By H. Yu, G. Chen, and Z. Guo
An intelligent knee-ankle-foot robot, which is compact, modular, and portable, has been developed to help stroke patients to carry out gait rehabilitation at outpatient and home settings. The robot was designed with a novel compact compliant force controllable actuator for safe human robot interaction. The modular design for the knee and ankle joint caters for patients with different degrees and types of impairments. A prototype of the robot has been built and tested in the lab to verify the mechanism design and control.
Stroke is becoming one of the leading causes of adult disability, such as gait impairment. Robots have been developed to overcome the limitations of manual therapy for rehabilitation, but most of them are bulky, expensive, and available only to big hospitals . A significant portion of patients still have residual gait impairments, such as knee hyperextension and drop foot, after discharge from hospitals. Therefore, there is a great need for a home-based wearable robotic systems for gait rehabilitation. Numerous robots have been developed specifically for the ankle joint to tackle the drop foot problem, such as the MIT ankle robot  or to aid the knee joint such as Tibion . However, research aimed at providing active assistive torque to both the knee and ankle was very limited due to the added mechanical design complexity. The Knee-Ankle-Foot Orthosis (KAFO) presented in  is lightweight, but not portable because of the tethered operation with pneumatic actuation system. We present an intelligent compact and modular powered knee-ankle-foot orthosis for chronic stroke patients to conduct gait rehabilitation at outpatient rehabilitation centers or at private homes. We developed a novel compact compliant actuator and linkage mechanism to achieve light-weight and modular design.
Figure 1 shows the schematic diagram and prototype of the robot. The modular robotic system consists of an ankle foot module and a knee module (Fig.1(a)). Each module is driven with the same compact compliant force controllable linear actuator. It is known from human biomechanics that the range of motion of the lower limb joints is within 90o during normal walking . Therefore, a simple rocker-slider mechanism is used to achieve a compact design, which is optimized based on the human gait kinematics . The structure of the system is made of lightweight carbon fiber composite material. The total weight for the mechanical module is about 3.5Kg. The robotic orthosis has a suit of sensors for gait pattern detection, muscle activity monitoring and assistive control (Fig.1(c)). Potentiometers are employed to determine joint kinematics. Foot pressure sensor combining with inertia measurement unit (IMU) are used to detect gait initiation and phases. sEMG electrodes are placed at the main muscle group to monitor the muscle activity pattern. The system will deliver the optimal assistance force based on the sensory information. Quantitative measurement of the recovery progress can be provided to evaluate the effectiveness of the rehabilitation therapy.
Figure 1: The prototype (b) and schematic diagram (c) of the portable powered knee ankle foot orthosis. The modular ankle and knee are driven by a compliant force controllable actuator (a), sensors are attached to detect and monitor the state of human body during the rehabilitation therapy.
Compliant actuator is required to provide safe and force controllable interaction to human limbs. However the series elastic actuators (SEA) widely used for rehabilitation robot design has to compromise between force transmission and compliance: large output and low compliance cannot be achieved simultaneously. We developed a novel SEA to overcome this limitation, extending a large range of output force while keeping a low intrinsic compliance . As shown in figure 1(a), we use a very soft linear spring at the output to handle the low-force range and ensure truly low intrinsic compliance and high force fidelity. We introduce a torsion spring directly between the motor and the ball screw in the high speed range. Due to its high reflected stiffness, the torsion spring can handle the high-force operation when the soft spring is compressed, ensuring continued force output and large bandwidth at high force range. As both springs are small, the actuator design is very compact. The actuator for this robot has a total weight of 0.85 Kg, but can transmit over 1000N force.
According to biomechanics analysis of human gait, the average torque is less than 30% of the peak torque for the knee and ankle joint . Thus the low-force range is selected to be 30% of the desired force so that the robot is working in the compliant range during most of the gait cycle. The torsion spring is used to control the high force and provide peak force after the linear spring is fully compressed.
Figure 2(a) depicts experimental results of the actuator force control with an excellent force tracking ability and large bandwidth . Specifically, Force error was kept below 5% in both low and high-force range. Force control bandwidth of 15 Hz for the low force range and 30 Hz for the high force range has been achieved. No disturbance or oscillation was observed when the linear spring fully compresses and controller switches. Preliminary impedance control strategy has been implemented for gait trajectory control. Figure 2(b) shows the gait trajectories of the ankle joint with different virtual stiffness, which illustrates that the trajectory is closer to the reference with higher stiffness. The robot achieved stable force and gait control.
Figure 2: The experimental results of the actuator and ankle robot. Force tracking in both low and high force range (a), impedance control on ankle joint with various virtual stiffness (b).
With a novel compact compliant actuator and compact linkage mechanism, we developed a modular and portable knee-ankle-foot orthosis for gait rehabilitation. To our best knowledge, this is the first lower limb exoskeleton with both powered knee and ankle joints driven by electrical motors. Moreover, our robot achieves truly intrinsic compliance and in the same time can have high force control bandwidth. We are currently develop more advanced gait control strategies based on the gait initiation and gait phase detection with IMU sensors  so that we can provide the optimal force assistance to compensate the specific gait deficiency of each individual patient.
This research is funded by the FRC Tier 1 research grant R397-000-114-133 and EDIC Seed Fund R-261-503-002-133, Faculty of Engineering, National University of Singapore.
For Further Reading
1. Riener, R., L’unenburger, L., Jezernik, S., Anderschitz, M., Colombo, G., and Dietz, V., “Patient-cooperative strategies for robot-aided treadmill training: First experimental results,” IEEE Trans. Neural Syst. Rehabil. Eng., vol. 13, no. 3, pp. 380â€“395, Sep. 2005.
6. Yu, H., STA Cruz, M., Chen, G., Huang,H., Zhu,C., Chew,E., Ng, Y. S., and Thakor, N. V., “Mechanical Design of a Portable Knee-Ankle-Foot Robot,” in Proc. IEEE Int. 2013 Conf. Robotics and Automation (ICRA), p. 2175-80, May 6-10, 2013, Karlsruhe, Germany.
7. Yu, H., Huang, S., Chen, G., Thankor, N., Toh, S. L., Cruz, M., Ghorbel, Y., and Zhu, C., “A Novel Compact Compliant Actuator Design for Rehabilitation Robots,” in Proc. IEEE Int. Conf. Rehabil. Robot.(ICORR), June 24-26, 2013, Seattle, USA.
8. Meng, X., Yu, H., and Tham, M. P., “Gait Phase Detection in Able-bodied Subjects and Dementia Patients,” in Proc.35th IEEE Annu. Int. Conf. Eng. Med. Biol. Soc.(EMBC), July 3-7, pp. 4907-4910, 2013, Osaka, Japan.
Haoyong Yu is an Assistant Professor of Department of Biomedical Engineering at NUS. He received his PhD in Mechanical Engineering from MIT in 2002. His current research at NUS focuses on robotics for neurorehabilitation, assistive technologies, robotics in surgery, and bio-inspired robots. Read more
Chen Gong is currently working towards the Ph.D. degree in Biomedical Engineering, NUS with his supervisor Dr. Yu Haoyong. His current research interests include rehabilitation robots system, compliant actuator and control theory. Read more
Guo Zhao is a Research Fellow in the Department of Biomedical Engineering, National University of Singapore. He received his Ph.D. degree in Mechatronics Engineering from the Institute of Robotics, Shanghai Jiao Tong University, China, in 2012. His research interests include neurorehabilitation exoskeleton, neuromuscular modeling, and neuromuscular-model based robotic control. Read more