Pre-crawling exercises have been demonstrated to improve the development of infants with Cerebral Palsy, allowing the infants to be ready for crawling exercise. However, manpower constraints and lack of therapy skills among parents limit the frequency and effectiveness of such exercises.

To address this problem, a device to facilitate pre-crawling exercises in healthcare and home settings was designed in this project. It enables therapists and parents to conduct safe, task-specific crawling training depending on the infant's condition and motion intention. The gait parameters of the infant are captured and stored in the cloud, while the infant's progress are tracked using machine learning.