Control architecture and network communication for a pediatric exoskeleton

Background and Purpose. Mobility limitations are common among adults with neurological disorders and children with cerebral palsy. These limitations affect the physical, psychological, and social aspects of their lives. Though wearable exoskeletons have been commercially available for gait assistance and rehabilitation in adults, similar robotic assistive exoskeletons for children are limited. Project Description. In this project, we aim to develop a custom, light-weight pediatric exoskeleton for children with mobility limitations, allowing personalization of care to promote their functional improvement and improve the quality of lives. The development of the pediatric exoskeleton requires expertise across diverse fields, including but not limited to: clinical rehabilitation robotics, electromechanical design, human-machine interaction, control theory and network communication. This work presents the control architecture and network communication for a pediatric exoskeleton. Inspired by the theory of shared-control paradigm in human locomotor systems, we designed a hierarchical control architecture for the pediatric exoskeleton. The control architecture includes both a high level controller for detecting users' intents and a low level controller that allows the exoskeleton to follow desired trajectories received from the high level controller. The system has been categorized into smaller modules and a Controller Area Network (CAN) has been designed for the network communication. Outcomes. An overview of the control architecture and network communication will be presented. The design, which is modular and scalable, consists of several key modules in the current stage. The hardware selection process for these modules will also be included (e.g., Host PC, PC104/Plus, master controller, user interface module, joint control module and insole sensor module). Discussion. The locomotion control strategy for the pediatric exoskeleton, which provides gait assistance in children with cerebral palsy, will be discussed. The discussion will also outline steps for future work to implement the hierarchical control architecture for further personalization of care, including adjustment of gait assistance in a real-time setting.