Socially Assistive Robots (SARs) continue to be a promising, emerging tool as part of therapeutic and clinical health care applications, though widespread adoption of these robots in day-to-day health care is yet to be realised. These social robots may provide a motivational, complementary tool to traditional (non-digitised) interventions (either in home-based or clinical settings) to improve engagement with (and efficacy of) therapies. Additionally, through personalised interactions and therapies (i.e. adapting the level of therapy provided to individual users), SARs can be an excellent approach towards person-centred care, to improve (mental and physical) health-related outcomes for both pre-clinical and clinical populations.
However, working with these target populations - (pre-)clinical persons with existing or potential vulnerabilities and risk factors (such as elderly citizens with Mild Cognitive Impairments or dementia, or children with neurodivergence, e.g. ASD) - introduces a set of niche theoretical and methodological constraints and challenges. Existing HRI research in healthy human populations may not necessarily generalise to these clinical and therapy-based contexts. These contexts are also confounded by an often under-considered set of ethical issues (such as issues of selfhood, agency, and privacy), as well as factors related to cultural norms and expectations, socio-economic status, race, and gender.
Therefore, researchers interested in clinical, therapy-based or health-related applications of SARs should carefully consider the research that their work is grounded in. We should develop (and build on) a relevant body of scientific literature that helps us refine our understanding of these niche applications and target audiences, in a way that allows ongoing research in these healthcare-related domains to become more ecologically valid; thus increasing the viability of SARs in the long-term.