Smart home technology adoption in Denmark: Diffusion, social differences, and energy consumption

In recent years, smart home technologies (SHT) have been increasingly entering the domestic sphere. Although previously designed for luxury homes (Darby, 2018; Gram-Hanssen & Darby, 2018), a range of SHTs are now designed for ordinary everyday activities such as cleaning and cooking (Aagaard, 2022). Defined as internet-connected technologies to provide services within the domestic sphere, and to the energy system (Gram-Hanssen & Darby, 2018), SHT covers many everyday practices, devices, and services, yet it seems difficult to live up to the underlying expectations related to energy efficiency, comfort, convenience, and controllability, etc. (Hargreaves et al., 2018; Strengers & Nicholls, 2017). The optimistic potential of SHT seems to originate in its initial purpose of being designed for luxurious and convenient living (Darby, 2018) with industry visions of passive users (Aagaard, 2021), and when implemented in mundane, dynamic, and complex everyday practices, these promises risk becoming disruptive and inconvenient elements of everyday life (Hargreaves et al., 2017a), for example by necessitating ‘workarounds’ to accommodate the technology in routine everyday practices (Larsen & Gram-Hanssen, 2020).

Diffusion of SHT is thought to be key to achieving energy efficiency, and energy reductions, and generally contributing to decarbonization and future energy transitions (Li et al., 2021; Sovacool & Furszyfer Del Rio, 2020). Moreover, SHT plays a key role in visions of smart grids and smart cities (Lund et al., 2017), and the research interest in smart home adoption appears to be increasing (Li et al., 2021).

Although the SHT diffusion rate has been described as low (Marikyan et al., 2019) and SHT might not be perceived as mainstream (Chang & Nam, 2021), the SHT market appears to be expanding (Sovacool & Furszyfer Del Rio, 2020). According to Eurostat data from 2022 (ISOC_IIOT_USE), around 10% of European households have SHT¹ in the form of home appliances such as robotic vacuum cleaners, home security systems, and energy management systems. This percentage is higher for Danish households than the European average, with 20% having smart home appliances, 17% with home security systems, and 15% with energy management systems. According to the Eurostat survey, 55% of households across EU countries have smart TVs, the most widespread SHT, while this figure is 66% in Denmark. Denmark can therefore be considered among the frontrunner nations when it comes to the uptake of SHT, alongside Norway and the Netherlands. However, this seems to be a rather recent trend. Based on survey data conducted by Statistics Denmark, the adoption of smart TVs increased from 40% in 2015 to 70% in 2021, robotic vacuum cleaners increased from 9% in 2015 to 12% in 2021, and intelligent control of heating or electricity increased from 16% in 2020 to 22% in 2021 (Hohnen & Hansen, 2022). This indicates that the popularity of smart technologies has increased recently, although, from an energy consumption perspective, the adoption of smart heating control appears as the most interesting increase, often as part of home energy management systems (HEMS) (Mahapatra & Nayyar, 2022).

Whether SHT is expected to improve efficient and flexible energy use, underpinning convenient and comfortable everyday living, or improve health, safety, and entertainment in future homes, the diffusion across various types of households and the impact on household energy consumption are important topics to investigate.

This paper investigates the research questions: which households tend to acquire SHT and what impact do these technologies have on their energy consumption? This is investigated using data from a representative survey of 1,468 Danish households in combination with administrative data.

This paper thereby addresses three aspects of SHT adoption: the diffusion of SHT across types of appliances and services, the differences in adoption of SHT across socio-demographic groups, and the relationship between SHT and energy consumption. The situation in Denmark offers a unique opportunity to combine survey data with energy consumption data and register data. The register data comes from administrative records, where information on socio-economics and housing for all people living in Denmark are registered using a unique identification number.

This data allowed analysis of diffusion, differences in uptake, and the relationship with energy consumption for the same sample of households. However, the case of Denmark and the use of survey data limit the study to investigate the smart appliances that were asked for in the survey and to a certain extent to the context of a prosperous Nordic welfare society.

Despite the limitations, the analysis contributes new evidence on SHT adoption in three ways. First, rather than investigating intentions to buy, awareness of products, or specific preferences, the variables indicate ownership of SHT, which reflects actual purchases. Although the data are self-reported, this can contribute to better evidence of which households tend to acquire SHT. Second, despite minor biases, the survey sample is representative of the total population in Denmark. This means that the analysis reflects ‘ordinary’ households rather than the likely biased samples in pilot studies or trials. Thus, the sample might reflect what Rogers (2010) describes as an early majority, which follows the stage of early adopters, Third, the analysis is based on more extensive data than previous studies based on surveys (e.g., Arthanat et al., 2019, 2020; Parag & Butbul, 2018; Sanguinetti et al., 2018a, 2018b; Shin et al., 2018) in terms of the number of observations and variables. This is also one of the first studies to specifically investigate the Nordic region.

To give a background for this study, this section presents relevant literature on the definition, diffusion, and adoption of SHT. It may not be clear exactly what the term SHT refers to. When defining SHT, three features seem to recur. The first feature is the connectedness of devices. To communicate, technologies must be digitally connected in some way, for example being integrated into systems or networks, including the Internet of Things (IoT). In particular, information technology such as smartphones, sensors, touchscreen panels, or digital voice assistants like Google Home and Amazon Alexa is important for devices to connect with lights, heating systems, and other domestic appliances (Tirado Herrero et al., 2018). This also includes how a technology is set up to control, monitor, or communicate with another device, for example, to anticipate and respond in certain ways (Aldrich, 2003).

The second feature is the residential location. This simply means that SHT, or the services provided, are located within the domestic sphere and residential housing, but can be controlled by residents or others outside the home (Gram-Hanssen & Darby, 2018).

The third feature is the services that SHT provides or is expected to provide. Combining the four features, SHT refers to the automation of connected technologies aimed at providing services within the domestic sphere or as Furszyfer Del Rio (2022) defines it: SHT refers to “[…] appliances that need to be digitally connected, provide some degree of automation and deliver enhanced services to occupants”. When considering SHT as a part of smart grids, the complex communication networks and remote monitoring may also provide services to electricity system operators (Gram-Hanssen & Darby, 2018), for example, the technical potential to deliver energy saving or time-shifting of energy demand (Ford et al., 2017).

Previous literature mentions a list of services related to SHT. Services related to energy demand are among the most well-described (Li et al., 2021), for example in terms of savings, flexibility, and control of energy demand, including electricity use, heating, and cooling (Darby, 2018; Shin et al., 2018; Sovacool & Furszyfer Del Rio, 2020). Services related to SHT also include safety and security (Shin et al., 2018), convenience, controllability, and comfort (Strengers, 2013; Strengers & Nicholls, 2017), entertainment and enjoyment, as well as more generally financial benefits, aesthetics, functionality and privacy (Sovacool & Furszyfer Del Rio, 2020). Finally, SHT can also symbolize wealth or a commitment to sustainability (Schill et al., 2019).

This variety of services reflects how the diffusion of SHT holds many promises (Skjølsvold et al., 2015; Sovacool & Furszyfer Del Rio, 2020; Strengers & Nicholls, 2017). The technology industry seems to promote visions of increasing levels of comfort and convenience for (passive) users (Aagaard, 2021; Strengers & Nicholls, 2017), whereas energy policy seems to envision SHT as a tool to provide energy efficiency and flexible energy use (Strengers & Nicholls, 2017). For example, the European Directive on energy efficiency in force (European Parliament 2018) incites member states to promote smart technologies (Article 2a) using a smart readiness indicator to measure a building’s capacity to use information and communication technologies to accommodate the needs of occupants and the energy grid (European Parliament Directive, 2018, recital 30). However, Janhunen et al. (2019) argue that cold-climate buildings do not fit well with the smart readiness indicator, as it is defined in the EU directive.

SHT requires adaptation and familiarization from householders (Hargreaves et al., 2017a), which might lead to an initial interest in functionalities, but if the practical benefits are not clear, SHT risks being perceived as gimmicky and unnecessary (Sanguinetti et al., 2018a, 2018b). SHT may also increase household labor (Strengers & Nicholls, 2017) and lead to unequal gender roles (Aagaard & Madsen, 2022; Furszyfer Del Rio et al., 2021), while its implementation relies on specific (technical) competencies (Aagaard et al., 2023; Larsen & Gram-Hanssen, 2020; Madsen et al., 2023). In addition, SHT might contribute to increasing levels of energy consumption (Darby, 2018; Hargreaves et al., 2017b; Peffer et al., 2011; Strengers & Nicholls, 2017), yet the direct effect of SHT on electricity or heating consumption is still unclear.

Several studies have described the adoption of SHT using the theory of diffusion of innovation by Everett M. Rogers (2010). Following this theory, SHT adoption can be described in five steps representing different groups of technology adopters (from innovators to early adopters, early majority, late majority, and laggards). Pantzar (1997) presents an alternative to Rogers’ diffusion of innovation theory by focusing on how SHT adoption can be seen as a process of domestication of everyday life technologies. This follows three steps, where SHT starts as fashionable objects of desire, then becomes objects that are reasonable to acquire in terms of functionality, and finally, objects of routine, where acquisition is ordinary and requires no justification (Pantzar, 1997). The element of domestication in SHT adoption links to studies on the smart home. Hargreaves and Wilson (2017) distinguish between three views on the smart home (functional, instrumental, and socio-technical), and Hargreaves et al. (2018) outline four distinct motivations to adopt smart home products: 1) saving energy, 2) interest in new technology, 3) protecting the environment, and 4) a desire for improved control. Moreover, Gram-Hanssen and Darby (2018) describe four perceptions of a (smart) home: 1) a controlled and secure space, 2) a site of activity and practices, 3) a place of relationships and continuity, and 4) an expression of identity and values. These studies emphasize how existing dynamics in the home, for example, related to aspects of functionality, control, and social relations, are important in the adoption and implementation of SHT. In a similar vein, Madsen et al. (2023) show how three ways of approaching and implementing SHT in everyday practices (i.e., the critical, the compliant, and the committed) represent embodied competencies acquired through previous practice and experience.

Sovacool and Furszyfer Del Rio (2020) point out social differences in the adoption of SHT as an important topic for future studies. However, from previous studies, there is already some knowledge on this. A general finding is that the adoption of SHT tends to reflect economic and technological developments, with a following bias towards urban areas and developed countries (Tetteh & Amponsah, 2020). Focusing on specific contexts, a review of socio-demographic variations in SHT adoption shows some clear differences across different socio-demographic groups. SHT adoption seems to be correlated with a higher income (Sanguinetti et al., 2018a, 2018b; Shin et al., 2018) and not living alone (Arthanat et al., 2019). The age of occupants also seems to be an important parameter, where younger households tend to buy and use SHT more than older households, which is reflected in an observed lack of adoption among older households (Arthanat et al., 2019, 2020; Pal et al., 2018). This is in line with an Israeli study finding that SHT was favored by younger households (Parag & Butbul, 2018). On the contrary, a South Korean study found that younger consumers had lower intentions to buy SHT (Shin et al., 2018). Finally, a Danish study indicated that some occupants were not willing to adopt Smart Energy Technologies even when they perceived such technologies as being useful in lowering energy consumption (Billanes & Enevoldsen, 2022).

Although the previous studies reviewed above provide valuable knowledge on the process of adoption of SHT, then more research is needed, not least given the high political expectations of the role of SHT in managing and reducing energy consumption. This study aims to contribute to the existing literature by providing new knowledge on social differences in the adoption of different types of SHT and investigating the impact of SHT on actual energy use. Thus, this paper contributes with empirical evidence on 1) the diffusion of SHT across types of appliances and services, 2) which types of households tend to have different types of SHT, and 3) how having SHT is associated with levels of energy consumption. While the majority of previous studies are based on intentions to buy, we contribute to the existing literature by analyzing survey data with declared purchases combined with energy consumption and administrative register data.

This section presents the data and variables used in the study and the methods that were used to analyze this data.

The analysis consists of three steps: first, we describe the diffusion of different forms of SHT; second, we identify differences in the adoption of SHT across social groups; third, we model the association between SHT uptake and energy consumption for appliances and heating.

Smart home technology (SHT) is increasingly entering homes to provide services such as energy management, security, comfort, and convenience in everyday life. For example as part of HEMS with purposes of energy saving and demand control (Mahapatra & Nayyar, 2022). However, there is still little evidence of the actual uptake of SHT, its determinants, or its influence on energy consumption.

According to Eurostat data, Denmark is among the frontrunners in adopting SHT for security and household appliances. Using Denmark as a case study and based on a representative survey questionnaire among 1,468 Danish households combined with register data from 2019, this study investigated (i) the diffusion of various types of SHT in Denmark, (ii) the social differences in the uptake of SHT, and (iii) the correlation between SHT and energy consumption.

First, it shows that home security and entertainment were the most common SHT services, followed by robotic help.

Second, the analysis also showed that older age groups were less likely to adopt SHT, especially for entertainment and electricity management. The group under 40 years old was more likely to have and use household robotic help (lawnmowers and vacuum cleaners) than other age groups. However, as an exception, it appears that the group aged 71 or older was more likely to have and use heating management as an SHT service. The overall picture is in line with previous studies that also indicate a stronger interest in SHT among younger age groups (Arthanat et al., 2019, 2020; Pal et al., 2018; Parag & Butbul, 2018).

In terms of income and education, there are only a small number of significant correlations. First, the highest income group was more likely to have smart home security. Second, individuals with a technical education tended to have smart heating management. Third, a longer education seemed to be associated with having robotic help in the household, and fourth, living with a partner was associated with a tendency to have and use SHT for entertainment. There are some examples in previous research that link SHT adoption to a higher income (Sanguinetti et al., 2018a, 2018b; Shin et al., 2018) and not living alone (Arthanat et al., 2019), and this study partly underlines this.

Third, from a policy perspective, smart heating control has been promoted as a way to save energy, but how realistic this has also been questioned, and it could be that the opposite is true (Darby, 2018; Hargreaves et al., 2017b; Peffer et al., 2011; Strengers & Nicholls, 2017). The unique data opportunities in Denmark allow us to combine survey data with energy consumption data, separated for respectively energy for heating (gas or district heating) and electricity for appliance use. To our knowledge, this has been the first opportunity to combine such representative quantitative data. This paper is therefore very important in shedding light on the correlations between SHT and energy consumption. The results indicate that there is no correlation between the actual level of energy consumed for heating and having smart heating control. Policy assumptions about any potential energy reduction following smart heating control should therefore take this into account. From this, it seems that having smart heating control is not associated with neither saving nor consuming more energy for heating. This could be the result of two opposing tendencies. First, it is well-known that different types of efficient heating technologies imply changing norms of comfort and that even higher levels of comfort are prioritized over any possible savings, as with performance gaps (Hansen & Gram-Hanssen, 2023), and smart heating control could likely have similar implications. Future studies that include temperature levels in homes with smart heating could be relevant. Second, is the opposite case where the savings result from making it easier to control heat and only use it when needed (Hansen et al., 2022a, 2022b). This paper did not include electric heating, though in a future with expectations of more electric heating, the time of use of electricity, and reducing peak consumption and shifting towards times with high renewable production will be increasingly relevant from a grid perspective (Smale et al., 2017; Torriti, 2015). In this case, smart control of heating will likely be beneficial.

Concerning electricity for appliance use, we investigated the association between SHT adoption and energy consumption, in the form of heating and electricity. We found a modest association between smart electricity management, such as smart lighting and smart plugs, and higher electricity consumption. However, surprisingly smart electricity management was associated with higher heating consumption, which was not the case for smart heating management, such as smart thermostats. This indicates that SHT services of heating management are not linked to higher or lower heating demand, whereas SHT services of electricity management are linked to higher heating demand, and with less confidence, also to higher electricity consumption. Finally, we found a correlation indicating that having home security systems is associated with higher electricity consumption. The home security systems analysis indicated an 8.8% higher electricity consumption, which is higher than what is reasonable to believe can be ascribed alone to the electricity demanded by the security system itself.

Although this analysis took other household and dwelling characteristics into account, the associations with electricity consumption might reflect other household practices that are indirectly linked to SHT, for example, more energy-intensive lifestyles in general. However, from a policy perspective, it is important to emphasize that this study cannot confirm that smart energy management will contribute to lowering energy consumption, neither related to heating nor to electricity for appliances. Our analysis indicates that SHT for energy management can be associated with higher heating consumption, but this requires further investigation.

Despite the richness of the dataset used in the paper, one limitation of this study is that it does not cover all forms of appliances and services linked to SHT, for example, smart TVs and smart fridges, nor does it capture varying degrees of smartness of the selected smart appliances. Also, ownership of electric vehicles which is not accounted for in the analysis may have influenced the result, although the uptake of EVs at the time of the data collection was low. Furthermore, our data may quickly appear outdated as the market for SHT is expanding rapidly. Future research could address these limitations with updated forms of SHT services. Finally, future studies should include the use of SHT in a future electrified society where flexibility and time of use become increasingly important.