Table 1 Status quo and future research... | Emerald Publishing

Table 1

Status quo and future research agenda on RTSR in times of COVID-19 and beyond

Topic	What we know	Future research agenda
Transformative potential of social robots as Entertainer Social enabler Mentor Friend	Robot roles Current state of knowledge indicates that users and caregivers consider that robots should not aim to replace humans but could only perform certain tasks (Lehoux and Grimard, 2018). Social robots can take complementary roles (e.g. motivational coach) and assist human caregivers in improving older adults' physical and psychosocial well-being (Čaić et al., 2019) Children can develop friendly relationships with robots as teachers and progressively treat them as peers or see them as friends (Cañamero and Lewis, 2016) rather than toys (Kanda et al., 2007) Robot skills Evidence suggests that in various child-related contexts, social robots can be as successful as humans in terms of comforting (Meyns et al., 2019; Shahid et al., 2014). One size does not fit; all-specific attention should be paid to the development of the robot's social behavior and skills beyond a mere functional support for the person (Bedaf et al., 2019) Social robots can provide personalized services (Robinson et al., 2014; Sorell and Draper, 2014), emotional comfort (Lehoux and Grimard, 2018) and can help rebuild self-esteem (Leite et al., 2012) Fulfillment of psychosocial needs Social robots can strengthen older adults' sense of autonomy by making them less dependent to staff and formal care, closer to friends and families outside the facilitations and more functionally available in doing tasks (Pirhonen et al., 2020). Social robots have potential to reduce older adults' social vulnerability (Khaksar et al., 2016)	Service context What context-related factors play a role in robotic transformative service? What well-being-relevant service provider roles can be substituted vs. augmented by social robots? Robot design How does the morphology of social robots (i.e. human-like, animal-like and machine-like) impact their transformative potential for diverse groups of vulnerable consumers (e.g. children, older adults, chronically ill and bedridden patients)? Which robot design aspects (e.g. appearance, tone of voice and gaze) lead to optimal well-being outcomes? Robot social capabilities What is the impact of automated social presence on the transformative potential of robots? Can the fulfillment of psychological needs serves as an explanatory mechanism of how social robots unfold their transformative potential? Can social robots be equally successful in building positive social relations for adults as they are for children? Consumer attitudes There is strong empirical evidence in various service contexts that documents negative attitudes of consumers toward automation (e.g. Longoni et al., 2019) and robots in particular (Wang et al., 2015). (2)What role do these attitudes play in transformative robotic service interactions? (3)What is the effect of COVID-19 on consumer attitudes toward social robots and have these attitudes reversed in the new “1.5m-society”? Consumer acceptance What drives the acceptance of different social robot roles among vulnerable consumers (i.e. individual level) and their care networks (i.e. collective level)? Which haptic behaviors (e.g. touching and hugging) are deemed appropriate by vulnerable consumers? How can such behaviors be leveraged for increased consumer acceptance?
Barriers to robotic transformative potential	Dehumanization and loss of privacy Older adults fear social robots would substitute their nursing staff and even further increase their loneliness (Čaić et al., 2018). Ethical considerations include loss of human contact, increased feeling of objectification and loss of control, loss of privacy and personal freedom as well as deception and infantilization (Sharkey and Sharkey, 2012) Contextual factors of privacy, trust and perceived behavioral control have a negative impact on continued use of social robots (de Graaf et al., 2015) Need for personalization Personalization is critical to emotional exchanges with robots (Henkemans et al., 2017; Kim et al., 2013). Sandygulova and O'Hare, 2018) found that gender matching is important for robot preference. Children choose social robots with the same gender Social robots' success in decreasing older adults' social vulnerability depends on social robot enablement and mediation. Social robot enablement includes older consumers' trust in social robots, their perceptions about the costs related to social robots and their concerns about safety. Social robot mediation includes personalized service, delivery, entertainment and connectivity provided by social robots (Khaksar et al., 2016)	Ethical concerns To what extent can and should robots take responsibility over children and/or older adults as care givers? To what extent is deception an alarming ethical concern? Can we identify the robot design and behavior characteristics that lead to consumer deception? Research suggests that sometimes meaningful professional relationships or even (commercial) friendships might emerge from service interactions (e.g. Price and Arnould, 1999; Yim et al., 2008). Can the robot-as-friend type assume a similar role? And if yes, is it desirable and ethically tolerable to deploy social robots as human substitutes for consumers with low agency? Can one robot type (e.g. friend) cater to all consumer needs at once, or is the most effective transformative robotic service context dependent? Mistrust What are the robot design barriers that make robots less trustworthy and humane than real humans in child and older adults care? How do different robot morphologies affect consumer trust? How will robots be equipped with truly empathetic AI being perceived and adopted? And how can we mitigate the effects of the Uncanny Valley (Mori, 1970)? Lack of personalization What are the preferred gender options for different robot types (e.g. mentor vs. friend)? Or should the gender be matched with the gender of a vulnerable consumer? Other What are other unintended consequences of robotic transformative service on different groups of vulnerable consumers (e.g. children, older adults, chronically ill and bedridden patients)?
Eudaimonic well-being	Psychosocial health Social robots increase children's positive mood after stressful tasks (Crossman et al., 2018). Social robots strengthen children's motivations and positive emotions in therapy (Meyns et al., 2019) In a healthcare context, children smile more and cry less in company of a social robot (Beran et al., 2015) Social robots improve children's openness and mood in hospitals, contributing to their self-management (Looije et al., 2016) Personal growth In children-related contexts, several studies show that social robots increase productivity, learning and success of children in doing daily tasks (Baxter et al., 2017; Khaksar et al., 2019; Wu et al., 2015). Playing educational games, social robots support children's social and cognitive development (Tanaka et al., 2007) Social robots instructing older adults in physical tasks and cognitive exercises improve their physical and mental health (Bhuvaneswari et al., 2013; Lopez Recio et al., 2013), prolonging older adults' ability to live independently	Measuring well-being How can well-being be measured for different vulnerability groups? And which, if any, new metrics need to be considered? What are the long-term effects of robotic transformative service on well-being outcomes? Will longitudinal studies support the claims of improved psychological well-being of vulnerable groups when compared to those without social robots during extended periods of social isolation? Can positive robotic well-being implications for older adults and children be generalized to other vulnerable consumers? Well-being trade-offs How can positive outcomes of interactions with social robots (e.g. eudaimonic well-being) outweigh negative consequences (e.g. privacy issues and dehumanization) for vulnerable consumers during COVID-19 and beyond? Under what circumstances is deploying social robots in vulnerable consumer settings detrimental to consumer well-being? For instance, may there be conditions (e.g. absolute isolation) under which robots might reinforce consumer subjective loneliness? Different dimensions of well-being To what extent are the hedonic and eudaimonic perspectives on well-being reconcilable in robotic transformative service? What consumer idiosyncrasies are important to consider for an effective robotic transformation of consumer well-being?

Topic	What we know	Future research agenda
Transformative potential of social robots as Entertainer Social enabler Mentor Friend	Robot roles Current state of knowledge indicates that users and caregivers consider that robots should not aim to replace humans but could only perform certain tasks (Lehoux and Grimard, 2018). Social robots can take complementary roles (e.g. motivational coach) and assist human caregivers in improving older adults' physical and psychosocial well-being (Čaić et al., 2019) Children can develop friendly relationships with robots as teachers and progressively treat them as peers or see them as friends (Cañamero and Lewis, 2016) rather than toys (Kanda et al., 2007) Robot skills Evidence suggests that in various child-related contexts, social robots can be as successful as humans in terms of comforting (Meyns et al., 2019; Shahid et al., 2014). One size does not fit; all-specific attention should be paid to the development of the robot's social behavior and skills beyond a mere functional support for the person (Bedaf et al., 2019) Social robots can provide personalized services (Robinson et al., 2014; Sorell and Draper, 2014), emotional comfort (Lehoux and Grimard, 2018) and can help rebuild self-esteem (Leite et al., 2012) Fulfillment of psychosocial needs Social robots can strengthen older adults' sense of autonomy by making them less dependent to staff and formal care, closer to friends and families outside the facilitations and more functionally available in doing tasks (Pirhonen et al., 2020). Social robots have potential to reduce older adults' social vulnerability (Khaksar et al., 2016)	Service context What context-related factors play a role in robotic transformative service? What well-being-relevant service provider roles can be substituted vs. augmented by social robots? Robot design How does the morphology of social robots (i.e. human-like, animal-like and machine-like) impact their transformative potential for diverse groups of vulnerable consumers (e.g. children, older adults, chronically ill and bedridden patients)? Which robot design aspects (e.g. appearance, tone of voice and gaze) lead to optimal well-being outcomes? Robot social capabilities What is the impact of automated social presence on the transformative potential of robots? Can the fulfillment of psychological needs serves as an explanatory mechanism of how social robots unfold their transformative potential? Can social robots be equally successful in building positive social relations for adults as they are for children? Consumer attitudes There is strong empirical evidence in various service contexts that documents negative attitudes of consumers toward automation (e.g. Longoni et al., 2019) and robots in particular (Wang et al., 2015). (2)What role do these attitudes play in transformative robotic service interactions? (3)What is the effect of COVID-19 on consumer attitudes toward social robots and have these attitudes reversed in the new “1.5m-society”? Consumer acceptance What drives the acceptance of different social robot roles among vulnerable consumers (i.e. individual level) and their care networks (i.e. collective level)? Which haptic behaviors (e.g. touching and hugging) are deemed appropriate by vulnerable consumers? How can such behaviors be leveraged for increased consumer acceptance?
Barriers to robotic transformative potential	Dehumanization and loss of privacy Older adults fear social robots would substitute their nursing staff and even further increase their loneliness (Čaić et al., 2018). Ethical considerations include loss of human contact, increased feeling of objectification and loss of control, loss of privacy and personal freedom as well as deception and infantilization (Sharkey and Sharkey, 2012) Contextual factors of privacy, trust and perceived behavioral control have a negative impact on continued use of social robots (de Graaf et al., 2015) Need for personalization Personalization is critical to emotional exchanges with robots (Henkemans et al., 2017; Kim et al., 2013). Sandygulova and O'Hare, 2018) found that gender matching is important for robot preference. Children choose social robots with the same gender Social robots' success in decreasing older adults' social vulnerability depends on social robot enablement and mediation. Social robot enablement includes older consumers' trust in social robots, their perceptions about the costs related to social robots and their concerns about safety. Social robot mediation includes personalized service, delivery, entertainment and connectivity provided by social robots (Khaksar et al., 2016)	Ethical concerns To what extent can and should robots take responsibility over children and/or older adults as care givers? To what extent is deception an alarming ethical concern? Can we identify the robot design and behavior characteristics that lead to consumer deception? Research suggests that sometimes meaningful professional relationships or even (commercial) friendships might emerge from service interactions (e.g. Price and Arnould, 1999; Yim et al., 2008). Can the robot-as-friend type assume a similar role? And if yes, is it desirable and ethically tolerable to deploy social robots as human substitutes for consumers with low agency? Can one robot type (e.g. friend) cater to all consumer needs at once, or is the most effective transformative robotic service context dependent? Mistrust What are the robot design barriers that make robots less trustworthy and humane than real humans in child and older adults care? How do different robot morphologies affect consumer trust? How will robots be equipped with truly empathetic AI being perceived and adopted? And how can we mitigate the effects of the Uncanny Valley (Mori, 1970)? Lack of personalization What are the preferred gender options for different robot types (e.g. mentor vs. friend)? Or should the gender be matched with the gender of a vulnerable consumer? Other What are other unintended consequences of robotic transformative service on different groups of vulnerable consumers (e.g. children, older adults, chronically ill and bedridden patients)?
Eudaimonic well-being	Psychosocial health Social robots increase children's positive mood after stressful tasks (Crossman et al., 2018). Social robots strengthen children's motivations and positive emotions in therapy (Meyns et al., 2019) In a healthcare context, children smile more and cry less in company of a social robot (Beran et al., 2015) Social robots improve children's openness and mood in hospitals, contributing to their self-management (Looije et al., 2016) Personal growth In children-related contexts, several studies show that social robots increase productivity, learning and success of children in doing daily tasks (Baxter et al., 2017; Khaksar et al., 2019; Wu et al., 2015). Playing educational games, social robots support children's social and cognitive development (Tanaka et al., 2007) Social robots instructing older adults in physical tasks and cognitive exercises improve their physical and mental health (Bhuvaneswari et al., 2013; Lopez Recio et al., 2013), prolonging older adults' ability to live independently	Measuring well-being How can well-being be measured for different vulnerability groups? And which, if any, new metrics need to be considered? What are the long-term effects of robotic transformative service on well-being outcomes? Will longitudinal studies support the claims of improved psychological well-being of vulnerable groups when compared to those without social robots during extended periods of social isolation? Can positive robotic well-being implications for older adults and children be generalized to other vulnerable consumers? Well-being trade-offs How can positive outcomes of interactions with social robots (e.g. eudaimonic well-being) outweigh negative consequences (e.g. privacy issues and dehumanization) for vulnerable consumers during COVID-19 and beyond? Under what circumstances is deploying social robots in vulnerable consumer settings detrimental to consumer well-being? For instance, may there be conditions (e.g. absolute isolation) under which robots might reinforce consumer subjective loneliness? Different dimensions of well-being To what extent are the hedonic and eudaimonic perspectives on well-being reconcilable in robotic transformative service? What consumer idiosyncrasies are important to consider for an effective robotic transformation of consumer well-being?