The Myosuit: Bi-articular Anti-gravity Exosuit That Reduces Hip Extensor Activity in Sitting Transfers

Abstract

Muscle weakness-which can result from neurological injuries, genetic disorders, or typical aging-can affect a person's mobility and quality of life. For many people with muscle weakness, assistive devices provide the means to regain mobility and independence. These devices range from well-established technology, such as wheelchairs, to newer technologies, such as exoskeletons and exosuits. For assistive devices to be used in everyday life, they must provide assistance across activities of daily living (ADLs) in an unobtrusive manner. This article introduces the Myosuit, a soft, wearable device designed to provide continuous assistance at the hip and knee joint when working with and against gravity in ADLs. This robotic device combines active and passive elements with a closed-loop force controller designed to behave like an external muscle (exomuscle) and deliver gravity compensation to the user. At 4.1 kg (4.6 kg with batteries), the Myosuit is one of the lightest untethered devices capable of delivering gravity support to the user's knee and hip joints. This article presents the design and control principles of the Myosuit. It describes the textile interface, tendon actuators, and a bi-articular, synergy-based approach for continuous assistance. The assistive controller, based on bi-articular force assistance, was tested with a single subject who performed sitting transfers, one of the most gravity-intensive ADLs. The results show that the control concept can successfully identify changes in the posture and assist hip and knee extension with up to 26% of the natural knee moment and up to 35% of the knee power. We conclude that the Myosuit's novel approach to assistance using a bi-articular architecture, in combination with the posture-based force controller, can effectively assist its users in gravity-intensive ADLs, such as sitting transfers.

Keywords: Myosuit; anti-gravity; assistance; exomuscle; exosuit; muscle-activity; textile; wearable.

Figures

Figure 1

Three-layer architecture of the Myosuit. The architecture is inspired by the bones (structural support), ligaments (passive support), and muscles (active support) of the human's musculoskeletal system. The garment layer is the interface between the Myosuit and the user and provides the overall structure of the suit. The ligament layer incorporates passive elements—rubber bands—to store energy and passively assist the joint's movement. The power layer uses an actuator, routed along the limb, to actively assist the movement of the joint.

Figure 2

Active and ligament layers of the Myosuit. The Myosuit's tendon actuators are attached at the shank and anchored to a waist belt at the hip and by a wrap at each foot. The tendons are routed along the leg using textile cable channels sewn to the garment layer. When the support forces are active, the tensile forces translate into extension torques at the hip and knee joint. The ligament layer provides an antagonistic structure to assist with hip and knee flexion. In this study, which focused on sitting transfers, the ligament was attached to the Myosuit, but not tensioned.

Figure 3

The Myosuit resembles a pair of pants that include a waist belt and thigh cuffs sewn onto a stretchable base material. One set of rubber bands, designed to aid with hip flexion, attach at the front of the waist belt and the front of the thigh cuffs. A second set of rubber bands, designed to aid with knee flexion, attach at the back of thigh cuffs and the base of the actuation unit. The waist belt and thigh cuffs can be adjusted to the user's body to ensure proper fit of the device and efficient transmission of forces. The waist belt houses the control unit and a set of batteries used to power the actuator units. Power cords connect the control unit to the actuators and provide the required power for the system. Dyneema cables, routed along the garment layer, define the actuation path of the Myosuit.

Figure 4

Tendon actuator. Two actuator units, each weighing 1,070 g, provide the assistive forces of the Myosuit. The actuator is placed on top of a carbon fiber shin-plate. Each unit incorporates a 70 W brushless DC motor coupled to an encoder. An inertial measurement unit, motor controller, and cooling fan are placed next to the motor with power and communication leads extending from the actuator to the central control unit. A Dyneema (0.6 mm) cable is fixed at the motor shaft (6 mm) and connects via a pulley system to the upper end of the actuator unit. At the upper end, the actuator cable is connected to the multiarticular cable connecting to the waist belt. The pulley system transmission ratio is 1:4. The unit allows for a maximum cable travel of 0.24 m.

Figure 5

Control chart for the Myosuit. The goal of the Myosuit is to provide its user with anti-gravity support. This approach aims to support the extensor torques at the knee and the hip joint. The controller is based on a virtual leg that connects the hip and ankle joints. The length of the virtual leg scales with the knee angle βknee. The force delivered to the user is defined to provide the highest moments at a knee flexion angle of 80° and decreases as the angle increases. Controller inputs are the shank angle, αshank, trunk angle, αtrunk, and the length of the tendon. Using these inputs, the knee angle, βknee, is calculated in real-time and used to adjust the level of force delivered to the user based on the user's current posture. The target force Ftarget is then fed into a PID controller that is set to the desired force level.

Figure 6

Virtual leg model. IMUs placed at the shank and trunk measure the shank (αshank) and hip (αhip) angles. The motor encoder measures the length of the cable routed along the leg. These parameters are used to compute the knee angle βknee. This angle scales with the virtual leg length that is used to scale the anti-gravity assistance forces.

Figure 7

Experimental setup for the sitting transitions. Two force plates are used to evaluate ground reaction forces.

Figure 8

Example hip kinematics and ground reaction forces used to segment the recorded data offline for the analysis. Total ground reaction forces, hip angular velocity, and hip angular acceleration for the sit-to-stand and stand-to-sit transitions were measured using the force plates and the motion capture system. Circles indicate the time of the maximum change in vertical ground reaction forces. Starting from this event, hip kinematics were used to identify the beginning and the end of each transition (vertical line).

Figure 9

Myosuit characterization. Relation between movements of the knee and hip angles and changes in tendon length (measured as encoder counts). Separate measurements were conducted for each joint to define the relation between the change in knee angle and the encoder counts (A) and the hip angle and the encoder counts (B). For the knee angle, the relation obtained is 0.75 deg/rad; for the hip angle, the relation is 0.68 deg/rad. The relation between the cable force and the encoder counts was also determined experimentally (C) and gives us a measure of the suit's stiffness at 5.28 N/rad. These three relations were used to determine the knee angle βknee in real-time using the virtual leg controller.

Figure 10

Moment-angle curve of hip and knee joints during sitting transfers. The curves were experimentally determined. The relation (regression line) is used in the controller of the Myosuit. When forces are actively applied during movements with and against gravity, the relationship between hip and knee angles is assumed to be linear. This relation is used to scale the assistive forces of the Myosuit.

Figure 11

Experimental evaluation of the Myosuit during sit-to-stand (left) and stand-to-sit (right) transitions. (A) The βknee estimation from the Myosuit's controller is shown alongside the value measured with the motion capture system. The estimate from the Myosuit is delayed, on average, by 11.48% relative to the motion capture measurements. The flat line at 110° is due to a safety feature designed to prevent large forces from being applied while the subject is in a sitting posture. (B) Desired tendon force (gray) based on the calculated βknee and the actual tendon force as measured with the load cell (red). (C) Total knee moment based on motion capture data and ground reaction forces for the powered condition (solid black) and the transparent mode (dashed black). The knee moment of the Myosuit, based on tendon force and the knee lever arm, is shown in red. (D) Total knee power based on motion capture data and ground reaction forces for the non-assisted (dashed) and assisted (solid) conditions. The Myosuit power, shown in red, can be delivered for both sit-to-stand and stand-to-sit movements. (E) EMG recordings of the gluteus maximus (GLX, blue) and vastus lateralis (VAL, green) for the non-assisted (dotted) and assisted (solid) conditions.

Figure A1

EMG recordings of the gluteus maximus (GLX, blue) and vastus lateralis (VAL, green) for the non-assisted (dotted) and assisted (solid) conditions.