Ethics and acceptance of smart homes for older adults

ABSTRACT Societal challenges associated with caring for the physical and mental health of older adults worldwide have grown at an unprecedented pace, increasing demand for health-care services and technologies Despite the development of several assistive systems tailored to older adults, the rate of adoption of health technologies is low. This review discusses the ethical and acceptability challenges resulting in low adoption of health technologies specifically focused on smart homes for older adults. The findings have been structured in two categories: Ethical Considerations (Privacy, Social Support, and Autonomy) and Technology Aspects (User Context, Usability, and Training). The findings conclude that older adults community is more likely to adopt assistive systems when four key criteria are met. The technology should: be personalized toward their needs, protect their dignity and independence, provide user control, and not be isolating. Finally, we recommend researchers and developers working on assistive systems to: (1) provide interfaces via smart devices to control and configure the monitoring system with feedback for the user, (2) include various sensors/devices to architect a smart home solution in a way that is easy to integrate in daily life, and (3) define policies about data ownership.


Introduction
Globally, in 2010, it was estimated that 524 million people were over 65 years old which was approximately 8% of the world population. The World Health Organization (WHO) has estimated this number to grow to 1.5 billion by 2050, making it approximately 16% of the world's population. 4 Further, for some countries, the population growth in this demographic is even more dramatic: in the UK alone, in mid-2017, the age group 65 and above was 18.2% of the total population and this was then estimated to grow to 20.7% by 2027, with projections for a further increase to 26.5% by 2040. 12 This demographic shift toward an aging population increases health challenges, giving rise to a need for health-care technologies targeted to help an aging population. There is an increase in people living with complex chronic illnesses. Globally, around 50 million people have dementia, and there are nearly 10 million new cases every year with significant costs to health services, 5 while, in the UK, dementia costs GBP 26.3 billion per year averaging at GBP 32,250 per patient per year, including health and social care (public and private funded). In addition, between GBP 22.1 and GBP 40.3 million per year is spent on police costs for missing people with dementia. 6 More recently, due to COVID-19, additional health-care challenges have arisen for older adults 7 as they are at much higher risk due to the virus. 8 The number of deaths was highest among this demographic, specifically those who were aged 80 or above. 7 According to ONS, the death rate was higher in males over the majority of age groups. 7 The significant increase in at home deaths triggered by the pandemic highlights the need for better health-care monitoring with remote communication features to be able to connect older adults with family members or emergency services in the time of need.
To reduce this pressure, assistive technologies are being researched and developed for older adults to enable them to lead their daily lives independently without compromising their health and safety. Advanced technologies benefit families and caretakers offering affordable options to monitor, care and provide safety to their loved ones remotely.
In other words, the purpose of these technologies is to assist people of advanced age in their daily lives to achieve a good quality of life (QoL). QoL is characterized by various factors, such as social contacts, activities, health and family relations, 9 and therefore the World Health Organization (WHO) characterizes health-related QoL as physiological, social and mental well-being, 1 as illustrated in Figure 1.
Within this paper, we define a smart home as a home with a system consisting of one or several assistive technologies. We are focusing on the use of these technologies to improve QoL for the aging community, for example, using assistive technologies to ease the activities of daily life, health  1 World Health Organization, WHOQOL: Measuring Quality of Life. https://www.who.int/healthinfo/survey/whoqol-qualityoflife/en/ monitoring, and self-management systems to help with recording physiological details, fitness-related technologies to be physically active and track emergency situations such as falls. Figure 2 illustrates the functionalities of a smart home to provide various benefits to older adults. There is continuous research and development on functionalities of smart homes. [10][11][12][13][14] One of the main concerns is the adoption and acceptance of these technologies. The number of studies conducted for the acceptance and adoption for smart home technologies for health-care targeting older adults is surprisingly low. In this paper, we present a literature review about the acceptability and ethical issues surrounding smart home technologies for the aging community, addressing the possible issues and challenges.

Motivation and goals
Smart home technologies provide several benefits in terms of supporting a good QoL for older adults. Despite the continuous research and development and the availability of several products in the market, questions arise regarding the technology acceptance, adoption, and interaction. 3 This could be due to several reasons, such as age, gender, health status, physiological and cognitive abilities all of which can heavily impact the acceptance and adoption of assistive technologies. 15 Another reason for low adoption could be related to the system design not adapting to ethical concerns, user experience, user interaction, awareness about technology, or catering for individual user requirements and/or personalization.
A recent study in the U.S. found that there is a rise in the adoption rate of certain smart home technologies among older adults, which include water leak detectors, thermostats, carbon monoxide alarms, smartphones, and back-up generators. The older population in the study was reluctant to adopt novel technologies and did not prefer smart home technologies, such as remote home monitoring, voice assistant systems, camera motion activated systems, and smart lights. 16 An earlier study in 2005, Lau studied the adoption rate of personal emergency response services (PERS) in various countries. 17 They stated that despite the availability of these technologies for some time in the market, only a fraction of the population adopted these technologies with less than 5% of aged Americans adopting PERS. In the UK, the adoption rate was only 15% and below 20% in other countries. 17 Another reason for this could be simply due to the lack of smart home assistive technologies readily available. Liu et al. found that readiness for smart home and health monitoring technologies is still low. Around 56% of the studies regarding smart homes and home health monitoring technologies were carried out or tested in lab environments and were proof of concepts. Currently, there is no evidence that these technologies help address cognitive decline, QoL, or heart conditions for older adults with complex needs. 18 Many research studies focus on using models available to investigate the acceptance of assistive technology among older adults but it is still unclear the reasons for low adoption of these technologies. 19 Intelligent assistive technologies for dementia patients were found to be developed in the absence of ethical considerations which results in low prevalence. 20 While a further assistive technology study (BRAVEHEALTH) shows that participants had a positive attitude toward the technology but were still resistant to adopt the system due to concerns over reliability, security, privacy, and trust. For example, some preferred to engage personally with physicians rather than use videoconferencing. 21 Studies show that video conferencing is perceived useful with benefits such as convenience and time and cost efficiency. [22][23][24] It was found that most people had experience of using video conferencing for personal or work reasons but less than half people used it for health and rehabilitation reasons. 23 In another study, doctors and patients were willing to use videoconferencing, although they preferred face-to-face contact, subject to the nature of the complaint meaning VC is not a solution for all illness and clinical needs but more suitable for short visits with nonchronic conditions or in an urgent care setting. 24 In light of the recent COVID-19 pandemic, it is possible that attitudes around videoconferencing have further changed, especially as it has become a more normal activity in everyday life.
Hence, it is important to gain an understanding of the acceptability and adoption rate of the healthcare technologies, especially for smart homes. The main goal of our research is to identify the reasons for this low adoption by conducting this literature review and find the methods by which adoption and acceptance can be increased. In the end, we offer some recommendations that may help shape the future assistive technologies to make them more acceptable.

Methodology
To understand more about the low acceptance and adoption of the smart home technologies among older adults, we conducted a literature review. We searched in MedlinePlus, PubMed, IEEE Xplore, and Google Scholar using the following keywords: ("acceptance," "adoption," "perception," "awareness," "ethics," and "ethical implications") and ("elderly," "elderly population," and "older adults") and ("assistive technology," "intelligent environment," "smart home," "medical technology," and "health technology"). We refined our search through the development of stringent inclusion and exclusion criteria (Table 1).

Acceptance models
Research studies have used models to investigate the adoption and acceptance of a technology, one of which is named "The Technology Adoption Model/Technology Acceptance Model (TAM)." This model is based on perceived usefulness (PU) and perceived ease of use (PEU) where PU means the person perceiving the technology under investigation as useful and PEU refers to a person perceiving the technology is not complex to operate and therefore easy to use without a lot of effort. 25 In 2000, this model was extended as TAM2 which included two more factors which impact the acceptability of a technology. Social influence means a person's perception toward a technology formed by family, friends or social status. While the latter means a person's assessment of a technology based on how relevant it is for the goals, results and its level of quality along with its ease of use. 26 In 2003, the Unified Theory of Acceptance and Use of Technology was presented (UTAUT). This model merged contributions of various adoption theories and models of technology use. 25,[27][28][29][30][31][32][33] This model was further extended to UTAUT2 to further include a user's perspective considering cost, motivation and habits, for example. 34 Other models that talk about factors affecting adoption include Technology to Performance Chain Model (TPC) and Model of Acceptance of Technology in Households (MATH) [35][36][37] and Elderadopt. 38 Many researches have used these models which include. [39][40][41][42][43][44][45][46][47][48][49][50][51] Capability Approach Framework (CA) is designed to describe an individual's use of resources to improve their daily life. 52 However, this model was used by Nikou et al., to investigate the adoption of digital health care technology among older adults. 53 The Expectation Confirmation Model (ECM) was used by Marikyan et al., to study smart home users' technology adoption. 54 Some researchers are using a multimodel perspective which combines Technology Acceptance Model (TAM), Theory of Reasoned Action (TRA), and the Theory of Planned Behavior (TPB) models to understand usage intention of smart homes by the older adults. 55

Commercialized assistive technologies
Assistive technologies can support older adults to lead an independent, safe and secure life within the comforts of their own home. Different types of assistive technologies, such as remote health monitoring via video, sensors and other smart objects, fall detectors, door monitors, bed sensors, and Smart HVAC (Heating, Ventilation, Air conditioning) which can support the aging community, have been previously defined by Miskelly. 56 Table 2 lists various examples of commercially available assistive technologies fitting within these categories and more.

Inclusion Criteria
Exclusion Criteria Focused on older adults where certain studies also included younger age groups.
Studies conducted using robots as assistive technology.
Technologies related to smart homes Smart homes not consisting of participants of advanced age. Assistive technologies for self-management, easily integrable in daily life to promote independence.
Assistive technologies involved in hospitals or rehabilitation centers Older adults living independently.
Research focusing on other aspects such as robots, sound, speech detection and security aspects of smart homes. Studies were published in English Articles that did not fit the inclusion criteria. Studies about perception of smart home technology focusing on acceptance or adoption by older adults.
There are many more early stage research ideas being conceived than there are commercial products available for consumer use within the smart home technology space. Some examples of commercially available products are shown in Table 3. It could be argued that the steps involved in the commercialization of assistive technologies are slow or not able to meet consumer demands. Coughlin et al. discuss translation of invention to innovation stating that although assistive technologies have been available for some time, government and major corporations have only recently given priority to the implementation of technology for older adults community. Therefore, the availability of these assistive technologies may be limited due to the lack of the policies required to successfully convert them to commercial opportunity. 64

Findings
We organize our findings under two categories: (1) Ethical Considerations and (2) Technology Aspects. A summary of the findings from the surveyed studies is shown in Table 4.

Ethical considerations
Ethics play an important role when developing any technology involving humans. This can heavily affect the adoption and acceptance of the technology if not considered. This gives rise to ethical and legal concerns despite the benefits that can be obtained from various assistive technologies. 20,[96][97][98] We have identified five themes related to ethical considerations including privacy, perceived benefits, autonomy, cost, and support of social and natural environments, stigma, social pressure, awareness and other issues. Objects connected via smart phones or laptops such as smart pillbox, door or smart locks. Also, for auto detecting presence such as for turning on lights. Monitoring via Cameras 3 Home security for older adults and activity monitoring. Smart HVAC 4 Smart heating, ventilation, and air conditioning for temperature control. Saving energy as well as providing control via smartphone. Personal Emergency Response Systems 5 Personal health monitoring system which can be used to call help in case of an emergency such as wearable push button necklace, watch, belts. Smart Watch, 6 , 78 Monitoring multiple measurements such as movement, falls, heart rate and SpO2. Smart Phone Applications, 9 , 1011 Reminders, step counts, data sharing with family members.

Sensors 12
Monitoring activities and monitoring environment such as humidity sen sor, smoke alarm etc.
Monitoring activities such as placed on the fridge, shower, pillbox, or any object to monitor that activity.

Privacy
Privacy is one of the major concerns voiced by participants in various studies. 19,21,43,64,66,77,99,100 In a focus group study about commercially available assistive technology products, participants raised a concern about the technologies monitoring them 24/7. 64 They considered it as a loss of dignity in their own home, even though they acknowledged that the purpose of the technology was to ensure their safety and security. 64 In a further study, participants stated that sharing details about them was very interfering and privacy intrusive. 77 They showed negative responses to image capturing technology which made them feel uncomfortable. 101 However, a recent study in the U.S. found that older adults with internet experience and a positive attitude toward emerging technology were willing to use Internet-connected cameras for home monitoring. Approximately 48% of participants were willing to use cameras for home monitoring compared to other technologies. 92 They also mentioned that it was important for them to control information sharing as not all the participants were willing to share their data with their families, staff and health-care providers. 77 Participants in another study by Steele et al. stressed that they wanted systems to be unobtrusive. 102 Camera monitoring was found to be a useful method to provide complete activity monitoring and fall detection surveillance for older adults, but it was felt to be the most intrusive form of surveillance which made users feel more reluctant to adopt the system. 103,104 Interestingly, Birchley et al. mention Table 3. Current (smart home) assistive technology products and services.

Commercial Products
Description Country Cost Model Lively 13 Lively provides activity sensors which can be placed on common objects within a home for example pillbox, doors, fridge, keychain etc. It also provides a watch that includes a push but-ton for calling help in case of fall emergency. The watch can give reminders for medication, counts steps and shares this with family members.

US
Various upfront and subscription packages.
HomeCare 14 Develco provides home monitoring sensors un-der the HomeCare umbrella, e.g., Gateway, Window sensor, Smart Plug, Light bulb, Flood alarm, Smoke alarm, Humidity sensor, Motion sensor for fall detection and preventing fire or any emergency. This is built for developers to design the solution.
Denmark Individually priced sensors.
Geeni 15 Geeni provides home monitoring sensors e.g., smart plug and switches, sensor alarms, smart humidifiers and various sensors.
US Individually priced sensors.
Just checking 16 Just checking is a home monitoring system to monitor movement of older adults by sensors attached to walls and various objects in the house. It is also for door monitoring and activity monitoring.  The study found that privacy was the major concern which caused low adoption. But the need for these technologies could potentially override their privacy concern.

UK
Dhukaram et al. Participants showed a positive attitude toward the benefits of the technology but were not willing to adopt the system due to current privacy, trust, reliability and security issues (Continued) The participants of all ages in this study perceive data protection and health and safety as highly important. Men tend to perceive greater advantages of health control using medical devices more than women. However, this is marginal difference only. Men also pay less attention to anonymous and intimate ways of using the technology, but other aspects of privacy seem to be similar in all gender groups.
The study also shows that healthy people hold more importance for data protection, its storage and transfer for e-health usage than people with poor health. People with poor health tend to be less concerned about the permanent surveillance unlike healthy people.
(Continued)    that users consider unobtrusiveness as an assurance of data privacy, suggesting that unobtrusive technologies could help alleviate some privacy concerns. 105 Nonetheless, Demiris et al. report that older adults who evaluated the smart home "Tiger Place" at the University of Missouri (in Table 4) found it to be an unobtrusive environment but raised concerns about the data shared with third parties. 65 In addition to data privacy concerns, it is also important to consider data protection. Sánchez et al. raise concerns regarding identity theft from a smart home. Since the smart home data contain details about a user's behavior and daily routine, it becomes very crucial for this data to be protected. 106 It was also found that privacy concerns seemed to be more important for younger people than older adults. This may be due to the lack of knowledge among older adults about the data transmission and other technical details involved in a system. 74,97 Studies show that compared to those in poor health, healthy people also hold more value to data protection, its connection with third parties, its storing process, and policies. People with poor health were more focused on how the technology can benefit them. 79,80 Privacy being one of the main concerns among the participants could be overridden by a better understanding of the significant need for technology.
The need for the technology may lead people to adopt smart homes or assistive technologies if they fulfill their intended beneficial purpose. As mentioned above, people in poor health were more focused about the benefits of technology rather than privacy concerns. 80 We can conclude that the need for technology outweighs privacy concerns to some extent provided the expected or intended benefit is met by the assistive technology. 77,99,106 Despite this trade-off, older adults still require technology to be reliable and trustworthy. Guo et al. 74 found that privacy concerns are associated with trust, and they impact the intention of adopting the technology. For example, users not familiar with a given service provider are not willing to disclose their personal data for fear it will be misused, which in turn impacts the acceptance of service or technology. However, personalization and familiarization of the technology toward the user may positively impact users' trust in technology and therefore its acceptance.
A study conducted by Wilkowska et al. 79,80 found a gender difference when looking at various ethical aspects, including privacy. In this study, all age groups irrespective of health conditions deemed privacy to be of high importance. Only by a small marginal difference was noted; men seemed to be less concerned about the anonymous and intimate way of using technology while women were more concerned about the security and required more security features. Women are reported to prefer to use assistive technology in an anonymous way. To summarize, the main factors influencing views around privacy that these papers highlight are: obtrusiveness, camera surveillance, age, gender and personalization to improve trust. All age groups and genders are concerned about continuous surveillance and data privacy issues which creates doubts about adopting technology. Therefore, it is important to develop technology considering the privacy aspect as it may heavily impact the acceptance and adoption of technology.

Perceived benefits
Technology benefits are more evident to the caregivers 107 than to the older adults for whom the technology is designed. They feel they can do well without it, and therefore lack perceived benefit. 107 If the perceived benefit by the users is high, then they are more willing to adopt technology. Accordingly, older adults with high perceived benefit consider technology beneficial in supporting them for aging in a place and believe that technology will not only increase independence but also reduce burden on family and caregivers. This positive attitude toward the perceived benefits of technology helps in acceptability of the technology. 108 Therefore, it is important that the expectation of the users matched the benefits provided by the technology.
Systems should aim to deliver the benefits and functionality that users consider as desirable and avoid those which are not adequate or unnecessary. 56,109 For example, the basic need of the users expected from assistive technologies is to promote their QoL. In other words, users see a benefit in technologies that can support them to achieve daily tasks, maintain their health, provide safety, and enable independence. This can be achieved by designing systems that monitor health in an unobtrusive manner and alert contacts in case of emergency. These systems when connected to other e-health services can prove to be even more beneficial where doctors or GPs can be in touch with the user and monitor their health records. 110 For assistive technologies to play an important role in supporting older adults they must be reliable. 76,111 One of the geriatric participants in a study reported failing to continue to use a blood pressure monitoring device as they felt readings were inaccurate. 76 Other functionality issues, such as batteries running out or malfunctioning of the features, cause older adults to distrust the devices, making technology less appealing. 63 These occurrences take away the trust from the users, making it difficult for the systems to be adaptive.
There is some debate surrounding the factors affecting the adoption of technology. Wilkowska et al. found that men perceived the advantages of health control using medical technologies as more valuable than women did. 79,80 However, the difference is only marginal and perhaps this could be due to some women not perceiving the health technology beneficial for themselves due to no health conditions or perhaps lack of knowledge or simply marginal difference. It was found that males were more aware of IoT and the term "smart home" than females. 73 And, Ziefle et al. found that there was no relationship between diversity (age, gender or health status) and willingness to adopt technology. 15 Also, participants gave very little importance to the esthetic of the technology and were focused on the perceived benefits. 15 Assistive technologies need to be designed with the expectations and requirements of users in mind. The system does not need to be highly complex to provide benefits otherwise it will make it difficult for users to operate and accept assistive technologies. For example, a remote health monitoring system should not require a complex set of commands to start up or interrupts activities of daily life (ADL), which makes it inconvenient for the user.

Autonomy
Older adults may consider themselves as vulnerable due to their age-related health conditions. This does not imply that they do not have pride, self-respect, and dignity. 96 Studies show that some older adults think they will do well without assistive technologies as they feel ashamed and insulted. 63 They were not willing to wear devices to monitor their health as it would make them feel frail or needing special assistance. 63 Older adults also tend to prefer to live independently in the comfort of their own homes as long as possible. 107 Autonomy refers to this independence and control over one's life. 112 Assistive technologies that restrict and restrain older adults may not be easily acceptable. For example, smart homes monitoring activities of individuals need to be adaptive of their habits and behaviors (daily routines). 12 Older adults tend to find it difficult to change their daily routine and learn new patterns and interactions for using the systems. Therefore, products should be designed to accommodate their existing living patterns rather than enforcing new patterns. 107 Technologies that do not offer users control, or limited control, over the influence on their lives are not readily adopted by older adults. For example, participants agreed that the sensors in the smart home (Tiger Place) were not intrusive but there was a lack of control, such as duration of monitoring, which made participants hesitate to adopt such technology. 77 Control over technology provides a sense of independence, making the user feel less conscious of being under surveillance as they can turn it off when they prefer. 71 The perception of control over a system made older adults feel they also had control over their well-being and, therefore, impacted their intention positively for using such a system. 113

Cost
Studies show that cost is a reason why older adults are reluctant to buy a technology 19,64,76 and it can also be difficult in rural areas, developing countries or countries with low economic development and low incomes. 114 Many business facilities have transformed to automated systems but very few households incorporate smart home technologies due to high cost. 115 Smart homes contain a collection of several sensors and technologies, making them costly. As older adults are often on a restricted pension income, cost can be an important concern as shown in a study by Steele et al., 102 which highlighted cost as a frequently discussed subject in focus groups. They also found that older adults were more likely to accept the technology if the implementation and maintenance cost was covered by their children or the government. 102 Since these technologies may indeed be costly, the concern about who will pay for these technologies become crucial. In case of care services, people with health insurance may have these costs reimbursed. But if people have to pay for assistive technologies, then the question of whether or how this will be available to people with low income is something that certainly should be considered.
Health is an important basic need for every individual. Services and technologies could either be made less expensive or at a subsidized cost for a more inclusive society, where not only the rich have access to these technologies. For example, government bodies could initiate methods to provide support to the community of older adults by making assistive technologies more inclusive through the improvement of their affordability and their integration with existing services.

Support of social and natural environments
Socializing is an important part of every person's life and depriving people from social needs can cause mental and physical health problems. Social support is a crucial environmental part which helps in improving health. Loneliness plays a role in increasing functional disability among older adults. 116,117 It is important to include social interaction using smart home technologies to help avoid loneliness. 67,118,119 Yeh et al. found a relationship between dependence on activities of daily life (ADL) and people who experience loneliness. Meaning people who were socially isolated or feeling lonely were more dependent on using instruments or assistance for completing daily activities of life. 120 A study by Chen and Chan found that factors specifically relating to cognitive decline, social isolation, and fear of illness were largely overlooked in studies relating to technology for older people. 121 It has been found that older adults perceive that smart homes restrict them from social interactions, which leads to loneliness. For example, the technology may give them a feeling of being safe only when within their house. 65,122 If technologies do not provide comfort and support maintenance of social interactions, this can lead to difficulties familiarizing with and learning to operate the technologies. This in itself can be a source of anxiety for older adults. 63,122 In light of the COVID pandemic, older adults and vulnerable groups have been enforced into social isolation. This isolation increases the risk of health decline, which creates a necessity for health-care monitoring with features to allow communication with family and friends and call emergency services in case of severe health decline.

Stigma, social pressure, awareness and other issues
There are several other reasons which cause low adoption assistive technologies. 19 Older adults participating in a study were not keen on using smart home sensors as they perceived it to lead to obsessing over their health. However, their family members thought the opposite and expressed interest in monitoring and accessing health information, despite the older adult not wanting to worry about it. 66 Another reason is the pressure from family and friends which can hinder their need to adopt technology by creating negative emotions. 67,89 When this motivation from friends and family is positive, it can lead to adoption of technology by older adults. This is also the case when their own need is higher than the pressure from others. However, if social and learning support is not available from the relatives, it could lead to discouragement, embarrassment, or impatience. 89 Stigma can also lead to hesitation in adopting technology. 97 Another term for this could be Ageism which is a form of discrimination and discourages people from adopting technologies if they are told that this is specifically for frail, vulnerable and old aged people. 123 Awareness, experience or low interaction with technology, ownership of devices (early adoption), and personality can also impact adoption of technology. 73,89 It was found that males were more aware of IoT and the term "smart home" than females. 73 There is some debate around gender and adoption of assistive technology. Some studies show that older women seem to perceive technology as less beneficial than men in the same age bracket. This could be due to several reasons, one of which could be due to fewer female participants in the study or simply due to a marginal difference or a small sample. 79,80 Another reason for the gender gap in technology adoption could be simply due to societal influence. Older women had much lower access to employment and education than their male counterparts. STEM was less accessible to women in the past than it is today. 124 A study was conducted in 2018 on smartphone acceptance and their usefulness among males and females in Jordan and the UAE. Generally, there was no difference in terms of how both genders think about the significance of the ease of using smartphones. It was found that Jordanian females who were less exposed to smartphones perceived them to be less useful. 125 It shows that the gender gap varies due to cultural or social influence where women have less female role models to look up to, face sexism in education and at workplace, or simply have had less opportunities.
Women are more likely to withdraw from internet usage due to health and accidental barriers compared to men in a study on internet adoption among older adults. 126 Smart Home and other technologies do not involve diverse gender needs in research studies and products which can lead to less adoption than if these needs were considered. [127][128][129][130][131] Prior research was gender biased and it historically led researchers to carry out their observations on males in biomedical, social, or behavioral research. This has resulted in the death of information focused on females ranging from organisms, well-being, governments, to polices. 124 This discrimination is no longer acceptable, and researchers recruit both genders to test and validate their developments. However, older generations of women may still be unaware of technology and its benefits that younger generations enjoy today.
This stigma and unawareness can be avoided by approaching older adults through media which is mostly used by them, for example, newspapers. Training or help portals and services can be provided to the aging, this will also give a sense of independence to them. Research shows that smartphones and watches seem to be more acceptable among older adults 41,88,101 while some also accept camera-based systems if privacy is taken into consideration. 92 This could be used to advertise service and technologies to assist them in their daily lives. Workshops, seminars and general awareness can be spread among people with the help of cell phones which will also help remove stigma if advertised appropriately.

Technology aspects
Technology needs to be specially designed for use by older adults as they are more likely to suffer from various health conditions. Older adults can be less patient, have difficulties in learning new tasks, or have physical or cognitive problems. In a study by Steele et al., 102 aged participants expressed their concern about interacting with wireless sensor network technologies due to not being able to use all functionalities of the system. Participants emphasized that systems can be difficult to understand due to their age. Hence, technology needs to be designed in a way which does not induce anxiety and discomfort and that it can be easily adaptable.
We have identified three themes related to technological aspects including user context and requirements, learning and training, and design and usability.

User context and requirements
Despite the availability of various assistive technologies as shown in Table II, the rate of adoption is low. One of the reasons for this could be not understanding or stereotyping the users' needs and expectations 132 or simply not catering for the users' context and requirements. This makes the product less beneficial to the users and results in low acceptability and adoption. For example, some products are designed to be cost efficient for caregivers, instead of promoting the QoL of older adults. 107 It has been found that products currently available are not focused on the user's context and therefore do not consider the behavior, environment, or activities of the user. Systems need to provide information about the user's situation and environment to enable efficient remote monitoring and provide maximal benefit to the user, 107 whilst ensuring ease of operation for older users.
Understanding and catering for user requirements within this demographics are essential. For example, designing smart homes or intelligent environments where users have to alter their activities in accordance with the system, rather their own preferred way, should be avoided. Individual habits vary and forcing everyone to follow the same standards may lead to irritation since every individual is different. Therefore, it is important to design a system which adapts to individual needs. 12 When designing a system for a specific audience, it must involve characteristics specialized to that user group. 133 For example, some older adults might have issues with reading small text on the screen, hence enhancing the font size, and creating interactive visualizations which are easily understandable is important. Overall, it is important to understand the audience at which the product is targeted, so it is more personalized and provides the perceived benefit, making it adoptable and acceptable.

Learning and training
Older adults may have difficulty learning the new skills needed to interact with technologies due to their age and health conditions. This can result in difficulties when trying to accomplish regular activities of daily life. Hence, expecting them to learn and operate complex interfaces or configurations is not appropriate. Studies show that participants did not want to train or learn new technologies. 64,76 This causes fear, discomfort, and anxiety issues when trying to use such assistive technologies. 19,63 The aging community often regards technology as difficult to operate. Generally, they consider that they are not capable enough to learn new products. 107 Mostly, they agree that modern technology provides many benefits but because they consider themselves to be incapable of operating technology, they are hesitant to use and, therefore, benefit from it. 121 Compared to young people, older adults have more difficulty in learning to use new products. However, this could be improved by making products with simple interfaces which are easy to use, and understandable to older adults. 97,121 Design and usability Usability and design contribute to user experience, which indicates the level of ease of use, simplicity, and joy that a user can experience from interacting with the technology or product. It has been found that the ease of use, interaction with the device and controllability were the most important characteristics of a system among the middle-aged and older participants. 15,71,134 It is often assumed that a completely automated system might be easy to operate and interact with. Studies show that aged participants when given a choice between an automated and manual system, chose manual as it gives the feeling of having control over the system. 42 Systems may also create fear and irritation when a user is unable to control the system or interact with it. 71 Birchley et al. suggest that providing improper focus on choices to end users puts burden on individuals. 105 It was found that mobile health applications designed for people with dementia lacked some features, which impacted the user experience and its usage. 135 There is some debate around the level of importance of design. Gamberini et al. found that 50% of the problems reported in technology by users were due to usability. This could be resolved by adjusting the design or providing training. 136 In contrast, Ziefle et al. found that participants were less concerned about the design of the technology. They were more focused on the benefits that the technology could provide. 15 The esthetic design of the technology is another perspective for designers and researchers to keep in mind when designing and developing assistive technology. 88,137 For example, older women in a study considered device esthetics as one of the barriers to wearable sensors and it was also found that the wearable sensors that were prominent were not acceptable. 88 Technology can be designed in a way which is familiar to the older generation, this enhances its sense of identity and belonging to the home environment of the person. 137 Multimodal Interaction has also been found to be a positively evaluated technique for interacting with technology. This can include various interaction approaches, such as voice, keyboard, touch screens, 138 gestures, 139 or facial expression, 140 recognized by a system. 138 Designing systems in a user-friendly and interactive manner is important. Older adults need to interact with systems that are easier to operate and provide appropriate feedback, 66 making them easier for users to comprehend. 141 Careful design is crucial as bombarding users with too many unnecessary options or heavy designs can complicate the system and negatively impact user perception of the technology.

Design and data bias
Smart assistive technologies are being developed with the latest artificial intelligence (AI) to help their users achieve the full benefits of the technology. AI or machine learning uses training and test data to train and evaluate the systems. However, the data used to train the systems can be based on data sets which are nonrepresentative of the general case or the targeted audience. This creates bias which means the result is unjust for the audience which it is targeted toward. [142][143][144] A study was conducted for face recognition algorithms with influence of factors, such as race, age, and gender. It was found that the accuracy of the system was lower for females people with darker skin color and varied among age groups as well. 145 Many studies mentioned in Table 4 include various age and genders but there is a difference in the number of recruited participants for age and gender. This variation can also lead to biased results, for example, a technology just reviewed and tested by males may not necessarily work for females or technology designed for older adults but tested by younger audience may also impact acceptance of technologies as they will not provide benefits as expected by its intended audience. A smart home study was conducted in a living environment where the residents were students and living at the Missouri University of Science and Technology Solar Village. The purpose of the study was to study about the use and interest in adopting smart home technology. 146 The research is a great way to understand the perspective of residents regarding smart home technology. However, the same study needs to be conducted with residents who are aged 65+ to understand their perspective for research aimed specifically for older adults. Another example is results achieved on gender difference. For example, there is some debate around difference in perception about smart home technology between genders. However, this could be either statistically insignificant (only marginal difference) or perhaps differences in sample sizes. There may be a difference in need between the genders or differential financial constraints or awareness.
Data sets are a crucial part of AI applications. They need be large and aligned to the population they are designed for. 123 Data set accountability is also essential when determining what is going to be used in an application. 72 Another example is of patients with dementia where providing care varies on an individual basis. This variation requires to be aware of the person's condition to be able to provide a comprehensive care. A behavior pattern based on an incomplete data set or on average values may not produce expected results and undermine the care. 123 Data can be collected from real environments and shared with hospitals and researchers over years to collect real data and experiment research on that. 123

Recommendations
Having surveyed the state of the art in assistive technology for smart homes, we are able to recommend that researchers and developers keep both the user needs and user experience in mind when designing an assistive system through: (1) Providing interfaces to control and configure the monitoring system with feedback for the user, (2) Designing a system which can easily integrate in daily life without creating any constraint or difficulties for the user, and (3) Defining clear and concise policies about data ownership enabling transparency and increasing trust.
Privacy seems to be one of the major concerns which initiates fear of technology leading to lack of trust. Older adults appear to prefer unobtrusive assistive technologies, which are designed to protect their dignity and independence. Privacy concerns can lead users to opt-out of using technologies, despite the benefits they offer to them. Hence, assistive technologies must ensure independence, dignity, choice of control over technology, and information sharing. Data ownership policies need to be introduced, this will include in Europe the General Data Protection Regulation (GDPR) or equivalent policies that may protect and help people understand about the data transmission, storage, and sharing with third parties. Data protection is also crucial for data obtained from smart homes as the risk of identity theft is high, 106 and thus poorly secured smart home systems may be vulnerable to such attacks. Therefore, excellent security protocol measures must be followed.
Personalization is an important factor which helps in earning users trust. Deriving the system around user's behavior and activities may enable the technology to be more adaptive to the needs of the older adult. 74 It is crucial to design technology in a way that provides informative visualization of data and with simple interfaces which boosts ease of use and ease of comprehension. It is assumed that intelligent environments are easier to use and control. However, a fully automated system may take away the control from the user, making them feel vulnerable to the system. 71 Hence, providing users with a configuration panel to control the technology with appropriate feedback is necessary. 113 It can help improve the system if customer satisfaction feedback is taken continuously to improve the needs of the aged demographic. Study shows that older adults can be reluctant to adopting technology, which associates themselves as frail and vulnerable or has negative image of being too old attached to it. Therefore, self image should be taken into consideration when designing or marketing an assistive technology.
Products available are often standalone and are made independently of one another. Therefore, the integration or connectivity for exchanging information is limited. 107 For example, products defined in Table II and Table III cannot be connected to exchange information. Hence, the benefit is limited to each technology. Smart homes need to have devices that can communicate and exchange information, to provide meaningful data about the user and their environment. For this purpose, hybrid solutions need to be designed that contain a combination of various sensors and devices that communicate with each other. Systems need to be designed so that they comprise multiple functionalities to provide a solution under one platform. For example, giving users reminders in a visually friendly calendar, summarized data, trends, alerts, and information about the environment, 12 making technologies integrable, so they become part of the user's daily life rather than forcing people to change their routine.

Conclusion and future work
Smart homes consisting of several assistive technologies provide a variety of benefits for older adults who want to live independently in the comfort of their own home and improve their QoL.
Older adults are often portrayed as dependent and resistant to change. However, they are demanding users who seek an independent and socially connected life. Studies show that though people are willing to use these technologies, factors such as privacy, perceived benefit, autonomy, cost, and support for social environment and technology aspects (user context, user requirements, design, usability, learning, and training) as highlighted in the findings of the extensively surveyed literature listed in Table 4 must be catered for when designing such systems. Customer satisfaction feedback can be utilized to help minimize these problems. This would help in improving the system design in accordance with user-specific needs, hence making it more adoptable. Trust needs to be gained by personalization and focusing on the needs of the user while incorporating ethical and technical aspects. Finally, the key themes we have identified that smart home technologies need to consider are: (1) provide intended or expected benefit, (2) data ownership policies and data security, (3) Personalized systems to gain trust, (3) high reliability, (4) cost effectiveness, (5) promotion of autonomy (independence and control over technology), and (6) caters for user requirements and enhances user experience.
Most importantly, older adults should be involved in the evaluation of these products as end users. If older adults are not aware of, or satisfied with the products, then it is highly unlikely that they will be willing to accept and adopt these technologies. We would, therefore, recommend that future research includes investigating the perception of older adults on currently available commercial products and getting early feedback on state-of-the-art research.
There is a need to study the current rate of adoption and acceptability of assistive technology among older adult people. Understanding the perception of available assistive technologies among older adults as illustrated in Tables 2 and 3, is essential. It is crucial to evaluate what older adults are expecting from smart homes and assistive technologies. This data could help design future technologies to maximize the perceived benefits, acceptability, and adoption among the aging community.

Disclosure of conflicts of interest Statement
The author(s) report no conflict of interest.