Modeling of Sleep-Onset Contact Pressure Interaction for Childcare Support Systems

To support caregivers in promoting infant sleep, this study proposes a novel contact pressure interaction model that quantifies the comfort of infants during holding. Thin-film pressure sensors are used to capture the time-series pressure data exerted when a caregiver holds an infant. The data is analyzed using a Bidirectional LSTM neural network model and visualized optimal pressure patterns associated with sleep onset. By linking holding pressures with infant sleep behavior, the study provides a foundation for understanding how holding pressure influences comfort. In future applications, the proposed model can be integrated into childcare support robots, with the potential to reduce caregiver workload and enhance the quality of life (QOL) for both infants and families.