WO2016056998A1

WO2016056998A1 - Assistive gait device

Info

Publication number: WO2016056998A1
Application number: PCT/SG2015/050377
Authority: WO
Inventors: Kin Huat LOW; Peh Er Adela May TOW; Lei Li; Pang Hung LIM
Original assignee: Nanyang Technological University; Tan Tock Seng Hospital Pte Ltd
Current assignee: Nanyang Technological University; Tan Tock Seng Hospital Pte Ltd
Priority date: 2014-10-07
Filing date: 2015-10-07
Publication date: 2016-04-14
Anticipated expiration: 2017-04-07

Abstract

An assistive gait device comprising a motorized exoskeleton configured to be coupled to a person and independently move each leg of the person. The device also comprises a first motorized stabilizer provided on the left side of the person and a second motorized stabilizer provided on the right side of the person, wherein the two stabilizers are connected to the exoskeleton and each of the two stabilizers is moved independently relative to the exoskeleton and is able to selectively contact the ground. A controller and sensory system is also provided, that comprises an input device and is configured to coordinate and actuate independent movement of the exoskeleton and the two stabilizers, thus automatically effecting a walking movement of the person. In a preferred embodiment, movement is effected by forward movement of the two stabilizers, followed by walking sequence of the motorized exoskeleton, which is initiated by the person activating the input device.

Description

ASSISTIVE GAIT DEVICE

FIELD OF THE INVENTION

This invention relates to an assistive gait device for assisting a human person with impaired movement ability to walk.

BACKGROUND OF THE INVENTION

The complex functions of human sensory and motor systems enable people to coordinate different segments of the body to achieve different tasks such as stable locomotion. However, they are also vulnerable to diverse disease such as SCI (spinal cord injury), stroke, or aging. These diseases which create deficits in motor control and muscle weakness affect independent living of patients. Although rehabilitation process can enhance patient independence and improve their walking abilities, most of them still need assistive device such as orthoses or crutches to compensate for the degeneration of the motor control ability or muscle weakness.

The burden of care for person with neurological disease is well-known to all. In Singapore, persons with stroke, spinal cord injuries and brain injuries are often left with residual neurological deficits to varying degrees. A significant burden of care exists for those with moderate to severe disabilities, who are unable to ambulate on their own and required the constant assistance of a caregiver.

Current mobility aids such as canes and walkers are safe and effective for patients with milder degrees of paralysis. However, patients with greater degrees of paralysis or unsteady gait require devices offering more support to enable them to ambulate. In addition, current devices that offer greater amounts of support have limitations which hindered their acceptance both during patients' rehabilitation stage and even more so during patients' usage at home. Broadly speaking, the highly supportive but 'passive' devices are heavy and required high energy expenditure to walk. The level of assistance from caregivers remains high, and the lack of stability assurance meant that at least two or more caregivers' help is required. Such devices tend not to be used after a while because of poor fit and high energy expenditure associated. Furthermore, patients using such gait devices also require sufficient upper body strength to control upright posture and also manipulate crutches/walking frames to partially support body weight and keep balance. Consequently, most quadriplegics are excluded. Paraplegics using the device will also need to use a significant amount of upper limb strength to maintain stability in walking and standing.

Robotic exoskeletons have drawn much attention recently due to their potential ability to help stroke and spinal cord injury patients regain the ability of walking. However, the biggest challenge is the balancing of the exoskeleton and how it can balance is still an open question. Most of the time, patients using such exoskeleton devices require sufficient upper body strength to control upright posture and also manipulate crutches/walking frames to partially support body weight and keep balance. However, high energy cost and the high potential of falling during using these devices remains a problem.

In recent years, the number of exoskeletons for the legs has increased both in research areas as well as in the market. In application of power augmentation (mostly for military uses), the current exoskeletons focus on increasing the user's strength and endurance in carrying heavy load for long distance. In rehabilitation and assistive applications, the exoskeletons focus on assisting a person to walk which includes support of the body weight with the stance leg and transporting the trunk, and moving the foot forward by assisting the swing leg. In both functions, current exoskeletons do not provide the function of maintaining postural balance. In addition, most exoskeletons create a burden in maintaining postural equilibrium for the user [21]. The recent development of robotic orthoses or exoskeletons aims at the possibility of home-based rehabilitation together with assistive ability. Some well-known devices are summarized in Table 1 below:

upper body; limited foot support area which resulst in the system having a high chance of falling when there are perturbations.

Table 1

In addition, there are also the Mina [4], and Vanderbilt exoskeleton [5]. These devices require the user to balance themselves through the using of crutches. Consequently, most quadriplegics are excluded from using these devices since paraplegics using the device will also need to use a significant amount of upper limb strength to maintain stability in walking and standing.

At the same time, there are possibilities of falling when the human user fails to control the crutch properly. Thus, these devices are still constrained to be used in a controlled environment. The problem identified is the under-actuation of most of the exoskeleton devices (usually with only hip and knee flexion motion are active controlled). Thus such devices require the user to use the upper body to provide propulsion force and increase the stability of the system. Full body exoskeletons which power all important joints of the human lower limb is one of the solutions, such as Rex [6, 25]. However, due to its limited foot support area, it has to sacrifice walking speed and the human-like gait pattern in order to achieve stable walking. Several recently developed full-body exoskeletons with compliant actuation [7-9] may lead to a more stable and human-like gait. However, the ability to deal with large disturbances from the human user and the environment will still be limited due to the limited foot contact area in single stance phase. To solve this problem, wheeled systems have been developed. In these devices, balance is secured by a larger support polygon formed by the wheels. At the same time, mechanisms which can help the movements of the pelvis are added to help body weight shifting. Some representative devices are: KineAssist [10], and WalkTrainer [11] and Nature-Gaits [12]. Several advantages of wheeled systems such as simple structures, maneuverability and high efficiency on the prepared surface have led to their wide usage. However, the ability of wheeled systems to overcome irregular terrain is limited by wheel diameter. This limits their usage in different environments. SUMMARY OF INVENTION

The assistive gait device of the present invention comprises a robotic exoskeleton having balance stabilizer mechanisms while a trajectory generation method generates a dynamically stable and tunable gait pattern.

The novel balance stabilizer mechanism is designed to incorporate itself with a robotic lower limbs orthotic device. The stabilizer functions to maintain stability during walking, to facilitate body weight shifting, thereby reducing energy cost and number of helpers required.

The proposed balance stabilizer mechanism is able to provide active balance assistance for exoskeleton robots.

The assistive gait device consists of two modules. The first module is the exoskeleton leg which assists the patient to walk by supporting their stance leg and moving their swing leg to bring the foot forward. Each exoskeleton leg comprises a shank link and a thigh link. All the joints allow flexion or extension of the lower limb, while additional inversion/eversion and rotation degree of freedom are added to the ankle and hip joints to allow lateral body weight shifting and turning motion. The second module is the balance stabilizer which ensures the maintaining of balance of the user by providing essential support points. The balance stabilizer consists of three joints: two rotational joints and one prismatic joint and it is coupled to the trunk of the user to provide support forces.

According to a first aspect, there is provided an assistive gait device comprising: a motorized exoskeleton configured to be coupled to a human person and configured to independently move each leg of the human person; a first motorized stabilizer provided on the left side of the human person and a second motorized stabilizer provided on the right side of the human person, the two stabilizers connected to the exoskeleton, each of the two stabilizers configured to be moved independently relative to the exoskeleton and to be able to selectively contact the ground to stabilize the human person; and a controller and sensory system comprising an input device and configured to coordinate and actuate independent movement of the exoskeleton and two stabilizers to automatically effect walking movement of the human person upon the human person initiating a walking sequence by activating the input device. The exoskeleton may comprise, for each leg, a shank link connected to a thigh link via a motorized knee joint configured to allow knee flexion and extension, the shank link connected to an exoskeleton trunk via a motorized first hip joint configured to allow rotation of the thigh link about a mediolateral axis, and a motorized second hip joint configured to allow rotation of the thigh link about a vertical axis.

The two stabilizers may each comprise a motorized first rotational joint configured to allow rotation of the stabilizer about the mediolateral axis, a motorized second rotational joint configured to allow rotation of the stabilizer about an anteriorposterior axis, and a motorized prismatic joint configured to allow the stabilizer to lengthen and shorten.

The input devices may comprise at least one of: a push button, a touchscreen and a joystick.

The controller may comprise a control algorithm in which a stable trajectory is generated based on a Zero Moment Point criteria and a whole body Centre Of Gravity Jacobian method is used to filter unbalanced motion.

The walking sequence may be preceded by forward movement of the two stabilizers, the forward movement of the two stabilizers being activated as start of motion by the human person via the input device.

The walking sequence may comprise the exoskeleton swinging forward a foot of the human person and placing the foot on the ground.

BRIEF DESCRIPTION OF FIGURES

In order that the invention may be fully understood and readily put into practical effect there shall now be described by way of non-limitative example only exemplary embodiments of the present invention, the description being with reference to the accompanying illustrative drawings.

Fig. 1 is a front view of an embodiment of an exoskeleton with balance stabilizers.

Fig. 2 is a back view of an embodiment of an exoskeleton with balance stabilizers.

Fig. 3 is a side view of an embodiment of an exoskeleton with balance stabilizers.

Fig. 4 shows the embodiment of stabilizer module. Fig. 5 shows the embodiment of ankle joint module.

Fig. 6 shows the embodiment of back support module.

Fig. 7 shows the embodiment of active hip and knee joint module.

Fig. 8(a) illustrates movement definition of the pelvis

Fig. 8(b) is a schematic view of balance stabilizer mechanisms coupled with a human person. Fig. 8(c) is a schematic view of a six-DoF design balance stabilizer mechanism coupled with a human person.

Fig. 9(a) depicts a human with exoskeleton coupled thereon.

Fig. 9(b) is a perspective view of a balance stabilizer mechanism

Fig. 9(c) is a perspective view of the balance stabilizer mechanism of Fig. 9(b) coupled with the exoskeleton of Fig. 9(b)

Fig. 9(d) shows detailed designs of the exoskeleton of Fig. 9(a).

Fig. 10 shows a leg motion realization using the exoskeleton with balance stabilizers.

Fig. 1 1 shows pelvis rotation realization using the exoskeleton with balance stabilizers.

Fig. 12 shows lateral pelvis motion realization using the exoskeleton with balance stabilizers. Fig. 13 shows sit to stand realization using the exoskeleton with balance stabilizers.

Fig. 14 shows four different support configurations and their corresponding workspace.

Fig. 15 shows a test configuration of a robotic exoskeleton with balance stabilizer

Mechanism.

Fig. 16 is a schematic illustration of a control system with a controller and sensory system.

Fig. 17 is a flow chart of a control algorithm for the exoskeleton with balance stabilizers.

Fig. 18 is a flowchart of gait pattern generation algorithm.

Fig. 19 is illustrates planning of hip and ankle trajectories.

Fig. 20 is a graph of hip and knee motion angles during walking.

Fig. 21(a) shows ZMP trajectories before modification relating whole-body motion to COG motion.

Fig. 21(b) shows ZMP trajectories after modification relating whole-body motion to COG motion.

Fig. 22 shows a motion sequence for robotic exoskeleton with balance stabilizer mechanism. Fig. 23 are photographs of a clinical trial of the assistive gait device with a tetraplegic subject.

DETAILED DESCRIPTION

Exemplary embodiments of the assistive gait device 10 will be described below with reference to Figs. 1 to 23 in which the same reference numerals are used to denote the same or similar parts.

The system 100 consists of two parts: the lower-limb exoskeleton 3 and a couple of balance stabilizer mechanisms or stabilizers 1 , 2 as shown in Fig. 1. The lower-limb exoskeleton 3 is configurable to fit persons of different sizes by changing the length of the thigh link 6 and shank link 8. The knee joints 7 of the exoskeleton 3 are configured to allow knee flexion and extension (as indicated by the arrows 9 and 10 respectively) between the thigh link 6 and shank link 8. Both the thigh link 6 and the shank link 8 are coupled to the person's thigh and shank respectively using straps. The exoskeleton 3 further compromises an exoskeleton trunk 4 which is coupled to the person's body, also using straps. The trunk 4 is connected to the thigh link 6 through a hip extension/flexion joint 5 (extension and flexion indicated by the arrows 11 and 12 respectively) that allows rotation of the thigh link 6 about a horizontal axis (i.e., the mediolateral axis), and hip rotation joints 27 (rotation indicated by the arrows 28 and 29 respectively) that allow rotation of the thigh link 6 about a vertical axis, as shown in Fig. 6. The hip joint for abduction/adduction motion is locked such that abduction/adduction is not possible.

The last part of the exoskeleton 3 is the ankle foot mechanism 50 comprising two ankle joints 21 , 22. The first ankle joint 21 allows dorsiflexion/plantar flexion as indicated by the arrows 23 and 24 respectively. The second ankle joint 22 allows eversion/inversion as indicated by the arrows 25 and 26 respectively. In the exoskeleton 3, both the joints 21, 22 in the ankle foot mechanism 50 are passive, while the joints 5, 27, 7 in the hip and knee are set to be active (driven by motor).

The stabilizers 1, 2 are provided on the left and right side of the person respectively, and connected to the trunk 4 of the exoskeleton 3. Each stabilizer 1 , 2 is configured to move independently relative to the exoskeleton 3. Each stabilizer 1 , 2 is designed to have three degrees of freedom with two rotational joints 13, 14. The first rotational joint 13 is configured to allow rotation about the mediolateral axis as indicated by the arrows 16 and 17 respectively. The second rotational joint 14 is configured to allow rotation about the anterioposterior axis as indicated by the arrows 18 and 19 respectively. Each stabilizer also comprises one prismatic joint 15 to allow the stabilizer 1 , 2 to lengthen or shorten as indicated by the arrow 20. All joints 13, 14, 15 of each stabilizer 1 , 2 are active. The two rotational joints 13, 14 allow the stabilizer 1 , 2 to be placed at any support position required while the prismatic joint 15 driven by a linear actuator will provide the support and propulsion force. All the joints 13, 14, 15 are designed to be active, i.e., motorized, by using a harmonic drive 31 coupled with a maxon flat motor 30, as shown in Fig. 7. At each joint 13, 14, 15, a potential-meter 32 is installed to obtain the absolute joint angle. The stabilizers 1 , 2 are provided on the left and right side of the person respectively, and connected to the trunk 4 of the exoskeleton 3

Controlling CoM (Center of Mass) of the body is essential in achieving stable walking. The first step in designing the stabilizer mechanism 1 , 2 was to identify the factors which are related to body COM control. These are as summarized in [13]: human gait locomotion consists of six determinants, which are pelvic rotation, pelvic tilt, stance phase knee flexion, knee mechanisms, foot mechanisms, and lateral displacement of the pelvis. Three out of six of these determinants are pelvic related movements. Pelvic movement allows the swing of one leg to be initiated since the pelvic lateral displacement allows the centre of mass and body mass to transfer to the stance foot, thus allowing the other foot to swing [14]. hi addition, pelvic movement contributes to forward progression of the body and to vertical support of the trunk. As a result, control of pelvic motion is critical to achieve stable locomotion [15]. In general, the pelvis can be modelled as a body with six DoFs (degrees of freedom) consisting of translational and rotational movements in three axes as shown in Fig. 8(a).

In general, the pelvis can be modeled as a body with six DoFs consisting of translational and rotational movements in three axes as shown in Fig. 8 (a). The six DoF can be achieved with a simplified human leg structure shown in Fig. 8 (a): 3DoF at the hip joint, 1 DoF at the knee and 2 DoF at the ankle. To provide the necessary support, a pelvic assistive mechanism referred to as the balance stabilizer 1 , 2 is proposed and shown in Fig. 8 (b). By connecting the pelvis and the ground, the balance stabilizer mechanism 1, 2 can provide necessary assistive force through the interaction between the stabilizer 1 , 2 and the ground.

Based on a biomechanical study, an assistive gait device 100 with a stabilizer mechanism 1 , 2 was designed which can assist pelvic motion in order to help people to minimize their dependence on the crutch. As a result, a pelvic assistive mechanism or assistive gait device 100 with stabilizers 1, 2 is proposed and shown in Fig. 8(b).

An example of 6 DoF design is shown in Fig. 8c. The system consists of two SPU (spherical, prismatic, universal joint) chains for the balance stabilizer 1, 2 coupled with URS (universal, revolute, spherical joint) chains for the leg to which the system is coupled.

The number of degrees of freedom of the system or assistive gait device 100 can be computed as:

F = A(tf - / - l) + ./ = 6(8 -9 -l) + (2x 3 + 3 x3 + 3x l) = 6

The design with five degrees of freedom is chosen in this work as shown in Fig. 9(c). The reason of choosing the design of five degrees of freedom instead of six is explained below.

The detailed design of the exoskeleton 3 is shown in 9(d). As described above, the ankle joint 50 of the exoskeleton 3 has two degrees of freedom 21 , 22 and both of them are passive. The knee is an active single degree of freedom joint 7 which allows the knee flexion motion. The hip joints 5, 27 are designed to be two degree of freedom: hip flexion/extension 5 and hip rotation 27, and both of them are active while the hip joint for abduction/adduction motion is locked. The stabilizer 1 , 2 is designed to have three degrees of freedom with two rotational joint 13, 14 and one prismatic joint 15 which are all active.

In order to show that the balance stabilizer 1 , 2 coupled with the robotic exoskeleton 3 can achieve human-like walking, the six determinates of gait were used to evaluate the performance. The first three determinates are related to leg motion during walking which are shown in Fig. 10.

During normal walking, the pelvis also rotates from side to side about the vertical axis which can increase step length as shown in Fig. 1 1. The kinematic relationship between the hip rotation angle and incensement of step length can be expressed as:

AstepJength⁼Wt_rank*sin(0hip_rot) At the same time, pelvis rotation can be used to realize turning motion during walking. The kinematic relationship between the hip rotation angle and turning angle can be found as:

(^turning ~~ hip_rot-

The fifth determinant lateral pelvic motion is used for lateral body weight shifting. The realization of lateral pelvic motion is shown in Fig. 12 and the lateral displacement can be calculated as:

The last determination is pelvic tilt which is used in order to save the energy of walking. However, since the additional actuator at the hip abduction/adduction joint will make the design and control more complex, this determinate is not realized in the present system. And by making this joint locked, there will be an additional foot-ground clearance created when the system has the pelvic lateral displacement. This feature will help to reduce the requirement of large knee bending motion which is used widely in exoskeleton devices to increase the foot-ground clearance.

At the same time, the stabilizer 1 , 2 has the potential ability of assisting the human user in sit to stand motion as shown in Fig. 13.

Apart from mechanism design, the points of the support of the stabilizer mechanism 1 , 2 also affect the kinematic properties of the robot or assistive gait device 100 such as workspace, i.e., the amount of space taken by the person in the exoskeleton 3 and the balance stabilizers 1 , 2. Thus, it is worth exploring the workspace of the assistive gait device 100 under different support configurations. A numerical method was used to establish the workspace of the assistive gait device 100 and the results are shown in Fig. 14. There are four kinds of supporting configurations and the different configurations are based on the observation of crutch support pattern when people use crutches to walk. All the four workspaces can successfully cover the hip trajectory of human walking shown in Fig. 14(e). The support of the configuration shown in Fig. 14(b) was used in the present work.

For testing the performance of such a design approach, a three DoFs mechanism was built up (motion constrained in sagittal plane). The system consisted of two parts: robotic exoskeleton 3 and balance stabilizer mechanism 1 , 2 as shown in Fig. 15. The robotic exoskeleton 3 consisted of two powered anthropomorphic legs: one degree of freedom at the hip (active); one degree of freedom at the knee (active); one degree of freedom at the ankle (passive). At the hip and knee joints, there were two DC motors at each leg powering up the joints by using a maxon flat motor EC 90 together with a harmonic drive mechanism (ratio: 100:1). The balance stabilizer 1 , 2 design combined a single prismatic leg with a rotational joint which were all actively controlled.

The controller and sensory system design is shown in Fig. 16. NI motion control card PXI- 7350 was used to control the system or assistive gait device 100. One EVIU (Inertial Measurement Unit) is provided at the back of the support to provide the trunk angle. At each joint, a potential meter 32 is installed to provide the absolute joint angle feedback. Meanwhile, limit switches and flexi-force sensors are installed under the stabilizers 1 , 2 in order to detect ground contract. The controller and sensory system comprising an input device is configured to coordinate and actuate independent movement of the exoskeleton and two stabilizers. The controller and sensory system is configured to be activated by the human user wearing the assistive gait device 100 by means of the input device, such as push buttons or a touch screen.

The control algorithm is presented in Fig. 17. First, variables of the control algorithm are initialized (171). When the user activates start of motion (172) via the input device, the algorithm calculates the rotation angle and length of the stabilizers 1 , 2 based on an original ground support point (173). The stabiliser motors 30 controlling the stabilizer joints 13, 14, 15 are then run or activated (174) to adjust the stabilizers 1 , 2 by retracting (i.e. shortening) the stabilizers 1, 2 via the prismatic joints 15 and rotating the joints 13, 14 to swing the stabilizers 1 , 2 forward, and then extending (i.e. lengthening) the stabilizers 1, 2 via the prismatic joints 15 to contact a new ground support point that is forward of the original ground support point. The algorithm then gets the user's intention to start walking (175) when the user initiates start of a walking sequence via the input device. Upon initiation of the walking sequence by the user, the algorithm calculates trajectories and angles of motion for the leg joints 5, 7, 27 and the stabilizers 1 , 2 (176). The algorithm then plots the trajectories and runs the motors 30 of the exoskeleton 3 and the stabilizers 1, 2 according to the planned motions (177), which concludes one walking cycle (178). By configuring the control algorithm to be operable by the human user by means of buttons or any other appropriate input device such as a touch screen or joystick, the assistive gait device 100 of the present invention thus allows persons who do not have sufficient upper body strength or control to use conventional crutches to be able to walk when using the present assistive gait device 100, since the walking sequence by the stabilizers 1 , 2 and exoskeleton 3 can be initiated by the person simply by activating the input device.

Apart from mechanical structure, the gait pattern planning algorithm is also an essential part in achieving a stable locomotion. The flowchart of gait pattern generation algorithm is presented in Fig. 18 and described in greater detail below.

At first, gait pattern generation method which is based on human walking was used. Polynomial function is used as the basis function to represent the trajectories of joints and the via point hip and swing ankle joints are set based on real human walking pattern shown in Fig. 19.

Since the centre of mass of human body is located near the hip, movements of the hip would affect the system stability significantly. Thus, to plan hip trajectories with minimum jerk, a polynomial equation (shown below) of degree 5 was chosen to interpolate the initial and final coordinates of the hip. To solve for the six unknown coefficients, we considered the coordinates of the hip at the start and end to be fixed, and also the velocity and acceleration at the start and end of each step to be zero. y_hip ¹ ¾ = at⁵ +bt⁴ + ct³ + dt² + et + f

In planning swing leg trajectories, the positions of ankle joint were chosen to follow humanlike swing ankle motion as shown in Fig. 19. Five critical positions: starting, right after swing, maximum height, before landing and final were selected as the critical points. Therefore, a polynomial equation (shown below) of degree 6 was used. To solve for the seven unknown coefficients, we considered five critical coordinates to be fixed and velocity at the start and end of each step to be zero. The hip and knee motion is plotted in Fig. 20 and the detailed information of this gait pattern generation method can be found in [17].

J z_imkk = at⁶ +btⁱ +ct⁴ +dt' + et² +fi + g However, gait planning based on human motion have limitations such as the kinematics and dynamics of the robot are different from person to person. Thus, the trajectory may not be stable. As a result, it is necessary to re-optimize the trajectory by considering the stability of the trajectory. Sugihara et al. [18] introduced the concept COG (centre of gravity) Jacobian, which relates whole-body motion to COG motion and this strategy is employed in the present work to modify the ZMP (zero moment point) trajectory. According to [19], the zero moment point can be ex ressed as:

}'ZMP ^rCM -z ^rCM

g + ^rcM ^m(g + r_CM) where ^RCM is the position of the centre of mass and L_OOM is the angular momentum about the body's centre of mass. The acceleration of COM can be approximated as: rCM "cM ^{+ r}CM

rCM ~

T²

According to [20], the ZMP trajectory can be expressed into a compact matrix form as:

Then, the CoM trajectory can be calculated as:

^XC = ;¹(x_ZMP - b

The next step is to establish the relationship between the change of the center of mass and joints motion by using the CoG jacobian which can be expressed as:

° a?

In addition to the stance feet, the stabilizer is also fixed to the ground and can be added as constraints in the form:

where J_SL e D ^3χΛ' and J_{SR e} r ^,XN are the Jacobian of the left and right stabilizer in the world coordinate. By combining CoG Jacobian together with additional constraints, the relationship between the change in ZMP and the change of the joint angle can be expressed as:

The result of the algorithm is shown in Fig. 21. It can be shown that the ZMP trajectory is successfully shifted inside the support polygon.

The whole motion sequence is shown in Fig. 22. The robotic exoskeleton 3 with balance stabilizer 1 , 2 gait cycle begins with the double support phase, where both feet are in contact with the ground. During this time, the stabilizers 1 , 2 would reposition themselves to the new ground support point to prepare for the next walking motion. During the single support phase the swing leg would step forward and the cycle is again repeated for the other leg.

To test the concept of patients with poor upper limb being able to use the device 100, in a previous trial, the system had successfully assisted a tetraplegic subject to achieve 5m walking during clinical trial of the device 100. During the walking process, the stabilizer mechanism 1 , 2 proves its robustness of supporting the subject's body weight and provides stability during propulsion of the mechanical exoskeleton 3. At the same time, the motion planning algorithm is able to generate a stable gait and a smooth synchronization motion between the robotic exoskeleton leg 3 and the balance stabilizer 1 , 2 mechanism. Finally, the built-in sensing mechanisms were able to provide accurate system states feedback to allow the system to respond and achieve a stable walking motion. In the initial phase of trial of this device 100, there were many assistants to ensure the subject's safety.

Commercialization opportunities and/or the potential public health impact of the innovation

The impact of chronic disease lies not in death but chronic long term disability. Currently, stroke is the most common cause of severe chronic disability in Singapore. The risk of stroke increases exponentially with aging [26]. More than two-thirds of stroke patients have significant weakness of the limbs, impairing their capacity to do activities of daily living such as eating, dressing and walking [27-30]. More than 50% of patients were unable to walk immediately post stroke and 12% required assistance to walk [30]. In addition, other neurologically impaired persons including spinal cord injury persons with tetraplegia (weakness of both hands and legs), any person suffering from neuromuscular disease in which there is weakness of the limbs, etc., will be beneficiaries of such a device. The proposed device 100 will have tremendous implications in rehabilitation and assistive field which include a chance for dependent persons with poor lower limb and trunk function to achieve upright standing and walking and mobilization with reduced assistance from the caregiver and therapist (e.g. from three persons assisting to one person assisting), both during therapy sessions and at home. Implications would be that more dependent persons will be able to achieve walking and upright tolerance with reduced assistance from the caregiver and with frequent use, reduce the risk of chronic disease and co-morbidities resulting from immobility.

Whilst there has been described in the foregoing description exemplary embodiments of the present invention, it will be understood by those skilled in the technology concerned that many variations in details of design, construction and/or operation may be made without departing from the present invention. For example,

References

[1] K. A. Strausser and H. Kazerooni, "The Development and Testing of a Human Machine Interface for a Mobile Medical Exoskeleton," in 2011 EEEE/RSJ International

Conference on Intelligent Robots and Systems (IROS 201 1), 25-30 Sept. 201 1 ,

Piscataway, NJ, USA, 201 1 , pp. 4911-16.

[2] M. Talaty, A. Esquenazi, and J. E. Briceno, "Differentiating ability in users of the

ReWalkTM powered exoskeleton: An analysis of walking kinematics," in 2013 ΓΕΕΕ 13th International Conference on Rehabilitation Robotics, ICORR 2013, June 24, 2013 - June 26, 2013, Seattle, WA, United states, 2013, p. NSF Engineering Research Center for; Sensorimotor Neural Engineering (CSNE); at the University of Washington; IEEE EMB; IEEE Robotics and Automation Society.

[3] H. Kawamoto and Y. Sankai, "Power assist method based on Phase Sequence and muscle force condition for HAL," Advanced Robotics, vol. 19, pp. 717-734, 2005. [4] P. D. Neuhaus, J. H. Noorden, T. J. Craig, T. Torres, J. Kirschbaum, and J. E. Pratt, "Design and evaluation of Mina: A robotic orthosis for paraplegics," in Rehab Week Zurich 201 1 - 201 1 IEEE International Conference on Rehabilitation Robotics, ICORR 201 1 , June 27, 2011 - July 1 , 201 1 , Zurich, Switzerland, 201 1.

[5] R. J. Farris, H. A. Quintero, and M. Goldfarb, "Performance evaluation of a lower limb exoskeleton for stair ascent and descent with Paraplegia," in 2012 34th Annual

International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 28 Aug.-l Sept. 2012, Piscataway, NJ, USA, 2012, pp. 1908-11.

[6] (2014). Rex exoskeleton. Available: http://www.rexbionics.com/

[7] W. Letian, W. Shiqian, E. H. F. van Asseldonk, and H. van der ooij, "Actively

controlled lateral gait assistance in a lower limb exoskeleton," in Intelligent Robots and Systems (IROS), 2013 IEEE/RSJ International Conference on, 2013, pp. 965-970.

[8] H. Sang-Ho, T. Hayashi, A. Yagi, T. Noda, and J. Morimoto, "Design of hybrid drive exoskeleton robot XoR2," in Intelligent Robots and Systems (IROS), 2013 IEEE/RSJ International Conference on, 2013, pp. 4642-4648.

[9] (2014). Evryon exoskeleton. Available: http://www.evryon.eu/index.html

[10] M. Peshkin, D. A. Brown, J. J. Santos-Munne, A. Makhlin, E. Lewis, J. E. Colgate, et al., "KineAssist: a robotic overground gait and balance training device," in 2005 IEEE 9th International Conference on Rehabilitation Robotics, 28 June-1 July 2005, Piscataway, NJ, USA, 2005, pp. 241-6.

[1 1 ] M. Bouri, Y. Stauffer, C. Schmitt, Y. Allemand, S. Gnemmi, R. Clavel, et al., "The

WalkTrainer: a robotic system for walking rehabilitation," in IEEE International

Conference on Robotics and Biomimetics - ROBIO2006, 17-20 Dec. 2006, Piscataway, NJ, USA, 2006, pp. 1616-21. [12] W. Ping, . H. Low, and A. H. McGregor, "A subject-based motion generation model with adjustable walking pattern for a gait robotic trainer: NaTUre-gaits," in 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2011), 25- 30 Sept. 2011 , Piscataway, NJ, USA, 2011 , pp. 1743-8.

[13] J. B. d. M. Saunders, V. T. Inman, and H. D. Eberhart, "THE MAJOR

DETERMINANTS IN NORMAL AND PATHOLOGICAL GAIT," The Journal of Bone & Joint Surgery, vol. 35, pp. 543-558, 1953.

[14] J. Perry and J. M. Burnfield, Gait analysis: normal and pathological function, 2nd ed.

Thorofare, NJ: SLACK, 2010.

[15] R. R. Neptune, S. A. Kautz, and F. E. Zajac, "Contributions of the individual ankle

plantar flexors to support, forward progression and swing initiation during walking," Journal of biomechanics, vol. 34, pp. 1387-1398, 2001.

[16] S. A. Joshi and L.-W. Tsai, "Jacobian Analysis of Limited-DOF Parallel Manipulators," Journal of Mechanical Design, vol. 124, pp. 254-258, 2002.

[17] L. Li, K. H. Hoon, and K. H. Low, "Gait Analysis and Modeling of Robotic Orthosis with Balance Stabilizer," presented at the 3rd IFToMM International Symposium on Robotics and Mechatronics, Singapore, 2013.

[18] T. Sugihara and Y. Nakamura, "Whole-body cooperative balancing of humanoid robot using COG Jacobian," in 2002 IEEE/RSJ International Conference on Intelligent Robots and Systems, September 30, 2002 - October 4, 2002, Lausanne, Switzerland, 2002, pp. 2575-2580.

[19] M. B. Popovic, A. Goswami, and H. Herr, "Ground reference points in legged

locomotion: definitions, biological trajectories and control implications," International Journal of Robotics Research, vol. 24, pp. 1013-32, 12/ 2005. [20] N. Naksuk, C. S. G. Lee, and S. Rietdyk, "Whole-body human-to-humanoid motion transfer," in Humanoid Robots, 2005 5th IEEE-RAS International Conference on, 2005, pp. 104-109.

[21] Schiffman, J.; Gregorczyk, K.; Bensel, C; Hasselquist, L. & Obusek, J. "The effects of a lower body exoskeleton load carriage assistive device on limits of stability and postural sway", Ergonomics, Taylor & Francis, 2008, 51 , 1515-1529

[22] URL: http://www.eksobionics.com/. Last visited: 20/SEP/2014

[23] URL: http://rewalk.com/ . Last visited: 20/SEP/2014

[24] URL: http://www.cyberdyne.jp/english/products/HAL/. Last visited: 20/SEP/2014

[25] URL: http://www.rexbionics.com/. Last visited: 20/SEP/2014

[26] Venketasubramanian N, Chen CL. Burden of stroke in Singapore. Int J Stroke. 2008 Feb;3(l):51-4.

[27] Venketasubramanian N, Tan LC, Sahadevan S, Chin JJ, Krishnamoorthy ES, Hong CY, Saw SM. Prevalence of stroke among Chinese, Malay, and Indian Singaporeans: a community-based tri-racial cross-sectional survey. Stroke. 2005 Mar;36(3):551-6

[28] Ministry of Health, Statistics, Health Facts Singapore; Top 10 Causes of Hospitalization.

Available at: http://www.moh.gov.sg. Accessed 20th Sep 2014

[29] Wolfe CD. The impact of Stroke. Br Med Bull 2000;56:275-86.

[30] Jorgensen HS, Nakayama H, Raaschou HO, Olsen TS. Recovery of walking function in stroke patients: the Copenhagen Stroke Study. Arch Phys Med Rehabil 1995;76:27- 32 (a).

Claims

1. An assistive gait device comprising:

a motorized exoskeleton configured to be coupled to a human person and configured to independently move each leg of the human person;

a first motorized stabilizer provided on the left side of the human person and a second motorized stabilizer provided on the right side of the human person, the two stabilizers connected to the exoskeleton, each of the two stabilizers configured to be moved independently relative to the exoskeleton and to be able to selectively contact the ground to stabilize the human person; and

a controller and sensory system comprising an input device and configured to coordinate and actuate independent movement of the exoskeleton and two stabilizers to automatically effect walking movement of the human person upon the human person initiating a walking sequence by activating the input device.

2. The assistive gait device of claim 1 , wherein the exoskeleton comprises, for each leg, a shank link connected to a thigh link via a motorized knee joint configured to allow knee flexion and extension, the shank link connected to an exoskeleton trunk via a motorized first hip joint configured to allow rotation of the thigh link about a mediolateral axis, and a motorized second hip joint configured to allow rotation of the thigh link about a vertical axis.

3. The assistive gait device of claim 1 or claim 2, wherein the two stabilizers each comprise a motorized first rotational joint configured to allow rotation of the stabilizer about the mediolateral axis, a motorized second rotational joint configured to allow rotation of the stabilizer about an anteriorposterior axis, and a motorized prismatic joint configured to allow the stabilizer to lengthen and shorten.

4. The assistive gait device of any preceding claim, wherein the input devices comprises at least one of: a push button, a touchscreen and a joystick.

5. The assistive gait device of any preceding claim, wherein the controller comprises a control algorithm in which a stable trajectory is generated based on a Zero Moment Point criteria and a whole body Centre Of Gravity Jacobian method is used to filter unbalanced motion.

6. The assistive gait device of any preceding claim, wherein the walking sequence is preceded by forward movement of the two stabilizers, the forward movement of the two stabilizers being activated as start of motion by the human person via the input device.

7. The assistive gait device of claim 6, wherein the walking sequence comprises the exoskeleton swinging forward a foot of the human person and placing the foot on the ground.