Keywords

1 Introduction

Life expectancy has increased from 61.7 years in 1980 to 71.8 years in 2015 [1], and this trend has been observed more in higher income countries [2] and higher income populations [3]. A comprehensive study on cause-specific and all-cause mortality factors predicted an increasing trend of life expectancy by 2040 in 116 of 195 countries [4]. According to a report released by [5], the global population aged 60 years or over numbered 962 million in 2017 which was more than twice as large as in 1980 when there were 382 million older adults worldwide. In this report, it has been estimated that the proportion of older adults aged 60 years and older, compared to the total population, is expected to double between 2007 and 2050 and will reach to 2 billion by 2050 [5].

Aging in place has been recognized as one of preferred models of aging due to the growth of the older adult population [6]. Most of older adults prefer to continue living in their own homes [7]. Feeling attachment to their homes and independence, sustaining self-identity, memories of the past, quality of neighborhood, and sense of freedom are all underlying factors associated with this preference [7, 8]. According to [9], more than 20% of adults aged 65 to 84 and around 40% of adults 85 and older live alone in their households. However, while many older adults can easily continue to live independently, aging in place may be problematic for older adults at the higher end of the age spectrum (i.e., those 80 years of age or older) [7]. Also, people are likely to experience physical and mental changes as they age, which also pose difficulties in doing daily tasks independently [10]. A further complication for aging in place stems from having to manage chronic conditions or disabilities, which is prevalent among older populations. According to a recent report from the United States Census, about 39% of Americans aged 65 or older reported having at least one disability; and among those 85 and older, 73% had at least one disability, and 42% have three or more types of disabilities [11].

While aging in place alone can be difficult since many older adults suffer from chronic diseases, physical, visual, and hearing limitations [12, 13], a wide variety of technological advancements have been made to support the wellness and independence of older adults (e.g. [14,15,16,17,18]). Few studies have investigated the potential values that these technologies may being to the older population. For example, in order to find if there is any correlation between the use of home-based technology (HBT) devices and the quality of life among older people, [19] interviewed 160 older people 75 and older. Using regression analysis, results showed that the use of multiple HBT devices is associated with better quality of life of among the residents. A review study of 48 papers investigated the clinical outcomes of smart homes and home health-monitoring technologies between 2010 and 2014. Results from this review showed that home health-monitoring technologies reduced the symptoms of depression and frequency of visits to the emergency department in older adults with chronic illness and facilitated sharing data with clinicians. However, it did not help with the conditions of disease, disability prediction, health-related quality of life, and fall prevention [14].

Many home technologies are posed to help older adults to live independently and age in place, as well as to make daily tasks easier and more convenient for the general population. Enabled by recent technological advancements, there is now a greater variety of home technologies, varying in their degrees of automation, available to users across generations. However, unlike other domains such as aviation and surface transportation, where the concept of automation has been established with widely accepted taxonomies, there is no clear description of automation and its applications among home technologies. This study aims to address the gap by reviewing existing definitions and models of home automation and smart home technologies. With a survey of previously suggested concepts, this study seeks to identify key directions to work toward with the objective of creating and establishing a universal definition and taxonomy of home automation and smart home technologies.

2 Smart Home Technologies

The term “smart home” was first used in 1984 in a consortium hold by the National Association of Home Builders [20]. The terms, “Domotics” and “home automation” were used to describe in-home technologies of the upper class in the 80s and 90s. [21] defined the term “smart home” as “a residence equipped with technology that facilitates monitoring of residents and/or promotes independence and increases residents’ quality of life” [21]. Smart home technologies passively collect multiple types of data (e.g. physiological and location data) from the resident, and, depending on the goals, share it with the residents, family members, and/or care providers [22].

The purpose of smart homes is to enhance resident security, safety, health and quality of life by non-obtrusive monitoring of activities. Characteristics of case models of smart houses such as Gator Tech Smart House, Matilda Smart House, the Aware Home, Duke University Smart House, Drexel Smart House, and MIT Smart House were reviewed by [23]. They discusses that the interrelations between the residents, spaces, technologies, social behavior, and the communications are important and there is not a harmony between the users’ daily routines and the design of the house. A systematic review on smart home technologies showed that physical and functional health are the top 2 priorities [21] and less attention has been paid to the issues of social interactions.

While users across the age spectrum, researchers, developers and designers, and service providers alike are moving toward making the smart home a reality, there is not yet a comprehensive framework to describe the various forms of in-home technologies. Home technologies vary greatly in the level of automation and the areas of application. In order to facilitate a streamlined discussion among stakeholders, as well as to ensure a successful deployment of potential solutions, it is necessary to better understand the current state of in-home technologies and to work toward developing a framework defining various application types and levels of automation.

3 Technology Adoption

The older population has been traditionally regarded and stereotyped as late adopters or rejects of technological innovations. Recent studies, however, have found evidence to argue that older adults, especially Baby Boomers, are interested in new technologies, willing to learn about them, and in fact using a variety of them [24,25,26]. While older adults have been slower than younger generations to adopt new technologies, they are increasingly embracing new technologies and related applications. For example, Pew Research Center reported that smartphone adoption among older adults 65 and older has almost quadrupled between 2011 and 2016, and that the vast majority of the segment is using the internet [27].

The desire of older adults to age in place leads them accept in-home technologies [28]. However, due to the physical and cognitive constraints, older adults accept new technologies under the influence of specific factors [29]. Older adults’ willingness to adopt smart home technologies within continuing care retirement community living environment was explored through individual interviews and focus groups. Results of the qualitative analysis showed that perceived need was a critical factor in older adults’ willingness to adopt smart home technologies [22]. In 2010, Coleman et al. found the same thing: older adults accept a technology if they find a direct benefit of it. Older adults want technologies that are useful, trustable, help them to age in place, and improve their quality of life [14].

[14] in their review paper concluded that the level of technology readiness for smart homes and home health monitoring technologies is still low. Previous studies showed ten factors as the key determinants for older adults’ technology adoption: value [28], usability [30], affordability, accessibility, technical support, social support, emotion, independence (Hawley-Hague, Boulton, Hall, Pfeiffer, & Todd, 2014; 30], experience, and confidence (Lee and Coughlin, 2015). In addition to these factors, [28] found that older adults express concerns of stigmatization when the technology is too noticeable or obtrusive within their homes. They concluded that stigmatization and privacy are the two factors that do not persist over time. Finally, they concluded that the role of family and friends on older adults’ perception of needing a technology in technology adoption process should not be ignored [28].

When it comes to in-home technologies, privacy is one of the biggest user concerns. Older adults prefer to have a system that tells them when cameras are activated, with a very clear indicator which assures the users that their privacy is not being compromised (Hawley-Hague, Boulton, Hall, Pfeiffer, & Todd, 2014; 28, 30]. Older adults usually refuse video cameras at home but have not fear of being recorded by microphone [30]. However, [22] said that privacy concerns rarely impacted users’ adoption choices and they are willing to trade privacy with the potential benefits of smart home technologies.

Smart homes, also called “automated homes,” aim to automate the tasks for the residents [31]. We call a technology “smart” if that technology possess an awareness of its situation and is capable of reacting to it [32]. If smart home technologies are supposed to accomplish tasks autonomously, then what tasks should be automated and to what extent? A taxonomy of smart houses was developed by [33] (Fig. 1). In this taxonomy, smart houses are divided into three categories: controllable houses, programmable houses, and intelligent houses. Classes in this taxonomy, from the top to the bottom, imply the evolution of houses and the complexity of the systems. Although the taxonomy provides a good source of information and a discussion platform on the complexity of smart houses, it mainly focuses on the input data and lacks the detailed information on the role of users and possible human factors outcomes.

Fig. 1.
figure 1

A taxonomy of smart houses by [33]

Full size image

Although one of the fundamental aims of automation and autonomous systems is to eliminate burden [34], older adults are concerned with being dependent on smart home technologies. They believe that smart homes would decrease their independence rather than improving it [30]. For example, older adults fear that smart home technologies are tools that substitute for personal forms of care and communication [21]. To alleviate these concerns researchers need to know how autonomous smart home technologies should be designed for the target users. A framework that conceptualizes autonomy and identifies human automation interaction variables is necessary to answer these questions. Moreover, any changes in the level of autonomy of a system, influences the role of the users. To understand what roles should be given to users based on the environmental conditions and user characteristics, researchers need a framework to communicate the design decisions and predict the human-automation interaction outcomes.

4 Level of Autonomy

Autonomy has been conceptualized in different domains. [35] defined automation as “full or partial replacement of a function previously carried out by the human operator.” They suggested a continuum of levels for automation that starts from the lowest level of fully manual performance to the highest level of full automation. Similar approaches have been applied to the levels of driving automation for on-road vehicles and medical robots [36] that describe the allocation of responsibility on the user versus the technology.

In the area of human robot interaction (HRI), [37] provided a cohesive framework on robot autonomy. This framework allows designers and researchers to identify a robot’s autonomy level on a 10-point scale. In their definition of autonomy, they distinguished the psychological and artificial aspects of the term and defined robot autonomy as “the extent to which a robot can sense its environment, plan based on that environment, and act upon that environment with the intent of reaching some task-specific goal (either given to or created by the robot) without external control.” Sensing, planning, and acting have been considered as the important aspects of function allocation between a robot and a human in this framework.

To reduce risks for an autonomous flight management system, NASA engineers worked on developing a prototype to prove the utility of autonomy concepts [38]. They suggested instead of looking for complete autonomy from human intervention, the focus should be on determining how autonomous each function within the system should be. They developed a method for determining the appropriate level of autonomy for each function within a system [38]. Using the four-stage model of human information processing theory, they classified functions into: information acquisition, information analysis, decision and action selection, and action implementation. According to this model, a particular system can involve different levels of automations across the four classes of function at the same time.

Most existing frameworks use the stages from information processing theory as a measurable concept to assess functions within human automation interaction [35, 38] which raises a few questions: what if the underlying mechanisms for processing automation related information is different from non-automated information? Is there any other mechanisms above the information processing stages for analyzing automation related information?

Another aspect in the level of autonomy frameworks is a continuum from fully manual to fully automated. For example, there are 10 levels in LORA by [37], 8 levels in the level of autonomy assessment scale by [38], and 4 levels in adjustable autonomy in intelligent environment by [34]. Figure 2 summarizes some of the existing frameworks. Since any changes in the level of autonomy requires the user to adopt new demands, these continuums are efforts to understand user reactions to autonomy and facilitate the interaction between the user and the systems.

Fig. 2.
figure 2

A summary of existing frameworks on the level of autonomy.

Full size image

As previously mentioned, there are a lot of common perspectives in the existing frameworks on the level of automation. However, most of them do not explain causal relationships between variables and outcomes [37]. To understand the ability of the frameworks on predicting the outcomes, experimental evidences as evaluative criteria need to be provided, specifically in the domain of home automation in order to alleviate concerns of safety, security, and user trust.

5 Discussion

Homes play a key role in the health and lifestyle of older adults. It has been estimated that older people spend 80 to 90% of their time in their homes [19], and one out of five adults aged 65 to 74 and two out of five adults aged 85 and older live alone [9]. While living alone and aging in place might be a challenge for older adults, home-based technologies and automation have been the focus of research to identify needs and enhance their safety and security.

Researchers have been using technology adoption models to uncover factors that influence technology acceptance and usage. However, little is known about users’ acceptance and performance in interaction with automated systems in homes. For example, although home-based technologies are being automated to eliminate burden and help older adults accomplish the tasks faster and easier, some of the older adults avoid depending on automated systems for the fear of skill degradation in the long term [35].

Having a framework in the area of automated home technologies and level of autonomy can help researchers to understand users’ roles and predicting their performance. In this paper, we reviewed some of the existing taxonomies and frameworks on the level of autonomy. Most of the frameworks have used the information processing stages as one aspect of to show who, system or human, is responsible at each of the different functional stages. In addition, a continuum of fully manual to fully automate was another aspect in defining the level of autonomy. Although the origin of many of these perspectives are similar, there are differences in their taxonomies and definitions. To build a framework that researchers and stakeholders agree on, further research is required to understand the role of autonomy within smart home technologies and models for predicting human and system performance to support design.