Project Summary Individuals with lower limb amputations are at higher risk of falling compared to able-bodied and other clinical populations and are more likely to sustain life-altering injuries. The higher fall risk is primarily due to the loss of the muscles crossing the ankle, which are critical to maintaining balance control. Prosthetic devices are designed to provide appropriate stiffness for needed stability and support. While research has shown the optimal stiffness to maintain balance varies across ambulatory activities (e.g., straight walking versus turning), most clinically prescribed prosthetic devices are passive and only provide a fixed stiffness level. The one commercially available, powered prosthetic ankle-foot has not been shown to restore balance control. Thus, a prosthetic device that actively adjusts ankle stiffness across different ambulatory activities is critically needed to advance the field and improve balance control for those with lower-limb amputations. The goal of this project is to determine if automatic stiffness modulation can improve the balance control of individuals with lower limb amputation as they perform typical ambulatory activities of daily living. By matching the ankle stiffness to the task requirements, we believe we will significantly improve balance control and decrease fall risk for those with lower-limb amputations. In the proposed work, we will utilize an open source, lightweight, state-of-the-art hardware system (Open-Source Ankle) that includes novel hardware, actuation, sensing, computation, and control software and pursue three specific aims. In Aim 1, we will perform a human subject experiment to determine the influence of prosthetic ankle stiffness on balance control during a wide range of ambulatory activities that will provide the basis for our activity detection and stiffness modulation algorithms. In Aim 2, we will implement our activity detection and phase-varying stiffness modulation algorithms into the Open-Source Ankle. We will use machine learning techniques to predict different ambulatory activities and validate the ability of the Open-Source Ankle, fit with a commonly prescribed low-profile prosthetic foot, to modulate the stiffness profile throughout the stance phase of the different ambulatory activities. The outcome of this aim will be a semi-active prosthetic ankle-foot system with activity-dependent, phase-varying, and user-specific mechanical stiffness profiles. In Aim 3, we will perform a second human subject experiment to determine if automatic stiffness modulation improves balance control in real-world environments. The outcomes of this research will provide insight into the relationships between stiffness and balance and if a semi-active prosthetic system with automatic activity-dependent, phase-varying, and user-specific stiffness modulation improves balance control for those with below-knee amputations. This addresses a critical need in service member, veteran, and civil...