Data-Driven Variable Impedance Control of a Powered Knee-Ankle Prosthesis for Adaptive Speed and Incline Walking.

Most impedance-based walking controllers for powered knee-ankle prostheses use a finite state machine with dozens of user-specific parameters that require manual tuning by technical experts. These parameters are only appropriate near the task (e.g., walking speed and incline) at which they were tuned, necessitating many different parameter sets for variable-task walking. In contrast, this paper presents a data-driven, phase-based controller for variable-task walking that uses continuously-variable impedance control during stance and kinematic control during swing to enable biomimetic locomotion. After generating a data-driven model of variable joint impedance with convex optimization, we implement a novel task-invariant phase variable and real-time estimates of speed and incline to enable autonomous task adaptation. Experiments with above-knee amputee participants (N=2) show that our data-driven controller 1) features highly-linear phase estimates and accurate task estimates, 2) produces biomimetic kinematic and kinetic trends as task varies, leading to low errors relative to able-bodied references, and 3) produces biomimetic joint work and cadence trends as task varies. We show that the presented controller meets and often exceeds the performance of a benchmark finite state machine controller for our two participants, without requiring manual impedance tuning.

Improving Sit/Stand Loading Symmetry and Timing Through Unified Variable Impedance Control of a Powered Knee-Ankle Prosthesis.

Individuals using passive prostheses typically rely heavily on their biological limb to complete sitting and standing tasks, leading to slower completion times and increased rates of osteoarthritis and lower back pain. Powered prostheses can address these challenges, but have control methods that divide sit-stand transitions into discrete phases, limiting user synchronization across the motion and requiring long manual tuning times. This paper extends our preliminary work using a thigh-based phase variable to parameterize optimized data-driven impedance parameter trajectories for sitting, standing, and walking, with only two classification modes. We decouple the stand-to-sit and sit-to-stand equilibrium angles through a knee velocity-dependent scaling term, reducing the model fitting error by approximately half compared to our previous results. We then experimentally validate the controller with three individuals with above-knee amputation performing sitting and standing transitions to/from three different chair heights. We show that our controller implemented on a powered knee-ankle prosthesis produced biomimetic joint mechanics, resulting in significantly reduced sit/stand loading symmetry and time to complete a 5x sit-to-stand task compared to participants' passive prostheses. Integration with a previously developed walking controller also allowed sit/walk transitions between different chair heights. The controller's biomimetic assistance may reduce the overreliance on the biological limb caused by inadequate passive prostheses, helping improve mobility for people with above-knee amputations.

Predicting Individualized Joint Kinematics Over Continuous Variations of Walking, Running, and Stair Climbing.

Goal: Accounting for gait individuality is important to positive outcomes with wearable robots, but manually tuning multi-activity models is time-consuming and not viable in a clinic. Generalizations can possibly be made to predict gait individuality in unobserved conditions. Methods: Kinematic individuality-how one person's joint angles differ from the group-is quantified for every subject, joint, ambulation mode (walking, running, stair ascent, and stair descent), and intramodal task (speed, incline) in an open-access dataset with 10 able-bodied subjects. Four N-way ANOVAs test how prediction methods affect the fit to experimental data between and within ambulation modes. We test whether walking individuality (measured at a single speed on level ground) carries across modes, or whether a mode-specific prediction (based on a single task for each mode) is significantly more effective. Results: Kinematic individualization improves fit across joint and task if we consider each mode separately. Across all modes, tasks, and joints, modal individualization improved the fit in 81% of trials, improving the fit on average by 4.3[Formula: see text] across the gait cycle. This was statistically significant at all joints for walking and running, and half the joints for stair ascent/descent. Conclusions: For walking and running, kinematic individuality can be easily generalized within mode, but the trends are mixed on stairs depending on joint.

Human Perception of Wrist Flexion and Extension Torque During Upper and Lower Extremity Movement.

Real-world application of haptic feedback from kinesthetic devices is implemented while the user is in motion, but human wrist torque magnitude discrimination has previously only been characterized while users are stationary. In this study, we measured wrist torque discrimination in conditions relevant to activities of daily living, using a previously developed backdrivable wrist exoskeleton capable of applying wrist flexion and extension torque. We implemented a torque comparison test using a two-alternative forced-choice paradigm while participants were both seated and walking on a treadmill, with both a stationary and a moving wrist. Like most kinesthetic haptic devices, the wrist exoskeleton output torque is commanded in an open-loop manner. Thus, the study design was informed by Monte Carlo simulations to verify that the errors in the wrist exoskeleton output torque would not significantly affect the results. Results from ten participants show that although both walking and moving wrist conditions result in higher Weber Fractions (worse perception), participants were able to detect relatively small changes in torque of 12-19% on average in all grouped conditions. The results provide insight regarding the torque magnitudes necessary to make wrist-worn kinesthetic haptic devices noticeable and meaningful to the user in various conditions relevant to activities of daily living.

Modeling the Transitional Kinematics Between Variable-Incline Walking and Stair Climbing.

Although emerging powered prostheses can enable people with lower-limb amputation to walk and climb stairs over different task conditions (e.g., speeds and inclines), the control architecture typically uses a finite-state machine to switch between activity-specific controllers. Because these controllers focus on steady-state locomotion, powered prostheses abruptly switch between controllers during gait transitions rather than continuously adjusting leg biomechanics in synchrony with the users. This paper introduces a new framework for powered prosthesis control by modeling the lower-limb joint kinematics over a continuum of variable-incline walking and stair climbing, including steady-state and transitional gaits. Steady-state models for walking and stair climbing represent joint kinematics as continuous functions of gait phase, forward speed, and incline. Transition models interpolate kinematics as convex combinations of the two steady-state models, with an additional term to account for kinematics that fall outside their convex hull. The coefficients of this convex combination denote the similarity of the transitional kinematics to each steady-state mode, providing insight into how able-bodied individuals continuously transition between ambulation modes. Cross-validation demonstrates that the model predictions of untrained kinematics have errors within the range of physiological variability for all joints. Simulation results demonstrate the model's robustness to incline estimation and mode classification errors.

Shortcomings of human-in-the-loop optimization of an ankle-foot prosthesis emulator: a case series.

Human-in-the-loop optimization allows for individualized device control based on measured human performance. This technique has been used to produce large reductions in energy expenditure during walking with exoskeletons but has not yet been applied to prosthetic devices. In this series of case studies, we applied human-in-the-loop optimization to the control of an active ankle-foot prosthesis used by participants with unilateral transtibial amputation. We optimized the parameters of five control architectures that captured aspects of successful exoskeletons and commercial prostheses, but none resulted in significantly lower metabolic rate than generic control. In one control architecture, we increased the exposure time per condition by a factor of five, but the optimized controller still resulted in higher metabolic rate. Finally, we optimized for self-reported comfort instead of metabolic rate, but the resulting controller was not preferred. There are several reasons why human-in-the-loop optimization may have failed for people with amputation. Control architecture is an unlikely cause given the variety of controllers tested. The lack of effect likely relates to changes in motor adaptation, learning, or objectives in people with amputation. Future work should investigate these potential causes to determine whether human-in-the-loop optimization for prostheses could be successful.