Development of an Assistive Robotic System with Virtual Assistance to Enhance Play for Children with Disabilities: A Preliminary Study

Children with disabilities typically have fewer opportunities for manipulation and play, due to their physical limitations, resulting in delayed cognitive and perceptual development. A switched-controlled device can remotely do tasks for a child or a human helper can mediate the child's interaction with the environment during play. However, these approaches disconnect children from the environment and limit their opportunities for interactive play with objects. This paper presents a novel application of a robotic system with virtual assistance, implemented by virtual fixtures, to enhance interactive object play for children in a set of coloring tasks. The assistance conditions included zero assistance (No-walls), medium level assistance (Soft-walls) and high level assistance (Rigid-walls), which corresponded to the magnitude of the virtual fixture forces. The system was tested with fifteen able-bodied adults and results validated the effectiveness of the system in improving the user's performance. The Soft- and Rigid-walls conditions significantly outperformed the No-walls condition and led to relatively the same performance improvements in terms of: (a) a statistically significant reduction in the ratio of the colored area outside to the colored area inside the region of interest (with large effect sizes, Cohen's d > .8), (b) and a substantial reduction in the travelled distance outside the borders (with large effect sizes). The developed platform will next be tested with typically developing children and then children with disabilities. Future development will include adding artificial intelligence to adaptively tune the level of assistance according to the user's level of performance (i.e. providing more assistance only when the user is committing more errors).