Preliminary Design of an Environment Recognition System for Controlling Robotic Lower-Limb Prostheses and Exoskeletons.

Drawing inspiration from autonomous vehicles, using future environment information could improve the control of wearable biomechatronic devices for assisting human locomotion. To the authors knowledge, this research represents the first documented investigation using machine vision and deep convolutional neural networks for environment recognition to support the predictive control of robotic lower-limb prostheses and exoskeletons. One participant was instrumented with a battery-powered, chest-mounted RGB camera system. Approximately 10 hours of video footage were experimentally collected while ambulating throughout unknown outdoor and indoor environments. The sampled images were preprocessed and individually labelled. A deep convolutional neural network was developed and trained to automatically recognize three walking environments: level-ground, incline staircases, and decline staircases. The environment recognition system achieved 94.85% overall image classification accuracy. Extending these preliminary findings, future research should incorporate other environment classes (e.g., incline ramps) and integrate the environment recognition system with electromechanical sensors and/or surface electromyography for automated locomotion mode recognition. The challenges associated with implementing deep learning on wearable biomechatronic devices are discussed.

Inverse Dynamics Modeling of Paralympic Wheelchair Curling.

Paralympic wheelchair curling is an adapted version of Olympic curling played by individuals with spinal cord injuries, cerebral palsy, multiple sclerosis, and lower extremity amputations. To the best of the authors' knowledge, there has been no experimental or computational research published regarding the biomechanics of wheelchair curling. Accordingly, the objective of the present research was to quantify the angular joint kinematics and dynamics of a Paralympic wheelchair curler throughout the delivery. The angular joint kinematics of the upper extremity were experimentally measured using an inertial measurement unit system; the translational kinematics of the curling stone were additionally evaluated with optical motion capture. The experimental kinematics were mathematically optimized to satisfy the kinematic constraints of a subject-specific multibody biomechanical model. The optimized kinematics were subsequently used to compute the resultant joint moments via inverse dynamics analysis. The main biomechanical demands throughout the delivery (ie, in terms of both kinematic and dynamic variables) were about the hip and shoulder joints, followed sequentially by the elbow and wrist. The implications of these findings are discussed in relation to wheelchair curling delivery technique, musculoskeletal modeling, and forward dynamic simulations.

Modelling the deflection of rowing oar shafts.

The deflection of rowing oar shafts subjected to a static load was investigated. Two sets of sculling oars of different design stiffness were tested at three different lengths from 2.66 to 2.70 m. Loads up to 201 N were applied to the blade end of the oar shafts, and deflections were measured at six positions along the length of the shafts. The experimental results were compared with theoretical predictions obtained by modelling the oar shafts as homogenous end-loaded cantilever beams. The results show that the oar shafts are not uniform, in contradiction to the assumed model, but rather are most compliant near the sleeves and up to 80% stiffer towards the blades. The effect of oar shaft stiffness and length on the deflection angle at the blade end of the oar shaft was at most 1.18 ± 0.01°. The measured variation of stiffness along the shaft has implications for boat propulsion and rowing performance.