This paper outlines the assessment of infants' gross motor performance with a multisensor wearable and its fully automated deep learning-based analysis pipeline. The method quantifies the posture and movement patterns of infants from lying supine until they master walking independently.