Pacapaca: A Laptop-Mounted Companion Robot that Uses Social Mirroring and Affective Bonding to Nudge Upright Posture

초록

We present Pacapaca, a bezel-mounted, expressive micro-robot for laptops that promotes upright upper-body posture through embodied social mirroring. In contrast to screen overlays or wearable sensors, Pacapaca uses only the built-in webcam and a lightweight on-device pose-estimation pipeline to compute a Turtle-Neck Score (TNS). The score integrates (i) the relative size of the face bounding box, as a proxy for face–display proximity, and (ii) the vertical displacement of the mouth keypoint from an upright baseline. The robot bends its neck proportionally to the user’s forward tilt, offering a peripheral, non-disruptive cue of posture drift. When a user-defined threshold is exceeded, a tethered 3D-printed “carrot” drops as a salient yet playful escalation. Upon self-correction, the robot acknowledges recovery with a brief nod and chime. Taken together, the interaction evolves from measurement, to mirrored feedback, to playful escalation when necessary, and finally to confirmation, sustaining engagement while gently reinforcing healthier posture habits. We evaluated Pacapaca with 25 participants on ten social-interaction qualities. The system received high ratings for likability, clarity of emotional expression, feedback timing, and perceived liveliness; seven of ten items averaged above 6 on a 7-point Likert scale. Qualitative comments described the robot as “cute,” “engaging,” and “playful,” and participants expressed interest in continued use. Notably, the companion-like presence and peripheral modality were perceived as minimally intrusive, aligning with the aims of calm, ambient biofeedback. These findings illustrate how expressive micro-robots employing social mirroring can improve adherence to posture correction while remaining approachable and deployable in everyday environments. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.

키워드