‘good child’ to their ageing parents, mirroring the identities of being a ‘good parent’ to their child, as discussed by Cardoso et al. (2020).

Third, and lastly, we provide insights into the micropolitics of the elderly’s intense but ambivalent quest for independence, providing empirical evidence for Puntoni et al. (2020) claim that artificially intelligent objects such as smart objects are inherently political tools that not only help to unleash the ‘creative capacity of families’ (Cardoso et al., 2020) but also incur hidden costs by placing the burden of care on the relatives of care-receiving family members. From this perspective, the quest for independence may be seen as ultimately futile.

Theoretical background

Consumption assemblages

There is a long tradition of studying consumption and consumer culture through social constructionist and post-structuralist perspectives (Arnould & Thompson, 2005). These perspectives view discursive statements as constituting and shaping the material world (Feely, 2020). In what has been labelled the ‘material turn’ in analogy to the earlier ‘linguistic turn’, social constructionist epistemology and its focus on textuality and human agency have been supplemented with flat ontologies, where dualisms such as structure and agency are abandoned in favour of a monistic sociology (Fox & Alldred, 2018; Thompson, 2019) and discursive entities and materials coexist on the same ontological plane (Feely, 2020).

One such new materialist perspective builds on Deleuze and Guattari (1987) concept of assemblages as collections of heterogeneous human and non-human components. Instead of non-material perspectives that attempt to understand the social world through the metaphor of the organism with its hierarchically organised parts contributing by their assigned particular functions, in an assemblage theory all components are on the same level and their capacities and relations are fluid and emerging. Assemblage theory can, thus, capture both the ephemeral and fluid aspects of collections of human and non-human components (Diaz Ruiz et al., 2020).

Assemblage-theoretical perspectives are increasingly present in consumer research and have been used to study a wide variety of subjects, including consumers’ romantic experience of nature (Canniford & Shankar, 2013), the outsourcing of parenthood to paid services (Epp & Velagaleti, 2014), the dissipation of brand audiences (Parmentier & Fischer, 2015), the role of social media in branding (Rokka & Canniford, 2016), and alternative food markets (Cherrier, 2017). More recent studies range from food as an embodiment of care for elderly relatives (Trees & Dean, 2018), and the relationship of consumers with voice-controlled smart devices (Schweitzer et al., 2019) to recreational cannabis consumption (Huff et al., 2021).

Close to our context, Hoffman et al. (2018) consider consumer-object assemblages in order to explore the consumer experiences of both consumers and smart objects, differentiating between three levels of experience: basic experiences, which emerge from the paired capacities of consumers and objects; aware experiences, which are filtered and processed from basic experiences with the goal of recognition, organisation, and attention; and conscious experiences, which integrate and accumulate aware experiences into subject experience and reflection.

Hoffman et al. (2018) postulate consumer experiences to be enabling experiences based on agentic self-extension (the consumer enables the assemblage) and communal self-expansion (the assemblage enables the consumer), as well as constraining based on agentic self-restriction (the consumer restricts the assemblage) and communal self-reduction (the assemblage restricts the consumer). From the interactions of consumers with smart objects and the associated affective flows, consumer experience assemblages emerge. However, little is known about consumer experience assemblages comprising multiple consumers and smart objects, and investigating such assemblages has been identified as a promising direction for future research (Hoffman et al., 2018).

Following in these footsteps, we embrace assemblage theory as a relevant theoretical foundation for studying how the integration of personal health technologies (Fox, 2017) such as smart assisted living technologies into assemblages with multiple consumers affects the capacities of the components, their consumer experiences, and their reciprocal relations. This monistic perspective allows viewing such smart objects not as passive objects but as active agents similarly to the now classical example of the seat belt alarm (Latour, 1992). As such, they can be said to have their own experiences that can be paired with the ones of the human components, providing a promising way of unpacking the interactions between multiple consumers and smart objects, as well as, ultimately, the experiences and micropolitics of the entire assemblage.

Family tutelage: between consumption and care

Elderly consumers represent a heterogeneous group with great variations due to factors such as age, health status, and socioeconomic position. With increasing age and deteriorating health and social network, elderly consumers increasingly become exposed to five types of vulnerabilities: physical, financial, social, technological, and service vulnerability (Roy & Sanyal, 2017).

The division of elderly consumption among the members of the elderly consumption ensemble (Barnhart & Peñaloza, 2013) may include consumers making choices for others. In particular, McGrath et al. (2017) find that decision-making on and adoption of assisted living technologies cannot only rely on the elderly consumers themselves but often relies also on informal caregiving of relatives and formal caregiving of care providers, who to a certain degree become part of the extended family (Barnhart et al., 2014).

More generally (not restricted to the context of elderly consumers), Liu et al. (2019) review the literature on consumers making choices for others, finding four typical cases: gift-giving, joint consumption, everyday favours/pick-ups, and caregiving. In a case of this not too dissimilar to what is studied here, Schneider-Kamp and Askegaard (2021) explore the duality of caregiving and choice on behalf of others in the context of parents making choices for their children and conclude on the ensuing ambivalent role of choice in consumer empowerment as oscillating between liberating and disciplining.

Huff and Cotte (2016) reconceptualise families with elderly consumers from an assemblage-theoretic perspective and show how such a perspective provides insights into the interactions and interdependence between elderly consumers, their relatives, and non-human components of the family assemblage such as houses and other possessions. Trees and Dean (2018) employ this conceptualisation to study how food practices in families with elderly consumers can be viewed as embodied components of loving care relationships. However, the particular issues pertaining to the agency of smart objects are missing from these contexts.

With a focus on caregiving interactions in families with one child diagnosed with type-1 diabetes, Cardoso et al. (2020) describe how family members’ dependence on caregiving of relatives requires orchestration and coordination between the remaining family members, strengthening patterns of interdependency in the family. Likewise, Dean et al. (2020) uncover how family members taking over caregiving poses tensions to their autonomy and identity construction in the liminal spaces between their previous identity as family member and an identity as caregiver. However, the mediating role of non-human agencies is absent from these analyses.

At the same time, elderly consumers strive to avoid the stigma of old age by distancing themselves from stereotypical and stigmatic images pertaining to this age group (Rosenthal et al., 2020). But less is known about exactly how the elderly consumers’ negotiation of their identity is affected by and affects their relatives’ negotiation of identity, and research in that direction from a consumer and marketing perspective is being encouraged (Cardoso et al., 2020).

One of the major images of old age is dependence on formal caregiving, which can be argued to fuel the desire of elderly consumers to engage themselves with technologies such as assisted living technologies to be able to remain living independently at home. There is a long tradition of studying the adoption of technologies by elderly consumers in the consumer research literature dating back to Gilly and Zeithaml (1985). Mostaghel (2016) systematically reviews innovation and technology acceptance among elderly consumers, identifying a need for more research from a consumer rather than healthcare perspective to update and further our understanding of how the elderly navigate the social and behavioural aspects of technological innovations.

Perez et al. (2019) have shown how technological objects affect multi-generational families and the relations within, finding a role for younger family members as family-internal technology influencers educating particularly elderly members. Recently, the discussion is broadening and recognising the relevance of considering not only the utopian but also the dystopian potential of smart digital technologies (Hofacker & Corsaro, 2020) and examining the costs of among others elderly consumers interacting with artificially intelligent technologies (Puntoni et al., 2020). As such, the discussion of elderly consumption assemblages including smart objects has been opened, but the constitution, complexities, and intricacies of these assemblages are far from exhausted.

Addressing the gaps mentioned above, the research question of this article can be elucidated as follows: How do smart objects as active rather than passive components affect the capacities and relations of the components of family assemblages with elderly consumers? In particular, we are interested to investigate to what degree consumers have joint experiences in assemblages with multiple consumers and smart objects, how the identity negotiations and transformations of elderly consumers and relatives are connected in family assemblages with elderly consumers, and how independence and dependence unfold as the capacities of components in such assemblages are affected by the integration of smart assisted living technologies.

Context

The tax-financed Danish egalitarian welfare state (Christiansen & Markkola, 2006) offers its elderly citizens the free choice of receiving formal caregiving in the form of home care in addition to caregiving available in nursing homes for the elderly (Rostgaard, 2006). This freedom of choice is described as empowering but also conditioning, referring to the de-facto establishment of a do-it-yourself culture of responsible citizens that, ultimately, furthers agendas of welfare state conditionality (Rostgaard, 2011).

Denmark generally exhibits high levels of data governance, as well as technical and operational readiness (Eichler et al., 2019). The Danish elderly consumers are European leaders in most areas of digitalisation, with, for example, 83% of citizens older than 50 years having used the Internet actively during the past 7 days compared to a European average of 49% (König et al., 2018). Danish elderly consumers are generally open to innovative assisted living technologies such as robotic companions (Klein et al., 2013), with 80% of Danish nursing homes already having embraced such technologies (Koh et al., 2021). This notwithstanding, the particular type of assisted living technology is important for consumer acceptance, with invasive technologies such as cameras being consistently rejected by elderly consumers (Offermann-van Heek et al., 2019).

The study reported in this article has been an integral part of a collaborative research project evaluating the opportunities and pitfalls of Internet of Things-enabled welfare and health technologies for vulnerable consumers. The particular assisted living technology studied was an Internet of Things-based smart alarm system, which monitors movements in the homes of elderly consumers living independently with the goal of detecting anomalies and notifying informal caregivers such as relatives of the elderly consumer via their mobile phones. The alarm system detects movement in two distinct ways. First, an alarm can be triggered by detecting a movement of falling in front of a fall detection sensor. Such sensors are typically installed in hot spots for fall events such as bathrooms. Second, an alarm can also be triggered by a movement detection sensor when it fails to detect movement during a longer period. Such sensors are meant to be placed in central locations of the home where the elderly consumer can be expected to pass with sufficient regularity during the day.

The alarm system does not expect movement during the night and can be disabled by the elderly consumer by pressing a button when they leave the home. There is another button for manually triggering an alarm. When an alarm is triggered – automatically or manually – the alarm system communicates with a mobile companion app installed on the smartphones of the family members of the elderly consumer. Should none of the family members respond sufficiently fast, the alarm is forwarded to an alarm centre where the employees attempt to get into contact with either family members or formal caregivers that could check on the elderly consumer. As a last resort, the authorities can be informed of the incident and take appropriate action.

Methods

The data for this study were collected in the years 2018 and 2019 by the first author. After an initial rapport-building stage, a combination of two qualitative methods was employed during research visits to the informants’ homes: in-depth interviews with elderly consumers, their relatives, and professional in-home caregivers; and observations in the homes of the elderly consumers and their relatives. These methods have ethnographic character as the informants and their practices were studied over a prolonged period of time and in natural settings such as the homes of elderly consumers and their relatives. Initial interviews and observations were supplemented by follow-up interviews with the elderly consumers and the most relevant relatives to capture both stability and change in the practices surrounding the alarm system. Observational and reflective field notes were written during and after each home visit. Additionally, we performed two expert interviews with the inventors of the alarm system in order to add insights on the intentions underlying the system, as well as the healthcare professional and ethical considerations embodied by the technology.

Elderly consumers living alone at home and relying on the alarm system and the respective other members of their elderly consumption ensembles constituted an intrinsically serviceable sampling frame, forestalling any necessity to explicitly define a target group based on inherently imprecise definitions of elderly consumers by either or both of age and health status. For studying the smart alarm system, we employed purposive sampling (Bernard, 2011, p. 46), allowing us to access an otherwise rather hidden consumer segment. A potential pool of elderly consumers was built using referrals and sampled to ensure some variety regarding the ages and family constellations of the elderly consumers. We excluded informants without sufficient experience with the alarm system considered in our study.

The recruited sample of 18 informants consisted of 4 elderly consumers, 7 relatives, 5 in-home caregivers, 1 medical expert, and 1 technical expert. Table 1 provides an overview of the 14 informants comprising the four main assemblages studied, with informants identified relative to the elderly consumer. All elderly consumers and their relatives were active users of the alarm system, yielding a total of 11 direct users. The remaining 4 informants were the 2 experts and 2 caregivers not related to the four main assemblages.

Table 1. Overview over the informants comprising the four assemblages studied.

	Assemblage 1	Assemblage 2	Assemblage 3	Assemblage 4
Informants (gender, age)	elderly (F, 86)*+ daughter (F, 62)*+ son-in-law (M, 64)+ grandson (M, 27) caregiver (F, 56)	elderly (F, 84)* son (M, 65)* caregiver (F, 41)	elderly (M, 81)*+ daughter (F, 59)*+ son-in-law (M, 63)+	elderly (M, 77)+ son (M, 52)+ caregiver (F, 47)
Length of use	> 1.5 years	>1.5 years	>1 year	<6 months
Interviews	7	5	5	3

*Informants with whom a follow-up interview was performed.

+Informants who were interviewed and observed at home.

The primary data consist of the recordings and transcriptions of the 24 interviews, as well as observational field notes. The 14 initial in-depth interviews lasted between 73 and 112 minutes and were conducted using a thematic rather than structured interview guide to retain a maximal amount of openness to informants’ perspectives and interests in the subject matter. The 6 follow-up interviews were conducted 3–5 months after the initial interviews and lasted between 37 and 61 minutes. Observations were performed in connection with all home visits, commencing during the rapport-building stage before the interviews and typically outlasting the interviews by approx. 30–45 minutes. The field notes relied on comprehensive note-taking (Wolfinger, 2002) and contained both analytical and reflexive elements.

We obtained consent for recording interviews, as well as for including short, non-identifiable, translated excerpts from interview transcriptions in publications. We refrained from assigning pseudonyms and opted for avoiding the inclusion of sensitive information such as health status. Furthermore, we implemented Saunders et al. (2015) best practice recommendations for avoiding re-identification by merging related and splitting unrelated accounts. Details irrelevant to the analysis were left out or carefully redacted to maximise informant anonymity. We reported the data collection for the study to the Danish Data Protection Agency as part of the collaborative research project. Ethical clearing by the Danish National Committee on Health Research Ethics is not required for this type of study.

In the analysis phase, the authors collaborated closely in the spirit of a hermeneutic community, with the first author taking the expert perspective and the second author challenging this perspective, forming a hermeneutic ‘dialogic community’ (Arnold & Fischer, 1994). We followed an approach closely mirroring phronetic iterative data analysis (Tracy & Hinrichs, 2017), starting off with identifying first-level descriptive and second-level analytic codes in the spirit of grounded theory (Corbin & Strauss, 2014) and then iteratively refining codes, concepts, and categories by interpreting the data through a number of emerging theoretical lenses. At the saturation stage of this analysis process, we had employed assemblage theory for structuring and theorising change (Giesler & Thompson, 2016), utilised consumer-object experience assemblages to understand the interactions of consumers and smart objects (Hoffman et al., 2018), and mapped affective flows between components to stratify these consumer experiences (Feely, 2020).

Findings

In this section, we explore elderly consumption assemblages with smart alarms by identifying their components and relations in the first subsection, mapping the affective flows in the two main consumer experience assemblages in the following two subsections, and considering processes of change and stabilisation in the final subsection.

When technology moves in: from elderly consumption ensemble to assemblage

Every invention has a history, and such a history is often deeply personal to the inventor. In the case of the alarm system studied in this article, a particular critical incident in the family of the inventor started the whole story:

Back in 2012, my 80-years-old uncle was a widower but still lived alone in his home. One day, he had an accident. He fell in the bathroom late at night, and he broke his hip. He actually lay there for 13 hours before he was found in a miserable condition, which also turned out to lead to a less than optimal outcome of the hip surgery. He had pneumonia, he was severely dehydrated, and he was in deep shock. He thought he was going to die!

This incident proved to be a catalyst for reassessing the role and potential of material objects in the consumption of care inside the family:

That was also quite a shock for the family, and we started thinking and discussing how we could avoid such situations in the future. And we thought, well, nowadays, homes can take care of light, temperature, perhaps even lock the door – why shouldn’t a modern home be able to care for its inhabitant? So, we came up with the idea of making a smart alarm system for monitoring movement in homes.

The initial idea was to create an alarm system that does not depend on the elderly consumer being able to press an alarm button but instead relies on smart sensors to detect movement or its absence.

By integrating such a smart alarm system, the elderly consumption ensemble comprised of the elderly consumer, the relatives, and formal caregivers, thus, is extended to an elderly consumption assemblage that in addition to the human components also comprises a number of non-human material and immaterial components including shared family values and possessions, the home of the elderly consumer, the alarm system, and the mobile companion app.

Figure 1 visualises the path towards this assemblage from (a) Barnhart and Peñaloza (2013) elderly consumption ensemble via (b) Huff and Cotte (2016) evolving family assemblage to (c) the elderly consumption assemblage with the alarm system, clearly illustrating the combined human, institutional, and material agencies within this assemblage. The illustrations of the ensemble and the assemblages are models and, thus, have schematic character: actual existing assemblages of families with seniors might, for example, include multiple relatives and no formal caregivers. Furthermore, such assemblages should be viewed as fluid and can be expected to change regarding their composition and the nature and strength of their relations over time, with different components attaching, detaching, and reattaching to the assemblage.

Figure 1. An elderly consumption assemblage including the alarm system.

There are three types of relations in the assemblage: consumer-consumer relations, consumer-object relations, and object-relations. The consumer-consumer relations consist of the informal caregiving and family relations between the elderly consumers and the relatives, the formal caregiving relations between the elderly consumers and in-home caregivers, and the collaborative and coordinative relations between relatives and in-home caregivers. The consumer-object relations include the varying relations of the human components to shared family values, possessions, and the homes of the elderly, with, for example, the elderly exhibiting the strongest relation to their homes, which relatives and formal caregivers only visit sporadically and temporarily. Furthermore, the elderly consumers form consumer-objects relations with the sensors of the alarm system while the relatives form such relations with the companion app. There is also potential for such relations between in-home caregivers and the alarm centre. Finally, the alarm systems comprising sensors and communication units form object-object relations with the companion apps and the alarm centre, exchanging information about alarms and their handling.

My new roommate: the elderly consumer-alarm system experience assemblage

The elderly consumption assemblage is an instance of a consumer-object assemblage, from which multiple consumer experience assemblages emerge through the interactions of consumers with smart objects. One of the focal such experience assemblages is the elderly consumer-alarm system assemblage.

In this assemblage, we found closely paired consumer experiences of elderly consumers and the alarm system on the basic and the aware level. On the level of basic experience, the elderly consumer’s capacity to pass by the alarm is paired with the alarm system’s capacity to experience a change in voltage produced by the motion sensor. On the level of aware experiences, the elderly consumer is aware of the alarm while passing it as well as of the need to pass it. This is paired with the alarm system’s processing and filtering of the measured data from the sensors in order to discriminate motion events, categorise them as critical or non-critical movements, and be able to react to them based on the alarm system’s internal state on how much times has passed since the last non-critical movement.

The elderly consumers additionally have the ability to accumulate experiences with the alarm system over many episodes and reflect upon these in order to adjust their everyday practices. The (paired) capacities, affective flows, and resulting consumer experiences at the three different levels are summarised in Figure 2.

Figure 2. Stratification of consumer experiences emerging from paired capacities in the elderly consumer-alarm system experience assemblage.

Reflecting on our field notes regarding observations in the elderly consumer’s homes, generally, an impression of a closely integrated relation between elderly consumers and alarm systems emanates, with the field notes summarising that ‘the device had become something like a new roommate, or even a family member’. The elderly consumers were found to possess a precise awareness of the instructions on how to use and integrate the alarm system into their everyday practices, thereby feeling empowered to continue living independently as expressed by one of the elderly users of the alarm:

What do you mean? I am not that old. I am only 86 years old. Of course, I can figure out how this alarm works. And why should I not be able to live in my home, if I can still manage on my own?

The alarms system and being able to use it affects their negotiation of consequences of age, as it lessens their feelings of dependence on others and, thereby, allows them to deflect an old age identity.

The elderly consumers are to some degree aware of the limits of the alarm and reflect upon, adjust, and attempt to manage their everyday practices in order to avoid false alarms that unnecessarily create disturbances among their relatives. Situations of false alarms are, of course, not completely unavoidable. One of the elderly consumers engages with repeated episodes of false alarms:

We had a bit of a problem with it [the alarm] because, apparently, I’ve gone to bed too early according to what it was set to. When I do that, it gets confused in the mornings, when it hasn’t seen me move for more than 10 hours.

After discussing with her daughter whether to get the alarm adjusted to account for such situations, they together decided to assume that alarms in the early morning likely are false alarms and do not require immediate attention. The daughter describes how she deals with such alarms:

I just switch my phone into silent mode. I do hear that it makes a vibrating sound, but my husband doesn’t wake up, and I just respond ‘ok’ without acting further. I know it is because she went to bed earlier.

Here, the consumer’s ability to have conscious experiences offsets the alarm system’s inability to have such experiences. Ideally, a more advanced smart alarm system could, of course, accumulate aware experiences in the form of categorised events with the goal of self-adjustment to reduce the number of expected future false alarms. For the time being, however, the burden of avoiding false alarms seems to firmly rest with the relatives.

Some of the elderly consumers found it more challenging to integrate the alarm system into their everyday practices and avoid situations of false alarms. One relative summarises the main mechanism behind such alarms succinctly:

She [the grandmother] forgets to walk by the alarm or to switch it off, and then she doesn’t remember it later.

In addition, these elderly consumers sometimes also actively cause false alarms as related by another relative:

The other day I got a scare. I knew he was going out, but he had pressed the wrong button. He pressed the emergency button instead of the off button. But I knew he was attending a birthday, so I turned the alarm off.

These issues can be traced back to a lack of aware experiences from the side of the elderly consumers. Ideally, a more advanced smart alarm system could ‘nag’ such elderly consumers to enable their aware experiences and improve their pairing on the level of consumer experiences. As of now, a certain level of co-involvement in the lives of the elderly to be able to interpret alarms seems unavoidable.

Always on duty: the relative-companion app experience assemblage

Another focal consumer experience assemblage emerging from our elderly consumption assemblage is the relative-companion app experience assemblage. As with the elderly consumer-alarm system experience assemblage, we found that basic and aware experiences were closely paired. On the level of basic experiences, when the companion app receives an alarm, it creates a notification message and a notification sound that provides an auditory experience to the relative. On the level of aware experiences, the relative becomes aware of the alarm through their categorisation of the notification sound as signifying a notification and through subsequent recognising of the notification message as an alarm. Through an interaction with the phone, the companion app in turn becomes aware of the relative’s experience of the notification.

The relatives also have conscious experiences as they accumulate episodic experiences with both false and true alarms, investigating the underlying reasons and reflecting on ways of avoiding false alarms in the future. The companion app cannot really be said to be conscious or self-aware. Ideally, it could however reflect upon its awareness of a relative having reacted or not, making other relatives aware of whether the elderly consumer has received the help needed and informing formal caregivers and authorities in cases, where the elderly consumer has not received help. The (paired) capacities, affective flows, and resulting consumer experiences at the three different levels are summarised in Figure 3.

Figure 3. Stratification of consumer experiences emerging from paired capacities in the relative-companion app experience assemblage.

When we contrasted the interviews of relatives with our field notes, it became quite apparent that most, if not all, of the relations between relatives and the companion app were formed without considerable effort. In the words of one of the relatives:

The [alarm] system just works. It really does. We are so happy with it!

The most obvious explanation of the relatives’ satisfaction would be that relatives tend to be relatively younger consumers with extensive experience regarding smartphone usage from other consumer-mobile app assemblages and ensuing familiarity with such technology. And indeed, all of the relatives among our informants confirmed that they use smartphones regularly and found the mobile companion app to be sufficiently simple to use.

That is not to say that the relatives did not experience challenges surrounding the actual use of the companion app. One prominent challenge is the need to keep the smartphone ready to receive alarms at all times. The daughter of one elderly consumer speaks on behalf of her husband, her son, and herself:

It is good for us – now we always have charged phones! [laughing] We had to learn the hard way that at least one of us needs to have a charged phone. One that is not switched to silent mode. We have to make sure of that! We are more relaxed about the alarm. It is plugged in [power outlet] and always has power.

Similar challenges surface in connection to other technical aspects of the companion app such as updates and their compatibility with the smartphones, which pose a significant disturbance to the relatives’ everyday practices:

There was an update to the app, and suddenly it would not run on my phone anymore. I had to contact the company for technical help. And while they fixed the problem, for a couple of days, I was checking on my mother several times a day. I was so worried!

This dependence on the battery and functional state of relatives’ smartphones is reminiscent of the ‘hysteresis of the battery’ (Robinson & Arnould, 2020), observed in the context of consumer mobility. The major difference in our study is that having smartphones run out of battery is not merely an inconvenience but rather poses a direct threat to the safety of the vulnerable elderly consumer living independently.

Any assemblage of components is naturally an abstraction, as components cannot be assumed to be atomic but rather as themselves being assemblages of subcomponents. The safety considerations surrounding the relative-companion app experience assemblage expose the inadequacy of this abstraction in relation to the everyday practices of the relatives. Unfolding this assemblage, we find that in addition to the relative and the companion app, it also has to include at least the phone, its battery, and its charger. In addition, there are typically multiple relatives involved. These relatives interact not only with the app but also with each other, e.g. negotiating who reacts to a genuine alarm and ensuring that at least one relative is always able to receive alarms. The assemblage resulting from zooming in on the relative-companion app experience assemblage is visualised in Figure 4, with the additional components and consumer-object as well as object-object relations in dark grey.

Figure 4. Zoomed-In view of the relative-companion app experience assemblage.

The alarms as a double-edged sword: enabling and constraining experiences

Here we return to the point of departure of this article, namely the issue of dependence and independence. At first sight, adding the alarm system to the elderly consumption assemblage resembles an enabling consumer experience aimed at increasing the independence of elderly consumers.

Interestingly, while the desire for independence emanates from the elderly consumers, in most cases the agency regarding integrating the alarm system into the assemblage lies with the relatives:

It was my decision. She [the elderly relative] didn’t want to move to a nursing home. So, I said, then we need to instal this alarm. Such that we can care for her!

By adding the alarm system, the relatives imbue the assemblage with new affective flows and ensuing capacities of caring for each other. These new capacities enable the elderly consumers to remain living in their homes:

I don’t want to go to a nursing home. I’m good here [at home]. I have nice neighbours.

By being able to stay in their homes, the elderly consumers also are enabled to retain their established relations in their respective local communities.

Analysing these enabling consumer experiences through Hoffman et al. (2018) framework of agentic and communal experience, we identify agentic self-extension on the part of the relatives that enable the whole assemblage, as well as communal self-expansion of the elderly who are enabled in this way to retain some degree of independence and social embeddedness.

Other elderly consumers experienced the alarm system as a constraining obligation forced upon them by their relatives rather than as an enabler of independent living, and were unwilling to integrate the alarm system into their everyday practices:

I want to deliver this [the alarm] back. I feel too overwhelmed and controlled by the technology.

In a further discussion with this particular elderly consumer, an impression emerges that he experiences the addition of the alarm system to the elderly consumption assemblage as constraining his freedom to do as he pleases. The field notes condense his experiences in the observation that ‘he was feeling that the alarm was more a way to control his movements than a way of caring for him’.

While such an overwhelmingly negative experience represents a singular case among our informants, we also found there to be constraining consumer experiences of elderly consumers who were satisfied users of the system. Being aware of the alarm, elderly consumers felt an obligation to their relatives to avoid false alarms by ‘pleasing’ the alarms:

Sometimes, when I watch television, I suddenly realise that I have been sitting for more than three hours. So, I need to stand up and walk by the alarm in the corridor. To make sure it knows I am ok. I would rather stay on the sofa for an hour or two more, but I don’t want to scare my family.

While regular physical activity can be assumed to be beneficial for the elderly consumers’ physical well-being, the burden of avoiding false alarms due to inactivity increases their mental load. Similarly, several elderly consumers felt obliged to remember to switch off the alarm when they leave their homes.

Notably, the integration of the alarm system into the elderly consumption assemblage also increases the mental load of some of the relatives. Even though the very same relatives have typically been the agents of change by deciding to integrate the alarm system, they often experience being constrained by feeling obliged to be on duty constantly:

But what if my phone has died, and I do not get the alarm? Then she could be lying half-dead, and I wouldn’t know! So, I always have to keep the phone ready, and I also have to be careful how much I drink [alcohol consumption in social context]. I cannot afford not to hear the alarm!

Such remarks refer to the difference between experiencing choice and destiny, not unlike ‘tyranny of management’ evoked by Askegaard et al. (2002) in their analysis of ‘necessities’ induced by the possibility of cosmetic surgery. In the terms of Hoffman et al. (2018) framework, we might consider the constraining experiences of both elderly consumers and relatives as instances of communal self-reduction, where the assemblage as a whole constrains their capacities.

While there is an alarm centre, which has the task of contacting formal caregivers in cases where the relatives do not react sufficiently fast, the relatives seem to consider it unacceptable to rely on this feature. This is in line with the experiences of the formal in-home caregivers interviewed, even the ones that are assigned to multiple assemblages as emergency contacts:

Normally, the relatives take care of everything. False alarms or when it actually matters. I know that I could be contacted, but this has not happened even once during the last year.

This responsibility for being able to receive alarms and being ready to react to them is alleviated only by the possibility to share this obligation with other relatives:

My son also has the app installed. So, the responsibility is split into two. That is great! But, of course, we have to coordinate that at least one of us can do it [follow up on alarms] at all times.

Temporarily detaching from the elderly consumption assemblage, e.g. by switching off the phone, potentially constrains the capacities of the whole of the elderly consumption assemblage. This type of agency is only an option for relatives in practice when the detachment is essentially without effect as other relatives remain attached and still enable the capacities of the assemblage needed to ensure the safety of and care for the elderly consumer. Viewed through the framework of agentic and communal experience, agentic self-restriction of relatives is not necessarily an acceptable strategy of coping with their self-reduction.

Discussion

Having investigated how smart objects as active rather than passive components affect the capacities and relations of the components of family assemblages with elderly consumers, in the following three subsections, we discuss the joint and disjoint nature of consumer experiences in consumer-object assemblages with multiple consumers, the identity negotiations and transformations of elderly consumers and their relatives, and the hidden costs of supporting the independent living of elderly consumers through smart assisted living technologies.

Multiplicity of consumer experience assemblages: separate but together

The elderly consumption assemblage studied in this article constitutes an interesting instance of an assemblage of multiple consumers and smart technologies. Hoffman et al. (2018) encouraged empirical investigations into such assemblages, imagining scenarios such as the interaction of multiple consumers with a smart object like Alexa Voice Services. An open question regarding assemblages involving multiple consumers is to what degree these consumers have joint experiences with the smart objects. In the case of multiple family members interacting with Alexa Voice Services, the consumers typically cohabitate in the same home and experience not only their own but also the other consumers’ interactions with the smart object. Thus, they could easily be viewed as forming one joint consumer experience assemblage with the smart object.

Unlike this hypothetical scenario, the one we studied empirically in this article involves multiple heterogeneous consumers, i.e. consumers filling different roles within the assemblage, who attach and detach at different points of time at different physical locations to different components of the overall assemblage. While the elderly consumers have regular basic and aware consumer experiences paired with the sensors of the alarm system in their homes, the relatives have basic and aware consumer experiences paired with the companion app on a less regular schedule at whatever physical location they happen to be at. Furthermore, their capacities exhibit significant levels of heterogeneity in parallel to the heterogeneity of their roles as they attach to different assemblages. On the level of basic and aware experiences, these consumers arguably form their own disjoint consumer experience assemblages with their respective smart objects as part of the overall alarm system.

Notably, on the level of conscious experience, we have found that the consumer experiences of the elderly and the ones of relatives are connected as they jointly reflect on topics such as the frequency of alarms, the reasons for false alarms, and strategies of avoiding false alarms. On this level, the multiple heterogeneous consumers arguably form a joint consumer experience assemblage with the smart objects, at least when they simultaneously attach to the whole consumer-object assemblage. In this way, our findings not only provide empirical evidence to Hoffman et al. (2018) concept of consumer-object assemblages but also demonstrate that, for the case of multiple consumers, whether consumer experience assemblages should be viewed as joint or disjoint may depend on the level of the consumer experience under consideration.

Formal vs informal care: just a technicality?

Just as the experiences of consumers are connected on some level, so are the negotiations of identity of the elderly consumers and their relatives. Rosenthal et al. (2020) advocate a destigmatisation of old age in advertising through a focus on realistic rather than overly positive portrayals of old age. Smart assisted living technologies might provide an alternative means of destigmatisation by transforming the reality of the elderly consumers rather than their portrayal, aiding the elderly consumers in their negotiation of old age identity as independent and living at home.

This is in and by itself a positive development – but nothing exists in and by itself. Policymakers, to some extent, can be expected to endorse and encourage the interest in independent living as part of their neo-liberally inspired quest to optimise cost and resource allocation in the care for their elderly citizens (Rostgaard, 2011). Smart assisted living technologies impact the balance between consumer autonomy based on exercising consumer choice and access to care by enabling a shift from formal care provided by professional in-home caregivers to informal care provided by relatives, providing further evidence for a fundamental ambivalence connected to the search for autonomy in the care sector (Schneider-Kamp & Askegaard, 2021).

In our study, we observed a blurring of the boundaries between informal and formal care. Informal care had an episodic character before the introduction of the alarm system, with the relatives being relatively unencumbered by the elderly consumer’s demand for care for most of the time. With the smart alarm system, informal care has lost some of this ephemerality (Diaz Ruiz et al., 2020) and become a more systematic and, most importantly, more pervasive part of the relatives’ lives. This finding provides further empirical support for the shift of responsibility observed in other settings of hyperconnectivity (Schneider-Kamp & Kristensen, 2019). The greater extent to which relatives take over care functions and the ensuing tensions to their identity construction yield not only a liminal space between previous identities as family members (Dean et al., 2020) but also a liminal space between informal and formal care.

Here, by liminality, we refer to the observation that relatives are neither formal caregivers, as they perform care without contracts and remuneration, nor traditional informal caregivers, who typically care for their elderly relatives at certain pre-mediated points of time or when demand arises. While the role of a 24/7 informal caregiver is not dependent on the technology, our argument here – in line with our assemblage theoretical perspective – is that the presence of the technological device in and by itself alters the way this role as a 24/7 caregiver is perceived. The relative is no longer ‘just’ a relative but an actant in a new techno-social assemblage, oftentimes with both public and private agency. Consequently, the alarm system effectively turns them into a ‘new type’ of caregiver that is on duty 24/7, eclipsing the demands on the availability of formal caregivers, who typically cover care 24/7 through a shift system, if at all. While Cardoso et al. (2020) see the informal caregivers in their study exhausted due to the demands of orchestrating and delivering care in the family, we find that the mere act of constantly being primed to do so can likewise be exhausting.

Daughters of the elderly seem to bear the brunt of this liminal identity, with the existence of a gender bias regarding family caregiving that continues to be a part of both public and academic debate (Barnhart et al., 2014; Cardoso et al., 2020; Epp & Price, 2018). Caregiving is widely viewed as an inherently female responsibility, and Trees and Dean (2018) discuss how the daughters of the elderly consumers are part of a ‘sandwich’ generation, who is exposed to norms of being a good woman by not only striving to be a ‘good parent’ (Cardoso et al., 2020) to their children but also to be a ‘good child’ to their ageing parents.

The consumption of smart objects like Alexa Voice Services for the purpose of home automation (Hoffman et al., 2018) has been suggested to be predominantly ludic (Seregina & Weijo, 2017), with the consumers free to opt out of the assemblage through self-restriction at any time. While we acknowledge this finding, we also think of it as transitional, since we foresee the development of a deeper dependence and existential intertwinement with smart objects in the future, not necessarily removing but definitely adding the ludic character. Our findings, therefore, contribute with a new layer to the consumption of smart objects, as the liminal identity as a more-than-informal caregiver and good child effectively constrains the ability of the relatives to even temporarily opt out of the elderly consumption assemblage. Such self-restriction would be perceived as inappropriate in the face of matters of safety, life, and death. It is in this new layer that we begin to see the hidden costs of smart assisted living technologies, which Puntoni et al. (2020) want us to pay attention to, making themselves felt.

The burden of care: independence through smart assisted living technologies?

Cardoso et al. (2020) refer to the dependence relation between informal caregiver and care-receiving family member as a precariously balanced interdependence. While the integration of smart assisted living technologies might increase relatives’ capacities for informal care through a spatiotemporal extension of the elderly consumption assemblage, we also demonstrated that it may decrease the relatives’ capacities in their other everyday assemblages. Not only might consumer autonomy be considered an illusion in the face of technological complexity (Schneider-Kamp & Askegaard, 2021), but we also found that the choice of relatives to have their capacities reduced fostered disciplining aspects, and that these were offsetting the spatiotemporally liberating ones to some extent.

The disciplining aspects of the elderly consumption assemblage studied emerged as a result of how the smart object affected the micropolitics (Fox, 2017) of the assemblage. The smart alarm assumed the position of master in the assemblage (Schweitzer et al., 2019), particularly as it pertained to the relatives who took up a more reactive and subservient position as a consequence. Where previously, relatives were actively checking on the elderly consumers and their well-being, this task had now been delegated to a smart object supposed to activate the relatives. The smart object had, thus, taken over the role of a family manager (Letheren et al., 2019) of the distanced caregiving relationship between the elderly consumers and their relatives.

On this backdrop, we argue that full independence of the elderly consumer necessarily remains an illusion, as dependence relations just shift to other agents such as relatives, or even the batteries of the relatives’ smartphones. While the elderly consumer’s perception of independence increased due to a perceived decrease along one of the multiple dimensions of vulnerability, we found a significant increase in the burden of care being placed on the relatives of the elderly consumer. Dean et al. (2020) postulated that the liminality of the burdened relatives’ identity might have a detrimental effect on their mental health. Our findings shed light on mechanisms that might contribute to this effect.

We agree with Cardoso et al. (2020) that ‘ready-made market-based solutions’ such as the smart alarm system do not alleviate the need for orchestration in family assemblages, and that these assemblages jointly surpass the capacities of the individual consumers comprising them. We urge caution, though, to consider the hidden costs and of employing inherently political tools such as smart objects (Puntoni et al., 2020) to tap into the ‘creative capacity of families, and that of leading orchestrators, to find and adapt market resources to their own care needs and orchestration styles’ (Cardoso et al., 2020, p. 937).

While the elderly consumers experience increased autonomy, the relatives find themselves trapped in a micropolitical spiderweb, unable to escape from constantly being on duty and with digital disconnection an unattainable luxury (Humayun & Belk, 2020). In this way, the experience of ageing and its restrictive influence on everyday practices is extended from the elderly consumer to the relatives in their liminal role as neither-formal-nor-informal caregivers. Ultimately, this can be seen as a tension between the liberative potential of technological progress and a virtual regression to the gridlocked interdependence of the multi-generational family model of centuries past and, thus, another instantiation of a contemporary ‘tyranny of management’ (Askegaard et al., 2002).

Implications and future directions for research

The elderly consumption assemblage provides a fruitful blueprint for researchers to study the integration of smart non-human components into care assemblages, with the potential to inform social and health policy regarding the balance between formal care provided by care professionals and technology-supported informal care through relatives. Marketers can gain valuable insights into a number of critical issues surrounding the growing market for (smart) assisted living technologies. They can benefit directly from an increased understanding of whom to market these products and associated services to and are enabled to consider how such products might place the burden of care on relatives. Such a multiple-user perspective further helps to understand potential barriers to market success entailed by the micropolitics of family care.

Our study is limited by the sample size, even if the four assemblages studied were studied in depth. Furthermore, the type of smart objects that we studied did not have a potential for aware consumer experiences. While awareness is not necessarily fully achievable yet from a technological point of view, existing and future ‘smarter’ objects might potentially affect assemblages in different or additional ways.

We believe that our findings underline some fundamental ambivalences concerning the relations between technology and health (Lianidis & Askegaard, in press). Future research could supplement the study of consumer experience in the elderly consumption assemblage and the ensuing micropolitics with the study of the emerging ethics of consumer-object assemblages in general, and ethics of care in the elderly consumption assemblage in particular. An assemblage-theoretical perspective could also help to understand the potential of technological solutions such as smart objects to ameliorate the varied forms of vulnerability encountered among the heterogeneity of consumers.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The data collection for this study was part of the collaborative research project VIND co-funded by the European Regional Development Fund and the Danish state. The funding sources had no involvement in the study design, in the collection, analysis, and interpretation of data, in the wrtiting of the article, or in the decision to submit it for publication.