ABSTRACT
Digital technologies, including sleep-tracking devices and smart mattresses, are becoming increasingly prevalent for gaining certainty and insight into unconscious behaviours such as sleepwalking. Sleepwalking, also known as somnambulism, is difficult to diagnose because it is episodic and unpredictable. Drawing on ethnographic research in sleep medical facilities and online communities focused on sleep tracking, this study examines how algorithmic systems shape our understanding of sleepwalking and our approaches to living with it. The results reveal three key dynamics: the negotiation of algorithmic authority, the limitations of measuring and categorising sleepwalking and the importance of integrating algorithmic data with personal experience. Our findings show that algorithmic systems do not simply detect sleepwalking; rather, the relationship between sociotechnical processes, clinical routines and knowledge, self-tracking practices and user interpretations constitutes it. Thus, sleepwalking can only be understood through the negotiation of data, narratives and embodied experiences, which reveals the limitations of standardised algorithmic measurement and the necessity of integrating diverse epistemic resources. These findings make a valuable contribution to sociological theory by shedding light on the relationship between human action and algorithmic influence in healthcare.