Development of VariLeg, an exoskeleton with variable stiffness actuation: first results and user evaluation from the CYBATHLON 2016

Abstract

Background: Powered exoskeletons are a promising approach to restore the ability to walk after spinal cord injury (SCI). However, current exoskeletons remain limited in their walking speed and ability to support tasks of daily living, such as stair climbing or overcoming ramps. Moreover, training progress for such advanced mobility tasks is rarely reported in literature. The work presented here aims to demonstrate the basic functionality of the VariLeg exoskeleton and its ability to enable people with motor complete SCI to perform mobility tasks of daily life.

Methods: VariLeg is a novel powered lower limb exoskeleton that enables adjustments to the compliance in the leg, with the objective of improving the robustness of walking on uneven terrain. This is achieved by an actuation system with variable mechanical stiffness in the knee joint, which was validated through test bench experiments. The feasibility and usability of the exoskeleton was tested with two paraplegic users with motor complete thoracic lesions at Th4 and Th12. The users trained three times a week, in 60 min sessions over four months with the aim of participating in the CYBATHLON 2016 competition, which served as a field test for the usability of the exoskeleton. The progress on basic walking skills and on advanced mobility tasks such as incline walking and stair climbing is reported. Within this first study, the exoskeleton was used with a constant knee stiffness.

Results: Test bench evaluation of the variable stiffness actuation system demonstrate that the stiffness could be rendered with an error lower than 30 Nm/rad. During training with the exoskeleton, both users acquired proficient skills in basic balancing, walking and slalom walking. In advanced mobility tasks, such as climbing ramps and stairs, only basic (needing support) to intermediate (able to perform task independently in 25% of the attempts) skill levels were achieved. After 4 months of training, one user competed at the CYBATHLON 2016 and was able to perform 3 (stand-sit-stand, slalom and tilted path) out of 6 obstacles of the track. No adverse events occurred during the training or the competition.

Conclusion: Demonstration of the applicability to restore ambulation for people with motor complete SCI was achieved. The CYBATHLON highlighted the importance of training and gaining experience in piloting an exoskeleton, which were just as important as the technical realization of the robot.

Keywords: Exoskeleton training; Overground walking; Powered exoskeleton; Powered gait orthosis; Spinal cord injury; Variable impedance actuation; Variable stiffness actuation; Wearable robotics.

Fig. 1

VariLeg exoskeleton with user (motor complete thoracic SCI). A variable stiffness actuator (VSA) in the knee joint can mimic the stiffness modulation observed in individuals with unimpaired gait (M2/M1). The hip joint is actuated conventionally with an electric motor and a reduction gear box (M3). Cuffs on the leg and a torso orthosis fix the exoskeleton to the user. The user balances using crutches that also serve to pilot the device through push buttons (e.g. triggering steps). Left inset: Details of the foot interface including a spring-loaded passive ankle and ground contact sensing

Fig. 2

Stiffness modulation in the knee joint during gait. The expected human knee joint stiffness modulation during gait was estimated through an EMG-based model, which was verified in static (isometric) condition (adapted from [28]). A possible implementation of stiffness modulation could be to simplify this behavior into several regions with constant stiffness. The controller switches through these levels according to the gait phase. At the CYBATHLON 2016, we used a simpler strategy commanding a fixed stiffness setpoint. Nevertheless, the illustrated stiffness levels could be achieved in test bench experiments. Note that the gait cycle starts and ends with a heel strike of the same leg in this representation

Fig. 3

Schematic of the Variable Stiffness Actuation (VSA) unit and its expected stiffness range. The VSA (inspired from the MACCEPA and MARIONET systems) is illustrated on the left. The lever motor (M_lever) situated in the lever unit controls the lever position relative to the shank. The lever unit is connected to the thigh through the spring k, which can be pretensioned (by the pretension motor M_pretension). Varying pretension, which changes spring length x, results in a change of the stiffness. The stiffness also varies with the deflection α, describing the deflection of the lever unit from its equilibrium position. Stiffness in function of x and α is shown on the right. The mechanically available stiffness modulation range is indicated as a grey area. Holding a pretension continuously is limited by the motor’s continuous current limit indicated with the 100% line (yellow). The relative angle between thigh and shank (knee angle) β therefore depends on the lever’s equilibrium position, the load applied to the joint and its stiffness

Fig. 4

Overview of the control structure of the exoskeleton. The control architecture is divided into three parts: high-level control, low-level control and safety functions. The high-level control is replaying trajectories for the exoskeleton joint positions and the stiffness setpoint. The individual tasks have differring trajectories grouped in modes. The modes can be selected by the user pressing buttons on the crutches or by an operator with an external computer. The trajectories are executed by a low-level position control loop for each joint. The exoskeleton state is supervised by safety functions that stop the exoskeleton if, e.g., the redundant sensing disagrees or the motors receive a position request that is outside of the allowed range of motion. φ_rl, φ_rp, φ_rh designate the reference joint angles, defined by the trajectories (stiffness for φ_rp and walking, inclines or stairs respectively for φ_rl and φ_rh). φ_l, φ_p and φ_h are the angles measured with the position sensors that are fed back to the low-level controller and evaluated in the safety functions of the exoskeleton. I_l, I_p, I_h designate the current sent to the motor. l refers to the lever, h to the hip and p to the pretension motors

Fig. 5

Walking trajectory of the exoskeleton compared to unimpaired gait. The nominal exoskeleton walking trajectory commands the equilibrium position of the knee more towards extension in early stance compared to unimpaired gait. This ensures buckling occurs due to the compliance of the VSA when loaded and is not pre-programmed into the trajectory. Ground clearance of the swing leg was increased to prevent collisions of the foot with the ground

Fig. 6

Walking scaling, incline and stair climbing trajectories of the VariLeg exoskeleton. The walking trajectories can be scaled in length (shown in a) and height to adjust the step. Different trajectories for walking, inclines or stairs were implemented and can be selected via the crutch or a computer wirelessly connected to the exoskeleton. The incline trajectory (b) was created by rotating the walking trajectory and prolonging the knee extension during late stance. The stairs mode (c) climbs steps one foot at a time and measures the height of the first executed step, which is performed at maximal step height

Fig. 7

Results from MACCEPA characterization. Experimental results were compared to theoretical values. Stiffness is higher for higher deflections at high pretensions. Experimental torque fits match theoretical data within 2 to 3 Nm RMSE, whereas stiffness curves display larger errors of up to 30 Nm/rad deviation at the highest pretension

Fig. 8

Amount of training necessary to achieve skill levels for different tasks. Both users required a considerable number of training sessions to gain proficient walking skills. The sit-to-stand motion was mastered after more than 20 sessions. Only basic skills were acquired on stairs and ramps

Fig. 9

Performance of the VariLeg exoskeleton at the CYBATHLON 2016. The CYBATHLON 2016 obstacles presented in order of appearance during the championship (from left to right, top to bottom). Official time for clearance are indicated for the first and second run, if available. The sofa and the slalom obstacles could be cleared during the competition. The tilted path was only cleared during the safety check (i.e., the official test run prior to the competition)