A recent lower limb exoskeleton robot for gait rehabilitation: a review

4.1. Lokomat

The Lokomat, developed by Hocoma AG in Switzerland, is a bilateral gait rehabilitation robot with a body weight support (BWS) system. The patient's legs are moved in the sagittal plane using DC motors connected to ball screws. The hip and knee joints of the Lokomat are powered by engines with linear ball screws [28]. In contrast to the Lokomat, the latest generation of exoskeletons is designed to focus on low mechanical impedance. The Lokomat is equipped with cutting-edge active actuators located at the hip and knee joints, which work in concert to guide a patient’s limbs expertly. These sophisticated devices enable the Lokomat to ensure precise movement in the sagittal plane, meticulously following a predetermined trajectory designed for effective rehabilitation. This technology allows patients to experience tailored guidance that promotes proper gait patterns and enhances their recovery journey. Furthermore, the Lokomat features an impedance controller that modifies the “guidance level”, which defines the permissible degree of deviation from the intended limb movement pattern. This integration guarantees optimal support and rehabilitation for patients [9].

Woodway has developed KineAssist and Lokohelp in the U.S., while Maibu Robotics in China has created the MoonWalker. These innovative exoskeleton robots are engineered to lighten users' load while improving their running efficiency significantly. By utilising advanced materials and technology, these devices support the wearer’s movements, allowing for more incredible speed, less fatigue, and enhanced performance during physical activities. The MoonWalker includes customisable movement modules, adjustable intensity levels, visual control options, and real-time feedback to improve user experience and adaptability [9].

The updated version of Lokomat features an optional FreeD module that enhances therapy by enabling pelvic lateral translation and transverse rotation [28].

The Lokomat is the first gait orthosis designed to assist individuals with walking difficulties using a treadmill. It coordinates the movement of the hip and knee joints, and for the ankle joint, it employs an ankle-foot orthosis equipped with a plantar flexion stop spring [30].

The Lokomat’s trajectories can be customised to suit individual patients and their specific step lengths [28]. It is one of the more well-researched stationary robotic systems developed to support and automate treadmill training [31].

4.2. ReoAmbulator

ReoAmbulator, developed by Motorika Ltd. and marketed in the USA as the “AutoAmbulator”, is a robotic system that supports body weight during treadmill use [3]. Robotic arms are strapped to the patient’s legs at the thigh and ankle, driving them through a stepping pattern. A single-masked, randomised clinical trial is underway to assess the effectiveness of ReoAmbulator in stroke patients. ReoAmbulator was developed in collaboration with the HealthSouth network of rehabilitation hospitals [29].

The ReoAmbulator robotic system, also referred to as “AutoAmbulator”, is an example of a treadmill gait trainer designed for lower-limb rehabilitation therapy, as illustrated in Fig. 2(b). The ReoAmbulator system is widely used in research centres and medical hospitals for rehabilitation therapies and educational studies on motor function recovery. Central to this system is a powered leg orthosis with advanced robotic arms that help patients engage in gait motion while providing additional walking force, enhancing rehabilitation. The robotic arms are attached to the thigh and ankle, allowing for a controlled stepping pattern tailored to each patient’s needs. This technology not only aids physical movement but also promotes neurological adaptation for better locomotion. Studies show that robot-assisted gait training can significantly improve balance and gait, comparable to traditional manual therapies. The ReoAmbulator stands out among commercial rehabilitation orthoses due to its significant development and the introduction of various prototypes that enhance its effectiveness in diverse rehabilitation settings [33].

Multiple sensors continuously monitor and adjust power and speed according to the patient’s needs. An interactive display with various modes boosts motivation and adds variety to training exercises. The ReoAmbulator’s sensors can assess mobility before and after training, allowing Motorika to customise experiences and track progress. The exoskeleton is connected to a virtual reality simulator featuring diverse terrains and games to enhance engagement. It also provides verbal and audio feedback. One mode displays an ideal step pattern for the user to replicate, with adjustments possible on the spot. A side panel enables a professional to control the system directly. The ReoAmbulator is a specialised device intended for individuals engaged in gait rehabilitation. Its primary objective is to enhance various aspects of locomotion, including gait patterns, balance, weight-bearing, coordination, speed, and endurance during walking activities. This innovative tool supports rehabilitation by providing targeted assistance and promoting functional improvement [34]. This system operates in a tethered manner akin to the Lokomat [35].

4.3. Walkbot

The Walkbot is an advanced fixed-frame exoskeleton designed explicitly for gait rehabilitation. It is offered in three distinct models to accommodate varying needs: Walkbot_S (standard model), Walkbot_K (pediatric model), and Walkbot_G (next-generation model) [36]. The Walkbot series of gait rehabilitation exoskeletons is an innovative advancement by the South Korean company P&S Mechatronics. Notably, two of the three models are designed for pediatric use through their inherent design or by incorporating additional modules. The Walkbot stands out from other fixed-frame gait rehabilitation exoskeletons due to its powered hip-knee-ankle design, which enhances its functionality. It is an effective evaluation device and features the Walkbot_G model, equipped with an integrated pressure plate to monitor each step in real-time. Furthermore, the device’s full integration with virtual reality provides an engaging platform for interactive training, demonstrating its commitment to improving rehabilitation outcomes.

The Walkbot system is an innovative gait orthosis designed to precisely coordinate hip, knee, and ankle movements through advanced natural gait training techniques. It addresses foot drop and toe drag by controlling ankle dorsiflexion and plantar flexion [30].

The device is equipped with a suspension harness designed to provide body weight support (BWS), a motorised treadmill, and an actuator-controlled exoskeleton (Fig. 3(c)). The treadmill’s speed and torque can be adjusted to provide both assistance and resistance, allowing for customisation according to the patient’s level of locomotor performance as training progresses [37].

4.4. MRG-P110

The “HIWIN” Robotic Gait Training System, specifically the model MRG-P110, is an advanced, fixed-frame medical robot tailored for gait rehabilitation. Unlike traditional rehabilitation methods, this innovative device remains stationary, allowing patients to benefit from its support while their legs are gently guided through a walking motion. This unique approach enables the rest of the patient’s body to remain stable, minimising unnecessary movement.

The MRG-P110 facilitates rehabilitation by assisting users in effectively retraining their walking gait cycle. It operates like a sophisticated powered leg exoskeleton firmly anchored to the ground. As the patient “walks” across specially designed actuated plates, the exoskeleton provides the necessary power and guidance to promote proper leg movement.

Moreover, this state-of-the-art device is fully integrated with virtual reality technology, creating an interactive and engaging training environment. This integration enhances user motivation and provides valuable feedback to optimise the rehabilitation experience, helping individuals regain mobility and confidence in walking [40].

Fig. 4Lower limb exoskeletons are designed for gait rehabilitation at a more accessible cost

a) MRG-P110 [7]

b) Gait simulator

4.5. Gait simulator

The Gait Simulator is an innovative rehabilitation device specifically crafted to restore and enhance motor skills in the lower extremities, primarily in Russia’s medical care context. This advanced simulator plays a crucial role in the recovery process for patients who have experienced conditions affecting the brain and spinal cord.

The use of this specialised medical simulator is particularly beneficial for individuals dealing with various challenges, including paresis (weakness), paralysis (loss of movement), and muscle contractures (shortening of muscles). It is also indicated for patients suffering from the lasting effects of cerebral palsy, various neurological disorders, and any conditions that lead to compromised mobility or heightened spasticity in the legs.

The unique design of the Gait Simulator allows patients to initiate motion through their arms. Users can synchronise the movement of all simulator components by moving their arms (or even just one arm), leading to the involuntary movement of their otherwise stationary lower limbs. For patients who have sustained injuries to the cervical and upper thoracic regions of the spinal cord, additional supportive apparatus is available to ensure safety during rehabilitation.

In these cases, patients are carefully positioned upright, and this support enables them to participate in a comprehensive range of rehabilitation exercises, all aimed at fostering strength, coordination, and improved mobility in their lower extremities. The Gait Simulator thus offers a holistic approach to rehabilitation, empowering patients on their journey to recovery.

In the given examples (Table 1), five exoskeletons that compose the “lower body fixed rehabilitation” group are fully explained based on their features; only one is passive and is under our subsequent research.

5.1. Gait anatomy for the prototype

The lower limb gait system primarily comprises a network of bones and joints, as illustrated in Fig. 5. There are seven types of bones, nearly 300 muscles, and three joints: hip (provided 3 DOFs), knee (provided 1 DOF), and ankle (provided 3 DOFs) [44-46]. Lower body movements are categorised into three primary groups: flexion and extension, adduction and abduction, and internal and external rotation. This classification provides a clear framework for understanding the functional mechanics of lower body dynamics [4], [47-48].

Fig. 5The joints and bones of the lower body

Fig. 6The sagittal plane of the lower body

Lower body exoskeletons can be targeted at single-joint or multi-joint levels. While the single joint may be applied to one specific joint, the multi-joint may involve a combination of lower joints [49].

To make the gait rehabilitation exoskeleton have similar movement functions to the lower body, some researchers sometimes select the 15 DOFs, but this kind of design makes the system more complex [50]. Our research focused on this situation in the sagittal plane based on Table 2 as follows as a multi-joint:

1) Hip joint connected to pelvic and femur bone, allowing flexion/extension rotation [51].

2) The knee joint contains the patellofemoral joints, and its movements support the flexion/extension during walking [51].

3) The movements of the ankle joint primarily occur in the talocrural joint, allowing for plantar flexion and dorsiflexion [28], [51-52].

The gait patterns observed in patients exhibit a high degree of complexity arising from their compromised motor functions. However, we can characterise the gait process via step length, width, and speed parameters. Our studies focused on the effects of various parameters on joint angles and their relationship with gait parameters. Reference [48] offers an in-depth exploration of the essential movements involved in human walking gait, providing a detailed examination of each phase and the intricacies of how they contribute to efficient locomotion. This analysis highlights the dynamic interplay of various muscle groups and joint mechanics that facilitate smooth and coordinated walking patterns. The range of motion has a significant role in exoskeletons. In [4], the range of motion is discussed and represented in the table of human lower extremity ranges, including the joints, movements, and values of rotate angels, compared between the sources: American Agency of Orthopedic Surgeons (1965), American Medical Association.

Table 2DOFs, bones, and joint movements of the lower body in the sagittal plane

Joints	Bones	DOF	Joint movements	Movement activities	Examples
Hip joint	Femur	1	Flexion-extension		Honda Walking, Assist, HiBSO, PH-EXOS, Exosuit
Knee joint	Tibia, patella	1	Flexion-extension		Knee Exo/Carnegie Mellon University, EXO
Ankle joint	Fibula	1	Dorsiflexion-plantar flexion		MIT ankle exoskeleton, BIT Ankle, Knee soft exosuit

5.2. Mechanical design for the prototype

Mechanical design plays a vital role in the functionality of lower limb rehabilitation exoskeletons, as it ensures the efficient transmission of force and energy through its integration with the human lower body. The anthropometric dimensions are also a significant part of the design concept [53-54]. These objectives can be accomplished by carefully designing suitable robotic mechanisms and actuation systems. [28]. In the design process, we must consider motion trajectory design, the comfort of human-machine interaction, and the effectiveness of rehabilitation activities [55-56].

Fig. 7 shows the renewed prototype of exoskeletons, according to the anatomy structure of the lower body explained in Section 5.1. Electric motors drive the prototype's mechanical design.

Fig. 7Mechanical design of the prototype

The kinematic structure of exoskeletons is the central part of mechanical design and is distinguished depending on the category of its constituent elements. It has two types [20].

Soft exoskeletons. Soft exoskeletons consist of flexible and lightweight elements that aid human movement. They are based on passive-assistive control and are wearable, mobile, and portable [76].

Rigid exoskeletons. The rigid exoskeleton consists of actuators, joints, and kinematic joints that enable and enhance human movement. In the medical field, two distinct types of exoskeletons are commonly employed, with specific designs incorporating features from both types for improved functionality.

5.3. Control structure of the prototype

The design of lower limb exoskeletons involves several essential components that work together to enhance mobility and support for the user. These components include advanced sensing systems that detect movements and provide real-time feedback, powerful actuation mechanisms that generate the necessary forces for movement, and sophisticated control strategies that ensure seamless coordination and responsiveness between the exoskeleton and the user’s intentions. Together, these elements create a functional and efficient device capable of improving mobility and aiding rehabilitation. These elements are detailed in references [46] and [56-57]. The control strategy is mainly due to the complex interaction between humans and robots and the human-robot interaction’s cognitive and physical aspects [58].

The control strategy for lower limb exoskeletons encompasses several critical components, including defined gait patterns, specific training objectives, a standardised evaluation framework for rehabilitation, and relevant motion data. Exoskeleton control strategies used in gait rehabilitation can be broadly categorised into two primary types. These categories encompass various approaches designed to enhance mobility and improve walking patterns for individuals undergoing therapy: (1) trajectory tracking and (2) assistance as needed. This classification provides a clear framework for understanding the approaches utilised to enhance gait recovery. In trajectory tracking, gait pattern activities were preprogrammed, like a normal gait. Gait patterns include gait time, interjoin coordination, and hip, knee, and ankle angles. The primary purpose of rehabilitation gait training is to restore the functions of disabled patients to normal levels in the lower body.

The most prevalent methodologies employed in the control of exoskeletons are force-based control and position-based control. These approaches are widely recognised for their effectiveness in enhancing the functionality of exoskeletal systems [59]. The specific control systems for exoskeletons should be selected individually or combined; however, a combined approach complicates the evaluation of the system. In the development of rehabilitation, exoskeletons are usually controlled position-based; in some resources, it is discussed as a “gait trajectory control” [60-61]. This control system is based on desired prerecorded exoskeleton positions (e.g., joint angles).

Position control and trajectory tracking guide the patient's lower limbs by providing feedback that aligns with fixed reference walking trajectories based on joint angles. It includes an internal control loop using the error between the reference angle trajectories and the angles measured by the sensors at each lower limb joint. A positive point of this control strategy is the imposition of a predefined joint angle trajectory resulting in limited kinematic error, an influential factor in driving human motor learning [75].

Torque control enhances versatility, robust behaviour, and dependable and safe tasks in the presence of humans. Therefore, torque control is widely employed in exoskeletons. This control method uses the error between the reference torques and measured torques by sensors on each actuator. Since torque control closely interacts with actuators, it delivers a simple way of controlling energy flow from the exoskeleton to the user, which is helpful in biomechanics research [75].

Fig. 8The Human-exoskeleton collaboration system

Control system design has been developed in the last decade, but there are still some dissolved problems related to human-robot integration in control systems [28], [62]. One of the largest is whether the patient should organically combine with the robot. The solution appears to integrate the human body and robot properly.

Moreover, the torque control method is critical in implementing the rigid-body inverse-dynamics control strategy. Although position control is common in robotics, it is not appropriate for all tasks, particularly when robots need to interact physically with the environment.

5.4. Actuators for the prototype

Currently, electrical motors are used wider than hydraulic and pneumatic types [63-64], more discussed in [59-60], [65].

The pneumatic actuator operates using a pressurised air source, which can be advantageous; however, it may have some limitations, such as lower power output and potential challenges in control. The principle of operation of pneumatic actuators is like that of hydraulic actuators, but it uses compressed inert gases instead of liquids. It also allows compressed air from the tank to increase the pressure inside the piston or diaphragm, causing it to expand or reduce the pressure at the outlet and cause it to contract [73].

Hydraulic actuators are typically characterised by their substantial size and higher cost, which can be attributed to their significant power requirements and the necessity for stability [48]. They are mainly used in Exoskeletons, designed to increase the user's strength in industrial or military applications. The principal advantage of a hydraulic actuator resides in its capacity to generate substantial power output. Still, on the other hand, the complexity of working with high-power liquid pumps and the weight of the liquid pumps and pipes are its disadvantages [73-74].

Being controllable of electric actuators is more suitable than others. The most used electrical motors are servo motors capable of advanced position-based control and can control many degrees of freedom, from 1 to 22. There are also other types of electric motors: AC motors, Brush DC Motors, and Stepper motors [4]. DC motors are well-known for their high torque-to-weight ratio, low noise, and reliability. One of the more feasible electric motors is a battery used to power electric actuators [44], [59]. Electric actuators will lead to the development of modern exoskeletons, as their parameters are size, weight, torque, and velocity. In developing lower limb exoskeletons, it is recommended to prioritise the placement of actuators on a single joint. This approach can enhance functionality and efficiency in the overall design. This approach facilitates the simultaneous control of multiple joints and allows for coordinated movement across two or more joints. This approach can enhance the overall functionality and effectiveness of the exoskeleton system. The Series Elastic Actuator (SEA) drive marks a considerable technological advancement, characterised by its ability to enhance performance by integrating elastic elements. This design improves precision and compliance in motion control applications, making it an asset in various engineering fields. It provides exceptional control precision, enhances safety, mitigates inertia impact, and improves energy efficiency by reducing friction losses. This innovative solution is poised to meet high-performance standards in various applications [66].

Actuator-related forms of exoskeletons can use single-class actuators or different-class actuators. However, in practice, one-class actuators are mainly used. Although using multi-class actuators is theoretically correct, it is more complicated.

5.5. Sensors for the prototype

In the past decade, sensors have become the central part of the exoskeletons, making them more attractive and safer. According to the types of signals in medical exoskeletons, mechanical sensors help to control the position, altitude, force, torque, pressure, and acceleration based on mechanical signals (or physical signals) are measured by potentiometers, piezoelectric sensors, capacitive sensors, accelerometers, gyroscopes, a foot sensitive resistor [44], etc. Neural and brain signals (or biosignals) need to sense human body elements during the rehabilitation training [67] to analyse the patient’s recovery and plan the following training proposals. Some offered sensors for measuring the bio-signals: EMG sensors [44], EEG, fNIRS, and MMG [8].

In addition, it has a positive effect on the efficiency of exoskeletons. Exoskeletons can be categorised into various types depending on their specific sensor environments. These categories reflect the different ways in which these advanced devices interact with their surroundings and respond to external stimuli [8], [10].

Exoskeletons directly to the physical sensing system. The inertia, force, and displacement sensors control the exoskeleton and measure kinematic parameters. They help track the exoskeleton’s positions, movements, forces, and the user’s body. Inertial sensors measure the orientations, positions, and movements of exoskeletons and patients using accelerometers, gyroscopes, and magnetometers, often combined into IMUs. Force sensors are employed to accurately measure the forces and torques exerted on joints, which facilitates a comprehensive evaluation of human-robot interactions. The exoskeletons can contribute to the system’s accuracy and control using the displacement sensors.

Fig. 9Sensor’s cooperation

Environmental sensors for exoskeletons are essential in improving their interaction with the environment. Various sensors, including ultrasonic, infrared, and LiDAR, can seamlessly integrate these advanced mobility aids. These sensors facilitate critical functions such as detecting real-time obstacles, avoiding potential collisions, and enabling smooth navigation through diverse terrains. By providing essential data about the exoskeleton's surroundings, these sensors significantly improve safety measures and increase adaptability, allowing users to manoeuvre confidently in different environments, whether indoors or outdoors. Exoskeletons equipped with human signal monitoring can control biological signals, enabling the analysis and determination of the patient’s physiological state. These devices primarily utilise sensors such as EMG (electromyography), MMG (mechanomyography), and EEG (electroencephalogram) to offer valuable insights into the patient’s physical condition, levels of exertion, and cognitive processes.

In general, incorporating the three types of sensors into exoskeletons will make them more expensive and advanced but also shorten, more effectively, and safer the patient’s treatment process.