AI-driven hip exoskeleton control framework that rapidly generalizes to a broad range of users and real-world locomotor tasks - Abstract In order for wearable robotic exoskeletons to assist the American public throughout daily life, researchers need to develop a control framework that satisfies real-world use. Over 10% of adults have difficulty walking, which hinders their ability to perform daily activities, maintain independence, and have a satisfactory quality of life. To address this issue, wearable exoskeletons have the potential to augment the walking ability of a diverse array of community members throughout their daily lives. That is, if researchers can establish an exoskeleton control framework that is 1) easy to use and 2) adequately assists the walking needs of users throughout daily life. However, current exoskeleton controllers can only assist a few stereotypical movements or require hours of arduous expert tuning using specialized equipment. Thus, there is a critical need for an exoskeleton control framework that rapidly and easily tunes to the diversity of user movement patterns during real-world ambulation. Until such exoskeleton controls exist, many community members, especially those with distinct movement patterns and limited resources, will continue to lack the mobility to achieve independent community ambulation. Our long-term goal is to develop an exoskeleton control framework that is easy to use and can quickly tune to any user and effectively assist their daily ambulation. Here, we will progress towards our goal by developing and evaluating a hip exoskeleton control framework that leverages artificial intelligence to rapidly tune to both young and older adult movement patterns in minutes. We expect such exoskeleton tuning to improve user ability to navigate an outdoor course with hills, stairs, and turns better than current ‘state-of-the-art’ exoskeleton controllers. Further, we will mechanistically explain how each exoskeleton control framework affects user walking performance by pairing the outdoor testing with indoor lab tests that involve detailed physiological measurements. The first aim of this research focuses on developing ‘Trailblazer Exoskeleton Control’ - a versatile hip exoskeleton control framework that leverages artificial intelligence to interpret the movements of new users in minutes via robot integrated wearable sensors, thereby enabling a non-specific and task-agnostic control strategy. The second aim’s objective is to evaluate the ability of young and older adults using Trailblazer Exoskeleton Control and three alternative conditions to navigate an outdoor walking course. Our engineering innovation using artificial intelligence to develop a new and easy to use exoskeleton control framework will set the stage of wearable robotic exoskeletons to assist the movement patterns of community members across the lifespan.
