The future of elderly in the hands of Care Robots

Transcription

1 1 pages. In Proceedings of. The future of elderly in the hands of Care Robots José Ignacio Rocca Department of Computer Science University of Birmingham Edgbaston, Birmingham, UK B15 2TT ABSTRACT The increasing elderly populations in most parts of the world and the advances made in robotic technologies have raised the question about how the future of elderly care is going to be. The main motivation of this study is to analyse and envision a future in which older adults can live in their homes independently for longer with the support of assistive care robots. However, we want these robots to provide good quality in care and assistive practices enabling care receivers to live with dignity. Our study focuses in three main research questions, the number of care robots will be necessary in the future, the kind of relationship users will have with these systems and the main ethical issues around care robots for elderly. In order to answer these questions and provide further lines for future work, we conducted a state of art research in the field of care robots. Our main findings show us that probably a mixed approach with several robots will be adopted in the future. Some of these robots will assist in more routine tasks and others will provide companionship. Moreover, there are some important ethical issues, namely safety, privacy, and autonomy that need to be balanced in the development of the care robots. Finally there is a necessity for more user centeredness in the field of care robotics.

2 CCS CONCEPTS Human Robot Interaction KEYWORDS Robots and Elderly, Care robots, HRI. INTRODUCTION Motivation Between 2015 and 2030, the number of people in the world aged 60 years or over is projected to grow 56 per cent according to the UN World population ageing report 2015 [1]. As life expectancy expands most countries start to experience some socio-economic imbalances due to ageing of their population. The younger generations are not able to replace older workers. Although several European countries and the United States are experimenting this situation, Japan is the most affected country with 20% of its population being 65 or older. Moreover, as people get older more caregiver workers are required, but usually the number of formal and informal caregivers stays constant. The whole care system is strained by this situation and technology can provide cost effective solutions for tackling this problem. Older people should live in their homes for longer time with the support of several systems [2]. In order to respond to this trend, several governments and organizations have been investing in the industry of eldercare robotics, known as Care robots, which according to a Merrill Lynch report [3] could reach a global market of $17.4 billion by The latest advances in other technological fields such as, artificial intelligence, internet of things and ubiquitous computing have also boosted the development of several robots for elderly care. Nowadays it is possible to produce cheaper robots such us physical assistants, mobile servants or person carriers that provide more functionality and are being increasingly accepted by the customers. Furthermore, in 2014, the International Standard Organization (ISO) published a new standard to ensure the safe interaction between humans and care robots [4]. The main objective of our study is to analyse the current state of art of care robots as they may be one possible solution for the aging population. However, our focus is to understand what will be the requirements for these robots in order to provide good quality of life to their users in a home based environment. We want older adults to live longer in their own familiar environment and these robots must provide dignified quality care to support them. 2

3 Research questions In this context, we guide our study trying to answer the following research questions: RQ1. Should we have several Care robots for different task or only one multi-purpose robot? RQ2. Should we think these robots like cold artefacts like home appliances or more personal devices? RQ3. What are the more important characteristics and ethical issues for ensuring the patients dignity? In order to answer these questions we are conducting a state of art research in which we will analyse the relevant underlying theories, the empirical literature and the commercial products available. This study organizes as follows. The next section describes the methodology we used in order to gather relevant literature in the field of Care robots. Then, we present the current state of art, presenting our base definitions and theories. After that we discuss the answers to our questions and present further research lines for future work. Finally we present our conclusions. METHODOLOGY In order to answer the research questions stated above and properly construct a state of art research in the field of Care robots we conducted several searches using Google Scholar as our search service. Although some part of the relevant literature in this topic will be indexed by specialized sources like Medline or Cinahl (for example tests in care houses or medical institutions) we decided not to include these search providers as part of our study. First we created a combined search string to find relevant articles. As a result of analysing our research questions we defined the search string: (CARE AND ROBOT) OR (CAREBOT) OR (ROBOTS AND ELDERLY). After executing the search in Google Scholar we filtered relevant papers by analysing their title and abstract. We then proceed and start reading the full text version of these articles and though their citations we snowball other relevant previous articles. In table 1 we present the results of our search methodology by topic. On the other hand, as the topic we are studying has great commercial impact, we searched for popular robots and robot manufacturers in order to analyse the current artefacts and products developed by important companies. The current products are presented as examples of the different types of care robots. Topic Number of articles Elderly HCI - HRI 10 Care Robots 16 Table 1. Relevant literature 3

4 STATE OF ART Elderly users Elderly Although the terms elderly or old people may be problematic, in this paper we will use them to refer people aged 60 or over. In this section we will cover the different degenerative aspects of ageing that may affect the interaction between the users and the Care robots. Furthermore, not all users will have the same level of impairments and would bring different capabilities and abilities when interacting with the robots. Understanding this and providing these Care robots with intelligence to model the different kind of user is an analogy of what a human nurse does every day at work. Nurses and care givers adapt they behaviour to different patients in order to provide them with better treatment and experiences. Perceptual abilities The effects of ageing on the visual abilities have been studied several times. For example, most problems with vision tend to appear as soon as the early forties. With the years, there is a reduction within the width of our visual field and colour distinction, particularly between the blue and green range. Presbyopia, a loss in the near vision, is a common impairment in the elderly. Furthermore, several studies indicate that people aged 60 years or more are less able to detect motion and have less accuracy when estimating depths. A reduction in the speed of processing visual information has also been detected in this group [5]. Other perceptual effects of ageing are related with speaking and hearing abilities. Between years the 20% of people have some form of hearing impairment, which rises to 75% for people between years [6]. Hearing declines with years, and older adults miss attention getting sounds with peaks over 2500 Hz. Furthermore some studies indicate that older user may find female voices more difficult to understand because the higher pitch. These affections related with hearing may also affect speech which becomes less clear and distinct. With ageing speech also become slower with longer pauses and less precise words. Some patients with speech impairments could have suffered strokes. Older patients experiment a deficit in language planning and usually speak with hesitation, pauses and false starts. This was a brief resume of possible perceptual impairments that old people may suffer, in order to design an effective interactive Care robot system we should take into account these aspects. As an example, most voice command interface should be adapted to recognize a speech with longer pauses, hesitation or less precise words. Psychomotor abilities Older people is slower when performing complex motor tasks or when asked to track a target. Furthermore, with ageing people have less ability to control and moderate forces they apply to objects and also handwriting is affected as a consequence. People are less accurate when they have to report their body position in relation to their surroundings. These degenerative motor affections cause balance problems, more doubts when planning and executing movement and difficulty to receive and process new information during the execution of a movement [5]. Most common pathologies associated with ageing that affect motor abilities include, arthritis, 4

5 osteoporosis and Parkison s disease. Some Care robot systems could be focused on helping people with balance or mobility problems. Figure 1. RP 7 Robot Figure 2. MySpoon Robot Figure 3. RIBA Robot Attention and memory Older adults have more trouble in maintaining attention, focusing on the important items while performing a task, for longer periods. Selective attention which is the ability to extract relevant information from other distractions also decline with age. Moreover, trained or automatic responses that are executed with low cognitive load are more difficult to change in older people. Ageing also affects the capacity of processing items from the working memory to the short term memory. There is a decline in the ability to remember specific events and remembering how to perform different tasks [6]. Also pathologies such as Alzheimer s disease affect the normal brain function and cognitive abilities. These implications must be taken into account on the design of Care robots systems that not require the patient to execute complex tasks involving high levels of cognitive load. Alternatively, several Care robots could help patients to retain or remember information about their family, themselves or where some things are located in the house. Care robots Definition It is possible to find different definitions in the literature for the term robot, however there is a sort of consensus in defining a robot as a computer system that is physically embodied and present some level (partial or full) of autonomy. Moreover, a Care robot is a robot designed for use in home, hospital or other setting to assist in, support or provide care for sick, disabled, young, elderly or otherwise vulnerable persons [7]. In this paper we are focusing our research on Care robots for elderly users in a homelike environment of interaction. Types of robots Robots can be classified in different ways, for example according to the domain where they are being used we can classify robots as industrial, military, health systems or search and rescue. Other possible classification can be made in terms of the amount of interactivity expected to have with humans, and so we can classify robots as industrial or work robots or personal robots. Personal robots can interact and cooperate with humans for example in their homes or other unstructured and unpredictable context. However, in this study we are classifying the examples of care robots in terms of how the robot will be used among human actors and what responsibilities will the robot have. This classification is provided in [8]. In this sense, enabling robots are those which enable or enhance the performance of an action by the human. Both the robot and the human, work together towards a common goal. The human is in control of the robot. An example of this type of robot could be the RP7 or the Da Vinci Robot which let the doctor execute surgical or medical procedures remotely. Also the feeding robot MySpoon robot can be considered an enabling robot. Secondly, a replacement robot tries to substitute a human by executing the task by itself. Thus, the robot is completely responsible for executing a task. For example the RIBA Robot is completely responsible for lifting a person. Pearl was developed as a nurse-bot by Carnegie Mellon University and can navigate through nursing facilities providing assistance to the care givers. 5

6 Figure 4. Walk Assist Toyota Robot Figure 5. Mobiserv Robot Finally we have assistive robots which aid the human to execute some task without the direct control or input from the human. Among assistive robots we can find two sub categorizations. On one hand, there are rehabilitation robots that help a patient to overcome some kind of physical impairment, for example intelligent wheelchairs, artificial limbs or exoskeletons robots such as Walk Assist from Toyota. On the other hand, there are social assistive robots which give companion like pets or provide services to cue the patient to take their medicines. Mobiserv is an assistive social service robot which can prompt the patients to take their medicines and was developed in the context of a European research project. Robina from Toyota is also an assistive social service robot which can execute household tasks. On the other hand, Paro, Smiby or Aibo can all be considered as assistive social companion robots. These pet robots provide companion to their users through a set of different capabilities. Human Robot Interaction Human Robot Interaction (HRI) is an extensive and diverse research, engineering and design field. The latest technological advances have boosted this field and since 2006 IEEE host a specialist symposium in HRI. As a discipline that studies the interaction between humans and robots it is divided roughly in four areas of application [9]. Firstly, we have the area of human supervisory control of robots in routine tasks such as, assembly lines in factories or delivering packages. Secondly, the discipline identifies remote control of robots in nonroutine tasks or hazardous contexts such as search and rescue robots. Then another form of interaction is automated robots, for example an autonomous car, in which the human is a passenger. Finally the discipline identifies the human-robot social interaction, with robots that provide entertainment, education or in the case of this study assistance and care for elderly or children. As part of the relevant body of knowledge in the field of care robots, two systematic literature reviews in the area of HRI have been identified and are being analysed below. Possible scenarios In this section we analyse the most common tasks performed by care givers and then we will state what are the values and ethical aspects that are important when executing these tasks. We draw upon these descriptions a possible scenario for future collaboration between a patient and a Care robot. Figure 6. Robina Toyota Robot 6

7 Figure 7. Paro Robot Figure 8. Smiby Robot Care practices The basic care practices can be grouped and categorized following different approaches. For example, the site AGIS which provide several guides for home care, present the next list of common care practices for elderly [10]: Bedroom: assisting getting in and out of bed, making bed and change linen. Bathroom: bathing, grooming and toileting. Personal care: dressing, transferring and walking. Health: manage medications and nursing care. Meals: plan, prepare and serve food, feeding and washing dishes. Household: laundry and house cleaning. Shopping: buy and store supplies. Transportation: take to social activities and medical appointments. Social activities: reading, playing games and companionship. Alternatively, a survey studio [11] conducted on nursing houses classify most of these tasks as daily routine which include bathing, cleaning, dressing, toileting and grooming. Other practices that are reported in the study include preparing meals and feeding, managing and controlling medications, assessing medical needs, transferring and helping people moving around, housekeeping, giving companionship and personal attention. The study also identifies planning care and management activities. It is important to clarify that several of these activities are currently being executed with the assistance of non-robotic technology that helps older adults to live on their own for more time. These technologies are known as Telecare and Telehealth. Telecare is a set of sensors (for example floor sensors), alarms and webcams designed to alert the householder of potential risks and, in some cases, alert outside help. Telecare also allows virtual visiting from health care or social workers. This technology usually monitors unobtrusively the usage of the house and behaviour patterns of the householder. Alternatively, Telehealth purpose is to monitor a medical condition that has required or could require hospitalization or major interventions. This is done through a set of wearable sensors that can monitor vital signs and other relevant variables. Usually, when Telehealth systems detect an anomaly in some relevant variable, it triggers an intervention from people outside the house or the medical services. To summarize, as we can see, there are several activities and tasks that could be executed with the assistance of a Care robot system. Furthermore, currently there are two kinds of technology assisting older adults in living by their own and can be combined with the Care robot systems in order to provide better care and quality of life to the elderly. Care values and principles In this context we can raise the question: how can we determine if one of these practices or tasks is being executed according to the ethical values of proper care? Joan Tronto a specialist in the field of care ethics state 7

8 that good care is a result of both a caring attitude (caring about somebody) in combination with a caring activity o task (caring for something) [8]. There are four main moral principles for ensuring good care. Attentiveness for recognizing needs in the patient, responsibility (both from the care giver and the patient) in order to full-fill these needs, competence and skills to execute the tasks properly, and finally responsiveness from the patient or carereceiver in order to guide the care-giver and provide feedback. These principles go along the process of care giving, in each step of the whole process. All these moral principles and characteristics of good care practices are dependent on the context in which the task is being executed and the task itself. Also the task itself as we mention before, is dependent on the patient, as Tronto state, care is thought of as good care when it is personalized. Scenario Imagine your old aunt Debbie, who is living on her own in the same house where she raised her family. She is in her mid-seventies and has some common vision and hearing impairments. However, she is cognitive healthy and uses a walking stick to move around. A couple of months ago Debbie fell down and injury her knee and hip. Now, she does not feel secure to move around the big two-store house without assistance. Although the stairs are her main problem even on the ground floor there are some steps that are dangerous. Moving her to a care house is not an option. She feels comfortable living in the house where she spent her most joyful times in life. She does not want to be treated as a child, nor being a charge for anybody in the family so she decided to have some special technology help at home. RoboCare is a social assistive humanoid robot specially designed for helping elderly users in their home to move around and fetch things they might require. When Debbie wakes up in the morning RoboCare helps her to get out of her bed and walk down to the bathroom. The robot helps Debbie moving around the house just walking on her side, prompting for steps and being ready to catch her if anything goes wrong, just as her grandchildren would do. The robot is in some sense a stand-in for another human presence as is it in some way empathic and express a small set of emotions to the way in which Debbie behaves. RoboCare not only responds to voice commands such as pass me my glasses but also prompts the user to do things with its aid. The robot knows Debbie s routine and adapts its behaviour in order to fit that routine, for examples fetching the right medicines at the right time of day. Another interesting thing of RoboCare is the ability to distinguish between the user and other visitors responding differentially to Debbie. This scenario was created using some examples provided in [12] and helps us to envision which will be the possible requirements and needs for a care robot which assist patients moving around. The scenario, focus on the expected values of the practice of moving people around and Debbie s needs, not on the RoboCare capabilities. Systematic Literature Reviews As a result of our search methodology we identified several Systematic Literature Reviews in the area that provide a good start point for understanding of the current state of art in the field. 8

9 Firstly [13] reviews the evidence of human responses to healthcare robotics in general. The revision concludes that it may be possible to increase the acceptance of healthcare robots by properly assessing the needs of the humans that are going to interact with them and then matching the robot s role, appearance and behaviour. Secondly, [14], focus its revision in the companion function of social assistive robots. Their results show that there is some level of qualitative evidence on the positive effects of these Care robots with respect to the elderly. However, the lack of randomized control trials and little medical literature in the field make impossible to make a strong claim on benefits. Although we could not access to the full text version of [15] and [16], these articles focus their review on possible benefits that assistive social robots may have on the quality of care for people with dementia and other affections. They find some evidence on positive effects in socio psychological and physiological variables; however the quality of the studies analysed was mostly low. Then, [17] research if and how the evaluations of acceptance of socially assistive robots by older people take into account socio-demographic factors (age, gender, education, etc.). Their results indicate that this is true for only a small number of studies. They conclude that assistive devices should be adaptable to individual needs and consider socio-demographic factors in order to improve their acceptance. Additionally, two Systematic Literature Reviews were identified in the field of HRI [9], [18]. Although this field is wider than the scope of our study, the insights provided by these reviews are interesting. In [9] the authors, claim that in order to provide more complex social interactions between Human and Robots both the human and the robot will have mutual models of each other. Their findings indicate that up to the date, the Human Factors community and other relevant disciplines have very small contribution in HRI past literature so more interaction among the disciplines is needed. As can be clearly seen from the analysis of previous systematic literature reviews in the field of Care and healthcare robots the quality of the primary studies conducted is generally low. Furthermore, these reviews stress out the need for having more user centred focus while designing and building these Care robots systems. On the other hand, it is important to state that the overall quality of these systematic literature reviews is good. The scope of search engines is quite broad and includes also specific search services such as Medline or Cinahl. Most of the authors included also the search strings and filtering strategies used for defining the final set of relevant literature. Ethical aspects As we have been analysing there are some relevant ethical issues that should be discussed in order to define good care practices and ensure the dignity for the elderly. As stated in [19] addressing these ethical issues is important because there is a risk of developing robotic solutions to the problems of aging that result in a reduced rather than an improved quality of life for older adults. In [20] the authors describe six possible issues that could threat the dignity of elderly people with robotic care entities. We list these issues: 1. Potential reduction in the amount of human contact. 2. An increase in the feelings of objectification and loss of control 9

10 3. A loss in privacy 4. A loss in personal liberty 5. Deception and infantilization 6. The circumstances in which the elderly should be able to control the care robot. As they conclude, an effective design should balance the benefits of the robot against the ethical issues. Although the ISO standards are useful in terms of focusing the design on the safety and other issues, no ethical framework or guidelines has been discussed or normalized for Care robots How can these guidelines been formulated? What are the underlying theories about human dignities and care values that can be applied and combined in a set of guidelines? There are several definitions of human dignity in the literature and in fact, human rights and dignity are close concepts. We are not focusing our work in the definition of dignity. Instead we are presenting some studies that using some theories about dignity and care ethics try to establish guidelines or frameworks to design and evaluate care robots. Guidelines for design Some authors [21] have analysed the dignity issue from the recognition of persons approach. In this theory, three main kinds of need for recognition have been distinguished: the need for respect as a person, the need to feel self-esteemed, and the need to be loved. These three aspects should be taken into account by the Care robot (social assistive kind) in order to achieve some level of dignity. On the other hand several robot ethics authors [7], [19] have based their works on the Capability Approach (CA) which offers an alternative and more accountable perspective to human dignity. The focus of CA is social justice on the provision of what is needed for a life worthy of human dignity In fact, this concept has been used to calculate the Human Development Index (replacing GDP), as a way of measuring progress by the UN. A capability should be understood as the answer to the question: what a person is able to do and to be? There are 10 central capabilities in which a person should achieve a threshold level in order to live with dignity[19]. 1. Life 2. Bodily health 3. Bodily integrity 4. Senses, imagination and thought 5. Emotions 6. Practical reason 7. Affiliation 8. Other species 9. Play 10. Control over one s environment In [8] the author present a framework that can be used for designing and evaluating the ethical aspects of Care robots with a value sensitive approach. The framework is constructed over the Value-sensitive design approach tailoring this approach to care contexts and the main values of care ethics. 10

11 The framework is composed by 5 aspects: context, practice, actors involved, robot and manifestation of moral elements. The context represents the place where the robot is supposed to interact for example a hospital ward, a nursing home or the patient s home. The practice refers to the specific task the robot is intended to execute or assist such as, lifting a person, feeding, social interaction, bathing or personal cleaning. Then the actors involved aspect refers to the possible changes in responsibilities that the robot could impose to the actors involved. For example a lifting robot may affect the responsibilities and roles in of a nurse patient relationship. As we state previously, the type of robot refer to a classification in terms of how the robot will be used in the human context. Will it replace, enable or assist other actors. Finally the framework takes into account the manifestation of moral elements that are derived from good care practices and ethics. The author claims that the framework can help not only care ethicist but also engineers and designers. Although the framework is based upon good care values, for each practice a care giver or a Care robot would execute or give assistance, some of these values of good care may go in conflict. For example if the patient does not want to take the prescribed medicines, the overall wellbeing of the patient (which is expected in good care context) goes in conflict with the autonomy (ability to make own decisions) of the patient. What happen in real life is that a nurse or a care giver might try to persuade the patient to take the medicines and intuitively decide which value is more important. The team in charge of designing the robot will have to define which values are more relevant for the care practice and context of execution. Alternatively, in [12] the authors stress out in the importance of autonomy as a key value for Care robot systems and focusing more in a user-centredness approach rather than a carer-centredness approach. This is especially important if the user has risky habits or his/her lifestyle is disapproved by the family. Although some facts of older age are relevant, such as declining mobility or worsening memory, the older adults should be able to make their own decisions. Currently, even with the usage of Telecare or Telehealth solutions, the value of privacy is being affected. As said earlier, the possible benefits and ethical issues should be balanced in every case. The framework and guidelines presented in [12] focus in not restricting the capabilities, freedom and autonomy of older adults. DISCUSSION Research Question 1 One or several robots As it can be clearly seen the current spectrum of Care robots for assisting older adults to live in their homes for longer is wide. On one hand we have single function autonomous machines that enable the user to be more independent providing assistance to eat (like MySpoon robot) or to clean the house (vacuum robots). Then we can find robots that are connected to Telecare and Telehealth services in order to control different aspects of the householder s routine and alert outside help if it is necessary. Furthermore, there are robots that are worn by the user in order to increase strength or add physical capabilities. Finally, at the end of the spectrum we can find the robotic companions which main function is help users with a small set of tasks and give companionship. In this category, there are robots that emulate pets (baby seals, bears, cats, etc.) and humanoid robots. 11

12 Regarding our first research question, what is the best approach for future home Care robots? Should we focus our energies in trying to develop a main humanoid multi-function robot or should we imagine a multi component solution? This topic is discussed in [12] and it is possible to say that both approaches may have advantages and disadvantages. Firstly in terms of cost effectiveness a multi-robot approach will be cheaper than developing a general, multipurpose, social Care robot. This multipurpose general robot will have a lot of different capabilities and thus would be much more expensive for the user. Instead a combination of smaller specialized robots will be probably cheaper for the user. On the contrary, a multi-robot approach will provide less attachment and notion of presence for the householder in comparison to a single robot. This could affect some of the capabilities for achieving dignity, so it is important to balance the benefits and the possible implications for the quality of care we are providing. Then, it also would be interesting for robots to separate concerns. As it may happen with people and care givers in general, older adults do not like a single entity or person to be the centre of their daily life. For example, they do not want to talk about the news with the same person that bath them or groom them but enjoy chatting about the news with their grandchildren. They prefer separate concerns and topics. In this context, multi-robot approaches achieve this separation of concerns between the different care practices. The user can have a robot for bath related tasks and other robot for companionship. But, how these different systems are going to exchange information? Probably they will be connected to a Telecare or Telehealth central solution. Under this scenario, privacy would be an important value to protect and balance. As a consequence of all these aspects, it is not possible to state what will be the right approach for future care robots. However, generally in robotics technology, less is more. So we probably will face an scenario in which we will have a combination of specific purpose autonomous machines and some companion robots, all connected in an environment of a smart home with different sensors and safety devices. Another thing that is important to state is that these robots will probably operate under some care-giver, nurse or medical advice or control. The robots will provide assistance to the existing care and health system. Research Question 2 Cold or personal robots Should we think in robots as home appliances like a washing machine or more personal robots that can empathize and generate a link with the user? Again this aspect is discussed broadly in the literature and it is also related with the answer to RQ1. We can expect that specific robots such as, a robotic vacuum, will generate less attachment and presence to the householder and thus will be easy to replace it. Alternatively, a general purpose robot, which knows the user routine, likes and dislikes, will generate a stronger connection and link. If we think, in the possible scenario described in the previous question, we can suppose that some of the robots will have presence and attachment meanwhile some others will not. Another thing that it is important to stress out is that, in order to generate some level of empathy, attachment and prescience it is not necessary to create a humanoid robot. The Paro robot [22], a baby seal robot covered with fur, has proven some positive therapeutically effects with older adults. The robot is programmed to exhibit 12

13 a set of responses, such as moving its tail, emitting sounds or closing its eyes. In some way the robot learns from the user likes and dislikes. As some of the Systematic Literature Reviews previously analysed claim, the evidence on the benefits of the Paro robot can be questionable in terms of generalization but some qualitative positive effects have been reported. This example let us understand another important aspect of the whole presence and attachment issue. In order to generate empathy and a strong connection with the user, the robot should adapt its behaviour to the user routine by creating an internal model of the user. More complex interactions, ones of social kind, will require more information about the user, less complex interactions and capabilities, will require less information. For example, a vacuum robot will have to know when it is the right moment to vacuum the bedroom floor, or where the user is located in the house in order to not interfere with him or her. On the other hand, a companion social robot would need to know more information about the user, in order to chat about his or her family or friends. Research Question 3 Ethical issues Finally, some of the important ethical issues we have identified in the relevant literature are related with safety, privacy, loneliness and autonomy. In some context, practice and situation, robots can experiment ethical trade-offs between the values of good care and some dignity principles. It is necessary to prepare them in order to execute the tasks and practice required balancing the different values that are in stake. Some articles present frameworks that help designers to take into account the different ethical issues to avoid threats to the dignity of the user. These frameworks are created using different theoretical grounds. In my opinion, in order to take into account the different ethical issues that care robot have to balance, we need to combine several of these frameworks in a set of guidelines or standardized principles that can guide designers, engineers and ethicist in the development of better solutions. As stated by the HRI Systematic Literature Reviews, more user-centred focus is required. Currently the robotic industry is capability focused and this is generating big gaps with the final user needs. Key aspects and future work Finally, we want to stress out some important aspects that need to be taken into account. These aspects are also possible lines for future works in this field. First we need to focus in understanding the main needs of older adults living in their own homes. We need know the users and their social environment. Then we could try to develop Care robots that will provide assistance in everyday tasks. The capabilities of the robot should be developed by a multi-disciplinary team balancing the benefits of robotic technology with the ethical issues and good care practices. This also will increase the acceptance of Care robots in general. We also need to figure out, what will be the role of these robots in the future smart homes. How these robots will exchange information with several Telecare and Telehealth systems? Moreover, what information will be required by all these systems in order to provide the best quality care for the householder? Furthermore, care robots should handle different kinds of user models for adapting their behaviour. The current state of art in this topic could be analysed to gain insights of possible opportunities and gaps in the 13

14 literature. This also is important in order to study how these Care robots will be successfully adopted in the future. Some articles have studied this, but further work is required in this area. Finally, more rigour is needed in the field of Care robotics. As stated by several previous reviews the quality of studies with older adults is generally poor. One approach could be add more experimental settings like randomized trial control tests and replicating experiments in other populations. Other approach could focus in generating more in the wild studies and using more qualitative techniques in order to increase the ecological validity of the studies. CONCLUSIONS In summary, we have conducted a revision of the state of art in the field of Care robots. We tried to analyse the current literature guiding our study using three research questions. The current social and technological context makes us believe that Care robots may be the future solution to help older adults to live independently in their homes for longer time. However, there are several ethical issues that must be taken into account and balanced by the different kinds of robots. Our findings point out that safety, privacy, loneliness and autonomy are the most important ethical issues that need to be considered. Some authors provide good guidelines to design and evaluate care robots according to different theories of human dignity and care ethics. Finally the state of art in the field shows us that more rigour and user centeredness is needed in the development of Care robots. 14

15 REFERENCES [1] ONU, World population, ageing, Suggest. Cit. United Nations, Dep. Econ. Soc. Aff. Popul. Div. (2015). World Popul. Ageing, vol. United Nat, no. (ST/ESA/SER.A/390, p. 164, [2] M. A. Blythe, A. F. Monk, and K. Doughty, Socially dependable design: The challenge of ageing populations for HCI, Interact. Comput., vol. 17, no. 6, pp , [3] M. L. Bank of America, Thematic Investing Robot Revolution Global Robot & AI Primer, [4] ISO, ISO 13482:2014 Robots and robotic devices -- Safety requirements for personal care robots, [5] D. Hawthorn, Possible implications of aging for interface designers, Interact. Comput., vol. 12, no. 5, pp , [6] P. Zaphiris and M. Ghiawadwala, Age-centered Research-Based Web Design Guidelines, CHI 05, pp , [7] S. Vallor, Carebots and caregivers: Sustaining the ethical ideal of care in the twenty-first century, Philos. Technol., vol. 24, no. 3, pp , [8] A. van Wynsberghe, Designing Robots for Care: Care Centered Value-Sensitive Design, Sci. Eng. Ethics, vol. 19, no. 2, pp , [9] T. B. Sheridan, Human-robot interaction: Status and challenges, Hum. Factors, vol. 58, no. 4, pp , [10] Agis, Home Care Tasks Checklist, [Online]. Available: [11] H. Been and M. Z. Eslami, A survey on tasks performed in eldercare, pp , [12] T. Sorell and H. Draper, Robot carers, ethics, and older people, Ethics Inf. Technol., vol. 16, no. 3, pp , [13] E. Broadbent, R. Stafford, and B. MacDonald, Acceptance of healthcare robots for the older population: Review and future directions, Int. J. Soc. Robot., vol. 1, no. 4, pp , [14] J. Broekens, M. Heerink, and H. Rosendal, Assistive social robots in elderly care: a review, Gerontechnology, vol. 8, no. 2, pp , [15] R. Bemelmans, G. J. Gelderblom, P. Jonker, and L. de Witte, The potential of socially assistive robotics in care for elderly, a systematic review, in International Conference on Human-Robot Personal Relationship, 2010, pp [16] R. Bemelmans, G. J. Gelderblom, P. Jonker, and L. De Witte, Socially assistive robots in elderly care: A systematic review into effects and effectiveness, J. Am. Med. Dir. Assoc., vol. 13, no. 2, pp , [17] P. Flandorfer, Wittgenstein, D. O. Johnson, R. H. Cuijpers, J. F. Juola, E. Torta, M. Simonov, A. Frisiello, M. Bazzani, W. Yan, C. Weber, S. Wermter, N. Meins, J. Oberzaucher, P. Panek, G. Edelmayer, P. Mayer, C. Beck, Q. Meng, M. H. Lee, M. Heerink, V. Evers, M. Pfadenhauer, C. Dukat, J. Le, M. Mohamed, R. Bemelmans, G. Jan, P. Jonker, L. De Witte, L. Carlson, M. Skubic, J. Miller, Z. Huo, T. Alexenko, M. Heerink, B. Kröse, V. Evers, B. Wielinga, H. Robinson, B. A. Macdonald, N. Kerse, E. Broadbent, J. Pineau, M. Montemerlo, M. E. Pollack, N. Roy, S. Thrun, E. Mordoch, A. Osterreicher, L. Guse, K. Roger, G. Thompson, M. E. Pollack, A. Sharkey, N. Sharkey, H. Baharin, S. Viller, S. Rintel, B. Graf, M. Hans, and R. D. Schraft, Population Ageing and Socially Assistive Robots for Elderly Persons: The Importance of Sociodemographic Factors for User Acceptance, Int. J. Soc. Robot., vol. 6, no. 2, pp , [18] M. A. Goodrich and A. C. Schultz, Human Robot Interaction : A Survey, vol. 1, no [19] A. Sharkey, Robots and human dignity: A consideration of the effects of robot care on the dignity of older people, Ethics Inf. Technol., vol. 16, no. 1, pp , [20] A. Sharkey and N. Sharkey, Granny and the robots: Ethical issues in robot care for the elderly, Ethics Inf. Technol., vol. 14, no. 1, pp , [21] A. Laitinen, M. Niemela, and J. Pirhonen, Social Robotics, Elderly Care, and Human Dignity: A Recognition-theoretical Approach, no. September, [22] K. Wada and T. Shibata, Living With Seal Robots Its Sociopsychological and Physiological In uences on the Elderly at a Care House, IEEE Trans. Robot., vol. 23, no. 5, pp ,