Design of a haptic arm exoskeleton for training and rehabilitation

A high-quality haptic interface is typically characterized by low apparent inertia and damping, high structural stiffness, minimal backlash, and absence of mechanical singularities in the workspace. In addition to these specifications, exoskeleton haptic interface design involves consideration of space and weight limitations, workspace requirements, and the kinematic constraints placed on the device by the human arm. These constraints impose conflicting design requirements on the engineer attempting to design an arm exoskeleton. In this paper, the authors present a detailed review of the requirements and constraints that are involved in the design of a high-quality haptic arm exoskeleton. In this context, the design of a five-degree-of-freedom haptic arm exoskeleton for training and rehabilitation in virtual environments is presented. The device is capable of providing kinesthetic feedback to the joints of the lower arm and wrist of the operator, and will be used in future work for robot-assisted rehabilitation and training. Motivation for such applications is based on findings that show robot-assisted physical therapy aids in the rehabilitation process following neurological injuries. As a training tool, the device provides a means to implement flexible, repeatable, and safe training methodologies.     

Redundancy resolution of the human arm and an upper limb exoskeleton

The human arm has 7 degrees of freedom (DOF) while only 6 DOF are required to position the wrist and orient the palm. Thus, the inverse kinematics of an human arm has a nonunique solution. Resolving this redundancy becomes critical as the human interacts with a wearable robot and the inverse kinematics solution of these two coupled systems must be identical to guarantee an seamless integration. The redundancy of the arm can be formulated by defining the swivel angle, the rotation angle of the plane defined by the upper and lower arm around a virtual axis that connects the shoulder and wrist joints. Analyzing reaching tasks recorded with a motion capture system indicates that the swivel angle is selected such that when the elbow joint is flexed, the palm points to the head. Based on these experimental results, a new criterion is formed to resolve the human arm redundancy. This criterion was implemented into the control algorithm of an upper limb 7-DOF wearable robot. Experimental results indicate that by using the proposed redundancy resolution criterion, the error between the predicted and the actual swivel angle adopted by the motor control system is less then 5°.     

Event driven sliding mode control of a lower limb exoskeleton based on a continuous neural network electromyographic signal classifier

This study presents an event driven automatic controller to regulate the movement of a mobile lower limb active orthosis (LLAO) triggered with the information obtained from electromyographic (EMG) signals, which are captured from the user’s triceps and biceps muscles. The proposed controller has an output feedback realization including a velocity estimator algorithm based on a high order sliding mode observer. The output feedback controller implements a class of decentralized super-twisting algorithm. The controller must enforce the movement of the orthosis articulations following some defined reference trajectories. This strategy realizes a time-window dependent event driven controller for the active orthosis. The controller selects among four different routines to be executed by a patient. A differential neural network classifies the different patterns of muscle movements. This classifier succeeds in defining the correct EMG class in a 95% of the tested signals. This work senses the EMG signals from the biceps and triceps, considering a possible injury in the patient to be obtained from the quadriceps. Therefore, four upper limb routines are established to generate the corresponding classes and the four different main therapies for the LLAO. A fully instrumented and self-designed orthosis is constructed to evaluate the proposed controller including three rotational joints per leg and a mobile robot to execute translation movements.     

Adaptive fuzzy impedance control of exoskeleton robots with electromyography-based convolutional neural networks for human intended trajectory estimation

Among various uses of exoskeleton robots, the rehabilitation of stroke patients is a more recent application. There is, however, considerable environmental uncertainty in such systems including uncertain robot dynamics, unwanted user reflexes, and, most importantly, uncertainty in user intended trajectory. Hence, it is challenging to develop transparent, stable, and wide-scale exoskeleton robots for rehabilitation. This paper proposes an adaptive fuzzy impedance controller (AFIC) and a convolutional neural network (CNN) which uses electromyographic (EMG) signals for early detection of human intention and better integration with a lower limb exoskeleton robot. Specifically, the primary purpose of the AFIC is to manage the mechanical interaction between human, robot, and environment and to deal with uncertainties in internal control parameters. CNN uses EMG signals, inertial measurement units, foot force sensing resistors, joint angular sensors, and load cells to deal with signal uncertainties and noise through automatic feature processing in order to detect user’s desired joint angles with high accuracy. EMG is particularly effective here since it reflects the human intention to move faster than the other mechanical sensors. In the experimental procedure, signals were sampled at 500 Hz as two healthy individuals walked normally at 0.3, 0.4, 0.5, and 0.6 m/s for eight minutes while wearing a robot with zero inertia. Approximately 70% of the data is used for training and 30% for testing the network. The estimated angle from the trained network is then used as the desired angle in the AFIC loop, which controls the robot online as the desired trajectory. Pearson correlation coefficient and normalized root mean square error are computed to evaluate the accuracy and robustness of the proposed angle estimation with CNN and AFIC algorithms. Experimental results show that the proposed approach successfully obtains the torque of the robot joints despite uncertainties in changing the walking speed.     

Personalized voice activated grasping system for a robotic exoskeleton glove

This paper proposes a novel human machine interface (HMI) and electronics system design to control a rehabilitation robotic exoskeleton glove. Such system can be activated with the user’s voice, take voice commands as input, recognize the command and perform biometric authentication in real-time with limited computing power, and execute the command on the exoskeleton. The electronics design is a stand-alone plug-and-play modulated design independent of the exoskeleton design. This personalized voice activated grasping system achieves better wearability, lower latency, and improved security than any existing exoskeleton glove control system.     

Real-time call admission control for packet-switched networking by cellular neural networks

In this paper, novel call admission control (CAC) algorithms are developed based on cellular neural networks. These algorithms can achieve high network utilization by performing CAC in real-time, which is imperative in supporting quality of service (QoS) communication over packet-switched networks. The proposed solutions are of basic significance in access technology where a subscriber population (connected to the Internet via an access module) needs to receive services. In this case, QoS can only be preserved by admitting those user configurations which will not overload the access module. The paper treats CAC as a set separation problem where the separation surface is approximated based on a training set. This casts CAC as an image processing task in which a complex admission pattern is to be recognized from a couple of initial points belonging to the training set. Since CNNs can implement any propagation models to explore complex patterns, CAC can then be carried out by a CNN. The major challenge is to find the proper template matrix which yields high network utilization. On the other hand, the proposed method is also capable of handling three-dimensional separation surfaces, as in a typical access scenario there are three traffic classes (e.g., two type of Internet access and one voice over asymmetric digital subscriber line.     

Development of a novel compact hydraulic power unit for the exoskeleton robot

Exoskeleton robots require powerful and lightweight power supplies. Because of its high power-to-mass ratio and fast response, hydraulic systems can meet the requirement for locomotion robots. In this paper, a novel compact hydraulic power unit (CHPU) is proposed. A two-stroke IC engine, with a rated power of 2.4 kW at 13,000 rpm, is used as the prime mover. The engine drives a high speed (10,000 rpm) piston pump to allow the engine to operate at high power. A spring loaded reservoir has been developed to prevent the pump intake from cavitation and contamination. The payload flow rate is indirectly estimated using the displacements of the actuator. A Ragone plot analysis shows that the CHPU can maintain a high specific power over a long duration. A dynamic model for the CHPU has been developed based upon simplified engine operating characteristics and a set of experimentally identified parameters. A prototype of the CHPU has been constructed with a rated power of 1.45 kW and a weight of 16.6 kg. Experimental testing of the prototype confirms the dynamic model and the output capacity of the CHPU.     

Real-time discrimination of ventricular tachyarrhythmia withFourier-transform neural network

The authors have developed a method to discriminate life-threatening ventricular arrhythmias by observing the QRS complex of the electrocardiogram (ECG) in each heartbeat. Changes in QRS complexes due to rhythm origination and conduction path were observed with the Fourier transform, and three kinds of rhythms were discriminated by a neural network. In this paper, the potential of the authors' method for clinical uses and real-time detection was examined using human surface ECGs and intracardiac electrograms (EGMs). The method achieved high sensitivity and specificity (>0.98) in discrimination of supraventricular rhythms from ventricular ones. The authors also present a hardware implementation of the algorithm on a commercial single-chip CPU     

Intelligent switching control of a pneumatic muscle robot arm using learning vector quantization neural network

Pneumatic cylinders are one of the low-cost actuation sources used in industrial and prosthetic application, since they have a high power/weight ratio, high-tension force and long durability. However, problems with the control, oscillatory motion and compliance of pneumatic systems have prevented their widespread use in advanced robotics. To overcome these shortcomings, a number of newer pneumatic actuators have been developed, such as the McKibben Muscle, Rubber Actuator and Pneumatic Artificial Muscle (PAM) Manipulators. In this paper, the solution for position control of a robot arm with slow motion driven by two pneumatic artificial muscles is presented. However, some limitations still exist, such as a deterioration of the performance of transient response due to the changes in the external load. To overcome this problem, a switching algorithm of the control parameter using a learning vector quantization neural network (LVQNN) is proposed in this paper. The LVQNN estimates the external load of the pneumatic artificial muscle manipulator. The effectiveness of the proposed control algorithm is demonstrated through experiments with different external working loads.     

Design and locomotion control of a hydraulic lower extremity exoskeleton for mobility augmentation

In this paper, a system design and three locomotion control algorithms are proposed for a hydraulic lower extremity exoskeleton to enhance mobility and reduce muscle fatigue caused by backpack loads. The range of motion of the exoskeleton and the capacity of the hydraulic power unit, which generates the hydraulic flow and pressure, are determined by analysing human walking data obtained using a motion capture device and force plates. For movement comfort, the mechanical structure and the joints of the exoskeleton are designed such that the motion of the wearer coincides with that of the exoskeleton. In addition, locomotion control algorithms for stable normal walking are described; these algorithms enable dual-mode control and transition control. Dual-mode control is comprised of an active mode in the stance phase and a passive mode in the swing phase. In the active mode, the exoskeleton is controlled to track the motion of the wearer, and in the passive mode, its active joints work as passive joints by blocking the hydraulic power supply from the hydraulic power unit. Transition control, which consists of a torque-shaping method and a pre-transition algorithm, is adopted to improve locomotion responses during gait phase transition. Finally, to verify the effectiveness of the locomotion control algorithms and the developed hydraulic lower limb exoskeleton, walking experiments are performed on a treadmill, at a speed of 4 km/h, while carrying 45 kg backpack loads. The assistance effect of the exoskeleton is also validated by comparing the electromyography (EMG) signals of four selected muscles, with and without the exoskeleton, for single stance and level walking while carrying the same 45 kg backpack loads.     

Real-time control of reactive ion etching using neural networks

This paper explores the use of neural networks for real-time, model-based feedback control of reactive ion etching (RIE). This objective is accomplished in part by constructing a predictive model for the system that can be approximately inverted to achieve the desired control. An indirect adaptive control (IAC) strategy is pursued. The IAC structure includes a controller and plant emulator, which are implemented as two separate back-propagation neural networks. These components facilitate nonlinear system identification and control, respectively. The neural network controller is applied to controlling the etch rate of a GaAs/AlGaAs metal-semiconductor-metal (MSM) structure in a BCl₃/Cl₂ plasma using a Plasma Therm 700 SLR series RIE system. Results indicate that in the presence of disturbances and shifts in RIE performance, the IAC neural controller is able to adjust the recipe to match the etch rate to that of the target value in less than 5 s. These results are shown to be superior to those of a more conventional control scheme using the linear quadratic Gaussian method with loop-transfer recovery, which is based on a linearized transfer function model of the RIE system     

Modeling and control of a curved pneumatic muscle actuator for wearable elbow exoskeleton

A novel curved pneumatic muscle based rotary actuator for the wearable elbow exoskeleton with joint torque control is proposed. Compared to the general utilization of the pneumatic muscle actuator (PMA) in a rotary joint, this novel structure weakens coupling relationship between the output torque/force and contacting displacement of the PMA so that it can be easily utilized in the tele-robotics with torque/force-feedback or the exciting application in rehabilitation for a wide range. By referred to two physical models, namely beam model and membrane model, the mechanics properties of this mechanical structure is analyzed. In addition a hybrid fuzzy controller composed of bang–bang controller and fuzzy controller is employed for output torque control with high accuracy as well as fast response. In a series of experiments, the actuator exhibits both good static and dynamic performances that well validated the models and control strategy.     

An original mechatronic design of a kinaesthetic hand exoskeleton for virtual reality-based applications

Within the new industrial era, the interaction between humans and virtual reality is spreading across our lives. The development of exoskeleton designed to enhance the immersivity of virtual reality environments has a potentially considerable social impact and arises as a hot research topic. The presented work dwells well with the subject by describing the mechatronic design process of a kinaesthetic hand exoskeleton system meant to reproduce proprioceptive stimuli coming from the interaction with a virtual reality. The presented prototype is a modular device, equipped with force and pose sensors, and driven by a Bowden-cable-based remote actuation system. Unlike similar devices, the proposed exoskeleton is specifically thought for VR interaction and is designed to be reversible while exerting up to 15 N per finger. For a more accurate rendering of kinetostatic finger stimuli, a procedure for reconstructing HMI force as a function of measured force and position signals by employing a system’s kinematic and dynamic model is presented, detailed, and followed by some preliminary tests. The results showed that the model can trace forces back to the end-effector with a percentage error below 15%.     

Development a multi-loop modulation method on the servo drives for lower limb rehabilitation exoskeleton

This paper proposed a multi-loop modulation method (MMM; 3M) on the servo drives applied for the lower limb rehabilitation exoskeleton (LLRE). This proposed 3M included the current, speed and position 3-loop with the auto-tuning, inertia identifier and external load torque observer. The controller gains by searching the optimal bandwidth and by identifying the inertia of the LLRE system were derived from the 3-loop auto-tuning. The controller gains could be instantly adjusted according to the system's oscillation during the LLRE gait motion. In particular, a concept of the phase margin (PM) was introduced in the calculation of the controller gain to ensure system stability. The proposed method could avoid system turbulence due to the excessive bandwidth. On the other hand, a torque observer and inertia identifier were adopted to compensate the external load. Compared the traditional parameter design method and traditional auto-tuning method with this proposed 3M in terms of tracking error and root mean square error (RMSE), the result showed that this proposed 3M provided better response and stability for gait tracking. Also, this proposed 3M could be adapted to variations in different loads.     

Real-time photoplethysmographic heart rate measurement using deep neural network filters

Photoplethysmography (PPG) is a noninvasive technique that can be used to conveniently measure heart rate (HR) and thus obtain relevant health-related information. However, developing an automated PPG system is difficult, because its waveforms are susceptible to motion artifacts and between-patient variation, making its interpretation difficult. We use deep neural network (DNN) filters to mimic the cognitive ability of a human expert who can distinguish the features of PPG altered by noise from various sources. Systolic (S), onset (O), and first derivative peaks (W) are recognized by three different DNN filters. In addition, the boundaries of uninformative regions caused by artifacts are identified by two different filters. The algorithm reliably derives the HR and presents recognition scores for the S, O, and W peaks and artifacts with only a 0.7-s delay. In the evaluation using data from 11 patients obtained from PhysioNet, the algorithm yields 8643 (86.12%) reliable HR measurements from a total of 10 036 heartbeats, including some with uninformative data resulting from arrhythmias and artifacts.     

Experimental characterization of the T-FLEX ankle exoskeleton for gait assistance

Designing robotic devices to assist or emulate the ankle is challenging due to the joint’s complexity and the fundamental role in walking. T-FLEX is an ankle exoskeleton based on vsa for rehabilitation and assistance of people with ankle dysfunctions. This device has presented promising motor recovery results for a stroke patient during a rehabilitation program. However, human walking applications require an electromechanical characterization to measure the device’s capabilities and determine the suitable configuration that responds to this complex task. This work presents T-FLEX’s experimental characterization carried out in a test bench structure. The results showed alterations in system times and actuators’ bandwidth because of the tendons’ force levels. Furthermore, this study determined the most appropriate T-FLEX configuration to obtain the best performance. Thus, this work also presents a preliminary validation under that configuration on a healthy subject in gait assistance to assess the device’s response in different velocities and measure the effects on the user. In conclusion, T-FLEX can assist the human gait for gait cycle duration greater than 0.74 s providing torque on the ankle of up to 12 Nm in propulsion and 20 Nm in dorsiflexion. Nevertheless, it should include an adaptable stage in the control architecture to counteract the stabilization time for providing the maximum torque at the right time.     

An asymmetric cable-driven mechanism for force control of exoskeleton systems

In this paper, an asymmetric cable-driven mechanism is proposed for accurate force control of exoskeleton systems with a compact structure. Inspired by the fact that the required forces in human motions are not symmetric in many cases, a spring-actuator type cable-drive mechanism is adopted, which enables a compact cable routing structure. The drive pulley is connected with the exoskeleton frame through a rotary series elastic mechanism to transmit the desired force to the human user. High performance in force control is achieved by advanced control algorithms, which combine a proportional and differential (PD) controller optimized by a linear quadratic (LQ) method with a disturbance observer (DOB) and a zero phase error tracking (ZPET) feedforward filter. The proposed system was tested for the elbow joint. Experimental results confirmed that the proposed system could generate and deliver accurate force to the human user even with external disturbances and modeling uncertainties introduced by human motions.     

Design and control of a wearable and force-controllable hand exoskeleton system

In this paper, a wearable and force-controllable hand exoskeleton system is proposed. In order to apply force feedback to the fingertip while allowing natural finger motions, the exoskeleton linkage structure with three degrees of freedom (DOFs) was designed, which was inspired by the muscular skeletal structure of the finger. Kinematic performance of the proposed linkage structure was verified by comparing with functional range of motion (ROM) which is required for activities in daily living. As an actuating system, a series elastic actuator (SEA) mechanism, which consisted of a small linear motor, a manually designed motor driver, a spring and potentiometers, was applied. Friction of the motor was identified and compensated to obtain a linearized model of the actuating system. Using a linear quadratic (LQ) tuned proportional-derivative (PD) controller and a disturbance observer (DOB), the proposed actuator module could generate the desired force accurately even with arbitrary finger movement. The performance of force transmission through linkage structure was verified by simulation and experiments. The proposed exoskeleton structure, actuator modules and control algorithms were integrated as a wearable and force-controllable hand exoskeleton system that could deliver force to the fingertips for flexion/extension motions.     

Real-time license plate detection and recognition using deep convolutional neural networks

Automatic License Plate Recognition (ALPR) is an important task with many applications in Intelligent Transportation and Surveillance systems. This work presents an end-to-end ALPR method based on a hierarchical Convolutional Neural Network (CNN). The core idea of the proposed method is to identify the vehicle and the license plate region using two passes on the same CNN, and then to recognize the characters using a second CNN. The recognition CNN massively explores the use of synthetic and augmented data to cope with limited training datasets, and our results show that the augmentation process significantly increases the recognition rate. In addition, we present a novel temporal coherence technique to better stabilize the OCR output in videos. Our method was tested with publicly available datasets containing Brazilian and European license plates, achieving accuracy rates better than competitive academic methods and a commercial system.     

PV powered bus shelters for the UK

Canada’s Carmanah Technologies has received a contract worth C$1.6 million ($710 000) from Trueform Group in the UK to provide solar-powered bus shelter lighting systems.This is a short news story only. Visit www.re-focus.net for the latest renewable energy industry news.