This paper explores the relationship between meal-assistance robots (MARs) and patient autonomy through a focused ethnography of a pilot experiment in a palliative care hospital. Drawing on the theoretical approach of Science and Technology Studies (STS), the study challenges the dominant dyadic model of human-robot interaction by proposing a relational framework that emphasizes the robot’s integration within the patient’s broader care network. While MARs currently have limited functionality for enabling autonomous eating in patients with severe mobility problems, participants reported enhanced well-being and expressed a strengthened sense of autonomy through their interactions with the robot. By integrating the notion of relational autonomy with a sociotechnical lens, the study demonstrates that the robot’s impact on autonomy is not solely a function of its technical capabilities but is co-produced within the care context. Central to these results is the notion that we call the robot’s heteronomy capacity—its ability to interact with and adapt to multiple agents and contexts. These findings contribute to a shift in the ontology of socially assistive robots (SARs), highlighting how robotic agency is dynamically produced through socially embedded interactions. The research underscores the importance of interdisciplinary approaches in MARs design and healthcare implementation, advocating for flexible systems that are open to relational heteronomy as a means of supporting patient autonomy and well-being.
Vallès-Peris et al. (Wed,) studied this question.