A System for Simple Robotic Walking Assistance With Linear Impulses at the Center of Mass

Abstract

Walking can be simplified as an inverted pendulum motion where both legs generate linear impulses to redirect the center of mass (COM) into every step. In this work, we describe a system to assist walking in a simpler way than exoskeletons by providing linear impulses directly at the COM instead of providing torques at the joints. We developed a novel waist end-effector and highlevel controller for an existing cable-robot. The controller allows for the application of cyclic horizontal force profiles with desired magnitudes, timings, and durations based on detection of the step timing. By selecting a lightweight rubber series elastic element with optimal stiffness and carefully tuning the gains of the closed-loop proportionalintegral-derivative (PID) controller in a number of singlesubject experiments, we were able to reduce the within-step root mean square error between desired and actual forces up to 1.21% of body weight. This level of error is similar or lower compared to the performance of other robotic tethers designed to provide variable or constant forces at the COM. The system can produce force profiles with peaks of up to 15 ± 2% of body weight within a root mean square error (RMSE) of 2.5% body weight. This system could be used to assist patient populations that require levels of assistance that are greater than current exoskeletons and in a way that does not make the user rely on vertical support. Index Terms-Aiding force, cable robotics, center of mass (COM), rehabilitation, walking. I. INTRODUCTION T HE field of wearable robots for assisting walking has witnessed an evolution from advanced, sophisticated devices that assist multiple joints [1]-[3] towards a greater focus on single-joint exoskeletons [4]-[9] and elegant, simple Manuscript

Metabolically efficient walking assistance using optimized timed forces at the waist

“…A force control algorithm was used to apply 32 conditions with sinusoidal force profiles with desired magnitudes, timings, and durations (Sinusoidal Force conditions; Fig. 1B) ( 17 ) as well as 4 conditions where the force had a constant magnitude throughout the step cycle (Constant Force conditions). Because the tether acts symmetrically on both legs, we could not differentiate between the effects of assisting during phases associated with one specific leg (e.g., first or second double support phase).…”

“…A cable robot (HuMoTech, Pittsburgh, PA, USA) applied force profiles to a waist belt. To avoid the tether going slack during portions of the step cycle, a constant backward force was simulated by inclining the treadmill ( 17 , 62 ) by a 2.8° angle. The reported aiding forces are the forces measured with a load cell minus the parallel component of gravity.…”

“…The force conditions were combinations of three desired durations, four desired onset timings, and different desired peak net aiding force magnitudes. The experiment was conducted using our robotic waist tether system ( 17 ), which was developed based on a commercially available actuation and control platform (HuMoTech, Pittsburgh, PA, USA; Fig. 1A, movie S1, and table S1).…”

Using a simple rope-pulley system that mechanically couples the arms, legs, and treadmill reduces the metabolic cost of walking

“…In general, the assistive force is best understood as a forward force applied at both the wrist-to-rope and foot-to-rope connection and thus, pulled the whole body in the forward direction. This effect is similar to applying an external force that pushes the whole body at the waist [ 22 , 29 ].…”

Differences between joint-space and musculoskeletal estimations of metabolic rate time profiles

“…Although there is no direct measurement of the time profile of metabolic rate during walking, we believe there could be opportunities to obtain additional information by conducting perturbation studies with devices such as exoskeletons [ 18 , 85 , 86 ] or tethers [ 87 , 88 ] that can be programmed to act during specific phases of the gait cycle. Optimizing coefficients in simulation models so that the estimations of the stride average metabolic rate match trends from “rich” datasets with perturbations to different parts of the gait cycle could potentially provide more confidence in estimations of the metabolic rate time profile.…”