Keywords: Technologies for neurodevelopmental disorders, Design and development in rehabilitation robotics, Socially interactive robotics - design and development
Abstract: Early detection of neurodevelopmental disorders in infants is critical for early intervention to improve their long-term function. Integrating natural play with quantitative measurements of developmental milestones may help to quickly and efficiently identify infants at-risk for developmental delays. Ailu is a sensorized toy designed to elicit and measure natural infant play interactions. Ailu is part of the Play and Neuro Development Assessment (PANDA) gym, whose purpose is to serve as a universal and quantitative screening tool for detection of delays. This case study describes design considerations made developing Ailu and evaluates Ailu’s potential in upper limb, lower limb, and parent-guided testing with a 3-month old infant. Ailu can encourage reaching, kicking, and grasping, and will be tested for distinguishing typical and atypical development with further infant trials.

Keywords: Technologies for neurodegenerative disorders, New technologies and methodologies in human movement analysis, Biomechanics and robotics in physical rehabilitation
Abstract: Upper limb (UL) compensation is a common strategy of patients with a high spinal cord injury (SCI), i.e., tetraplegic patients, to perform activities of daily living (ADLs) despite their sensorimotor deficits. Currently, an objective and sensitive tool to assess UL compensation, which is applicable in the clinical routine and in the daily life of patients, is missing. In this work, we propose a metric to quantify this compensation using a single inertial measurement unit (IMU). The spread of forearm pitch angles of an IMU attached to the wrist of 17 SCI patients and 18 healthy controls performing six prehension tasks of the graded redefined assessment of strength, sensibility and prehension (GRASSP) was extracted. Using the spread of the forearm pitch angles, a classification of UL compensation was possible with very good to excellent accuracies in all six different prehension tasks. Furthermore, the spread of forearm pitch angles correlated strongly with qualitative prehension grip scores and the task duration. Therefore, we conclude that our proposed method has a high potential to classify compensation accurately and objectively and might be used to quantify the degree of UL compensation in ADLs. Thus, this method could be implemented in clinical trials investigating the effectiveness of interventions targeting UL functions.

Keywords: Modeling and identification of neural control using robots, Neural processes in rehabilitation, Human-machine interfaces and robotic applications
Abstract: Although neurorehabilitation is centered on motor learning and control processes, our understanding of how the brain learns to control movement is still limited. Motor adaptation is an error-driven motor learning process that is amenable to study in the laboratory setting. Behavioral studies of motor adaptation have coupled clever task design with computational modeling to study the control processes that underlie motor adaptation. These studies provide evidence of fast and slow learning states in the brain that combine to control neuromotor adaptation. Currently, the neural representation of these states remains unclear, especially for adaptation to changes in task dynamics, commonly studied using force fields imposed by a robotic device. Our group has developed the MR-Softwrist, a robot capable of executing dynamic adaptation tasks during functional magnetic resonance imaging (fMRI) that can be used to localize these networks in the brain. We simulated an fMRI experiment to determine if signal arising from a switching force field adaptation task can localize the neural representations of fast and slow learning states in the brain. Our results show that our task produces reliable behavioral estimates of fast and slow learning states, and distinctly measurable fMRI activations associated with each state under realistic levels of behavioral and measurement noise. Execution of this protocol with the MR-Softwrist will extend our knowledge of how the brain learns to cont

Keywords: Technologies for neurodevelopmental disorders, Clinical evaluation in robot-aided rehabilitation, Control strategies in rehabilitation robotics
Abstract: Controlling two objects simultaneously during a bimanual task is a cognitively demanding process; both hands need to be temporally and spatially coordinated to achieve the shared task goal. Children with unilateral spastic cerebral palsy (USCP) exhibit severe sensory and motor impairments to one side of their body that make the process of coordinating bimanual movements particularly exhausting. Prior studies have shown that performing visually-coupled task could reduce cognitive interference associated with performing `two tasks at once' in an uncoupled bimanual task. For children with USCP, who also present with cognitive delay, performing this type of task may allow them to process and plan their movement faster. We tested this hypothesis by examining the grip force control of 7 children with USCP during unimanual and visually-coupled bimanual tasks. Results demonstrated that despite the visual coupling, the bimanual coordination of these children remained impaired. However, there may be a potential benefit of visually-coupled task in encouraging both hands to initiate in concert. The implication of the study for children with USCP is discussed.

Keywords: Neural processes in rehabilitation, Robotic platforms in neuroscience, Modeling and identification of neural control using robots
Abstract: Innovative research in the fields of prosthetic, neurorehabilitation, motor control and human physiology has been focusing on the study of proprioception, the sense through which we perceive the position and movement of our body, and great achievements have been obtained regarding its assessment and characterization. However, how proprioceptive signals are combined with other sensory modalities and processed by the central nervous system to form a conscious body image, is still unknown. Such a crucial question was addressed in this study, which involved 23 healthy subjects, by combining a robot-based proprioceptive test with a specific analysis of electroencephalographic activity (EEG) in the µ frequency band (8-12 Hz). We observed important activation in the motor area contralateral to the moving hand, and besides, a substantial bias in brain activation and proprioceptive acuity when visual feedback was provided in addition to the proprioceptive information during movement execution. In details, brain activation and proprioceptive acuity were both higher in case of movements performed with visual feedback. Remarkably, we also found a correlation between the level of activation in the brain motor area contralateral to the moving hand and the value of proprioceptive acuity.

Keywords: Neuromodulation, Robotic platforms in neuroscience, Biomechanics and robotics in physical rehabilitation
Abstract: Somatosensory Evoked Potentials (SSEPs) are an important tool for both basic neuroscience research and evaluation of therapeutic techniques. While a large body of work exists in the study of electrically induced SSEPs both as a metric for therapeutic performance and tool for physiological research, comparatively little work has explored stretch response SSEPs evoked via tendon tapping. The measurement of SSEPs necessitates both timing and stimulation intensity consistency. This work presents an evaluation of a simple tapping device for automating this procedure and a comparison to manual tendon tapping demonstrating significantly reduced variability in both timing and intensity. The variable intensity nature of automated tapping is then used to measure SSEPs in a single subject, with apparent modulation of peak-peak amplitude by stimulation intensity.

Keywords: Body-machine interfaces, Design and development in rehabilitation robotics, Control strategies in rehabilitation robotics
Abstract: Competitive rehabilitation games can enhance motivation and exercise intensity compared to solo exercise; however, since such games may be played by two people with different abilities, game difficulty must be dynamically adapted to suit both players. State-of-the-art adaptation algorithms are based on players’ performance (e.g., score), which may not be representative of the patient’s physical and psychological state. Instead, we propose a method that estimates players’ states in a competitive game based on the covariation of players’ physiological responses. The method was evaluated in 10 unimpaired pairs, who played a competitive game in 6 conditions while 5 physiological responses were measured: respiration, skin conductance, heart rate, and 2 facial electromyograms. Two physiological linkage methods were used to assess the similarity of the players’ physiological measurements: coherence of raw measurements and correlation of heart and respiration rates. These linkage features were compared to traditional individual physiological features in classification of players’ affects (enjoyment, valence, arousal, perceived difficulty) into ‘low’ and ‘high’ classes. Classifiers based on physiological linkage resulted in higher accuracies than those based on individual physiological features, and combining both feature types yielded the highest classification accuracies (75% to 91%). These classifiers will next be used to dynamically adapt game difficulty during rehabilitation.

Keywords: New technologies and methodologies in human movement analysis, Human-machine interfaces and robotic applications, Cognitive robotics in rehabilitation
Abstract: There is increasing interest in using virtual reality (VR) in robotic neurorehabilitation. However, the use of conventional VR displays (i.e., computer screens), implies several transformations between the real movements in 3D and their 2D virtual representations that might negatively impact the rehabilitation interventions. In this study, we compared the impact on movement quality and cognitive load of novel vs. standard visualization technologies: i) Immersive VR (IVR) head-mounted display (HMD), ii) Augmented reality (AR) HMD, and iii) Computer screen. Twenty healthy participants performed simultaneously a motor and a cognitive task. Goal-oriented reaching movements were recorded using an HTC Vive controller. The cognitive load was assessed by the accuracy on a simultaneous counting task. The movement quality improved when visualizing the movements in IVR, compared to the computer screen. These improvements were more evident for locations that required movements in several dimensions. We found a trend to higher movement quality in AR than Screen, but worse than IVR. No significant difference was observed between modalities for the cognitive load. These results provide encouraging evidence that VR interventions using HMDs might be more suited for reaching tasks in several dimensions than a computer screen. Technical limitations might still limit the beneficial effects of AR, both in movement quality and cognitive load.

Keywords: Body-machine interfaces, Human-machine interfaces and robotic applications, Assistive robotics
Abstract: Assistive robotic arms have shown the potential to improve the quality of life of people with severe disabilities. However, a high performance and intuitive control interface for robots with 6-7 DOFs is still missing for these individuals. An inductive tongue computer interface (ITCI) was recently tested for control of robots and the study illustrated potential in this field. In this study, we investigated the possibility of developing a high performance tongue based joystick-like controller for robots. In the first part, different methods for mapping the 18 sensor signals to a 2D coordinate, as a touchpad, are compared. In the second part, a new approach for emulating an analog joystick by the ITCI system is introduced and its performance is evaluated based on the ISO9241-411 standard. Two subjects performed the multi-directional tapping test using first a standard joystick, next the ITCI system operated by hand and finally by the tongue. Throughput was measured as the evaluation parameter. The results show that the contact on the touchpads can be localized by almost 1 mm accuracy. The effective throughput of ITCI system for the multi-directional tapping test was 2.03 bps while keeping it in the hand and 1.31 bps when using it inside the mouth.

Keywords: Body-machine interfaces
Abstract: Myoelectric Computer Interfaces (MCIs) are a viable option to promote the recovery of movements following spinal cord injury (SCI), stroke, or other neurological disorders that impair motor functions. We developed and tested a MCI interface with the goal of reducing abnormal muscular activations due to compensatory strategies or undesired co-contraction after SCI. The interface mapped surface electromyographic signals (sEMG) into the movement of a cursor on a computer monitor. First, we aimed to reduce the co-activation of muscles pairs: the activation of two muscles controlled orthogonal directions of the cursor movements. Furthermore, to decrease the undesired concurrent activation of a third muscle, we modulated the visual feedback related to the position of the cursor on the screen based on the activation of this muscle. We tested the interface with six unimpaired and two SCI participants. Participants were able to decrease the activity of the targeted muscle when it was associated with the visual feedback of the cursor, but, interestingly, after training, its activity increased again. As for the SCI participants, one successfully decreased the co-activation of arm muscles, while the other successfully improved the selective activation of leg muscles. This is a first proof of concept that people with SCI can acquire, through the proposed MCI, a greater awareness of their muscular activity, reducing abnormal muscle simultaneous activations.

Keywords: Human-machine interfaces and robotic applications, Control strategies in rehabilitation robotics, Body-machine interfaces
Abstract: Pattern recognition based myoelectric control has been widely explored in the field of prosthetics, but little work has extended to other patient groups. Individuals with neurological injuries such as spinal cord injury may also benefit from more intuitive control that may facilitate more interactive treatments or improved control of functional electrical stimulation (FES) systems or assistive technologies. This work presents a pilot study with 10 individuals with cervical spinal cord injury between A and C on the American Spinal Injury Association Impairment Scale. Subjects attempted to elicit 10 classes of forearm and hand movements while their electromyogram (EMG) was recorded using a cuff of eight electrodes. Various well-known EMG features were evaluated using a linear discriminant analysis classifier, yielding classification error rates as low as 4.3% ± 3.9 across the 10 classes. Reducing the number of classes to five, those required to control a commercial therapeutic FES device, further reduced the error rates to (2.2% ± 4.4). Results from this study provide evidence supporting continued exploration of EMG pattern recognition techniques for use by high-level spinal cord injured populations as a method of intuitive control over interactive FES systems or assistive devices.

Keywords: Human-machine interfaces and robotic applications, Body-machine interfaces, Robotic prostheses - neural interfaces
Abstract: Natural myocontrol employs pattern recognition to allow users to control a robotic limb intuitively using their own voluntary muscular activations. The reliability of myocontrol strongly depends on the signals initially collected from the users, which must appropriately capture the variability encountered later on during operation. Since myoelectric signals can vary based on the position and orientation of the limb, it has become best practice to gather data in multiple body postures. We hereby concentrate on this acquisition protocol and investigate the relative merits of collecting data either statically or dynamically. In the static case, data for a desired hand configuration is collected while the users keep their hand still in certain positions, whereas in the dynamic case, data is collected while users move their limbs, passing through the required positions with a roughly constant velocity. Fourteen able-bodied subjects were asked to naturally control two dexterous hand prostheses mounted on splints, performing a set of complex, realistic bimanual activities of daily living. We could not find any significant difference between the protocols in terms of the total execution times, although the dynamic data acquisition was faster and less tiring. This would indicate that dynamic data acquisition should be preferred over the static one.

Keywords: Brain-machine interfaces, Control strategies in rehabilitation robotics
Abstract: An assistive robotic manipulator (ARM) can provide independence and improve the quality of life for patients suffering from tetraplegia. However, to properly control such device to a satisfactory level without any motor functions requires a very high performing brain-computer interface (BCI). Steady-state visual evoked potentials (SSVEP) based BCI are among the best performing. Thus, this study investigates the design of a system for a full workspace control of a 7 degrees of freedom ARM. A SSVEP signal is elicited by observing a visual stimulus flickering at a specific frequency and phase. This study investigates the best combination of unique frequencies and phases to provide a 16-target BCI by testing three different systems offline. Furthermore, a fourth system is developed to investigate the impact of the stimulating monitor refresh rate. Experiments conducted on two subjects suggest that a 16-target BCI created by four unique frequencies and 16-unique phases provide the best performance. Subject 1 reaches a maximum estimated ITR of 235 bits/min while subject 2 reaches 140 bits/min. The findings suggest that the optimal SSVEP stimuli to generate 16 targets are a low number of frequencies and a high number of unique phases. Moreover, the findings do not suggest any need for considering the monitor refresh rate if stimuli are modulated using a sinusoidal signal sampled at the refresh rate.

Keywords: Wearable robotic systems, Exoskeletons, Assistive robotics
Abstract: A new tendon driven mechanism, embedded into a soft hand exoskeleton for rehabilitation and assistance, was proposed in this study. The proposed solution was a pulley flexion mechanism inspired by the human musculoskeletal system to enable a natural and comfortable finger flexion. A biomechanical constraint for the finger flexion motion states that the relation between the proximal interphalangeal joint angle of the finger should always be flexed around 1.5 times the distal interphalangeal joint angle. The study aimed to comply with this constraint, by simultaneously distributing the forces over the distal and middle finger phalanges. For evaluation, the voluntary and exoskeleton flexions were compared based on the relation between the proximal and distal interphalangeal joint angles. The results showed that during the exoskeleton flexion the relation between the interphalangeal joints complied with the biomechanical constraint, where the proximal interphalangeal joint angle was 1.5 times larger than the distal interphalangeal joint. This ensures that the mechanism flexes the finger comfortably. The proposed solution is therefore a promising design for a novel soft exoskeleton that will be used for training and assistance of patients with hand paralysis.

Keywords: Wearable robotic systems, Assistive robotics, Exoskeletons
Abstract: Wearable actuators in lower-extremity active orthoses or prostheses have the potential to address a variety of gait disorders. However, whenever conventional joint actuators exert moments on specific limbs, they must simultaneously impose opposing reaction moments on other limbs, which may reduce the desired effects and perturb posture. Momentum exchange actuators exert free moments on individual limbs, potentially overcoming or mitigating these issues.

We simulate unperturbed gait to compare conventional joint actuators placed on the knee or hip of the swing leg, and equivalent angular momentum exchange actuators placed on the shank or thigh. Our results indicate that, while conventional joint actuators excel at increasing toe clearance when assisting knee flexion, free moments can yield greater increases in stride length when assisting knee extension or hip flexion.

Keywords: Modeling and identification of neural control using robots, Neural processes in rehabilitation, Exoskeletons
Abstract: One key question in motor learning is how the complex tasks in daily life should be trained. Often, complex tasks are directly taught as a whole, even though a priori training of simple movement components has been shown to be more effective for complex tasks. The important implication of the part-whole transfer paradigm is that the training of most complex tasks could be simplified and, subsequently, devices used to train (e.g. rehabilitation robots) can become simpler and robot-assisted rehabilitation could become more accessible. However, often the recomposition of several simple movement components to a complete complex movement is forgotten. At least for the last training step a complex rehabilitation device may be required. In a pilot study, we wanted to investigate if a complex robotic device (namely, a robot with many degrees of freedom), such as the ARMin rehabilitation exoskeleton robot, is really beneficial for training the coordination between several simpler movement components or if visual feedback would have equal benefits. In a study, involving 16 healthy participants instructed in a complex rugby motion, we could show first trends on the following two aspects: 1) part-whole transfer paradigm seems to hold true and simple robots might be used for training movement primitives. 2) Visual feedback does not have the same potential as visuo-haptic guidance for movement recomposition of complex task. Therefore, complex rehabilitation robots may be still beneficial.

Keywords: Human-machine interfaces and robotic applications, Design and development in rehabilitation robotics
Abstract: Improving upon the therapist–device relationship is an important aspect that will increase the number of upper-limb robotic rehabilitation devices being used for therapy. One path to strengthen this relationship is for these devices to generate large data sets that rehabilitation therapists can use to enhance their patient assessment procedures. In this article, a national survey of Canadian therapists was conducted in order to learn about their data collection and analysis methods. A total of 33 responses were gathered from an online survey. These results show that there is a demand for the collection and visualization of various patient data, some of which cannot be easily collected with existing methods. It was also seen that there exists a large variation between therapists about which major steps constitute the general rehabilitation process. From these results, a set of fourteen general software requirements has been created. Insights from the survey regarding influences on software designs are briefly discussed. This research helps to enable the development of software systems that increase the interaction potential between therapists and robotic devices.

Keywords: Wearable robotic systems, Robotic orthoses - design and development, Design and development in rehabilitation robotics
Abstract: Recent technological improvements are consistently improving the efficacy of wearable mechatronic devices designed to support rehabilitation. However, it has been identified that there is currently a limited number of devices that can perform resistive motion tasks. To address this limitation, a Wearable Mechatronics-Enabled (WearME) Glove has been developed to support rehabilitative motion tasks. Using the WearME Glove, a control system was developed to enable the performance of resistive finger and wrist motion tasks. An initial evaluation of the device applied to rehabilitation tasks shows that average control errors of 2.4% and 1.5% were achieved for a resistive finger task and a resistive wrist flexion–extension task, respectively. In addition, an analysis of the resistive duration of each task showed that an average of 69%, 76% and 83% of the index finger, thumb and wrist motion duration, respectively, were being resisted by the WearME Glove. The results of this study show that the WearME glove can provide consistent resistance to the finger and wrist for different rehabilitation tasks.

Keywords: Body-machine interfaces, Wearable robotic systems
Abstract: According to the World Health Organization, stroke is the third leading cause of disability. A common consequence of stroke is hemiparesis, which leads to the impairment of one side of the body and affects the performance of activities of daily living. It has been proven that targeting the motor impairments as early as possible while using wearable mechatronic devices as a robot assisted therapy, and letting the patient be in control of the robotic system, can improve the rehabilitation outcomes. However, despite the increased progress on control methods for wearable mechatronic devices, a need for a more natural interface that allows for better control remains. In this work, a user-independent gesture classification method based on a sensor fusion technique using surface electromyography (EMG) and an inertial measurement unit (IMU) is presented. The Myo Armband was used to extract EMG and IMU data from healthy subjects. Participants were asked to perform 10 types of gestures in 4 different arm positions while using the Myo on their dominant limb. Data obtained from 14 participants were used to classify the gestures using a Multilayer Perceptron Network. Finally, the classification algorithm was tested on 5 novel users, obtaining an average accuracy of 78.94%. These results demonstrate that by using the proposed approach, it is possible to achieve a more natural human machine interface that allows better control of wearable mechatronic devices during robot assisted therapies.

Keywords: Biomechanics and robotics in physical rehabilitation, Human-machine interfaces and robotic applications, From lab to market - Regulatory issues
Abstract: Rehabilitation robots can provide high intensity and dosage training or assist patients in activities of daily living and decrease physical strain on clinicians. However, the physical human robot interaction poses a major safety issue, as the close physical contact between user and robot can lead to injuries. Moreover, the magnitude of forces as well as best practices for measuring them, are widely unknown. Therefore, a measurement setup was developed to assess normal and tangential forces that occur in the contact area between an arm and a splint. Force sensitive resistors and a force / torque sensor were combined with two different splint shapes. Initial experiments indicated that the setup gives some insight into magnitudes and distribution of normal forces on the splint-forearm-interface. Experiment results show a dependency of force distributions on the splint shape and sensor locations. Based on these outcomes, we proposed an improved setup for subsequent investigations.

Keywords: Wearable robotic systems, Assistive robotics - home robots, New technologies and methodologies in biomechanics
Abstract: Pathological tremor is caused by a variety of neurological diseases. Although it is not life-threatening, it brings great inconvenience to patients. Traditional treatments including medication, rehabilitation programs and deep brain stimulation (DBS) have shown limited effectiveness along with risks and side effects. In order to overcome the limitations of these treatments, a new method, wearable exoskeleton technology, is introduced which aims to provide a new solution for tremor management. Based on this method, a wrist tremor suppression exoskeleton (WTSE) is developed in this research. For controllable damping force, a magnetorheological (MR) fluid damper is designed and an embedded acquisition platform is used to acquire real-time tremor information. The total weight of the WTSE is 262.13 g and the maximum continues damping force reaches 8 N. The prototype is wearable and the damping force is real-time adjustable. According to the preliminary results, the signal acquisition system can obtain reliable data and the WTSE can reduce the amplitude of acceleration and angular velocity of simulated tremor by 60.39% and 55.07%, respectively.

Keywords: Wearable robotic systems, Design and development in rehabilitation robotics, Biomechanics and robotics in physical rehabilitation
Abstract: Functional impairment of the hand, for example after a stroke, can be partially improved by intensive training. This is currently done by physiotherapy and the optimal intensity of hand rehabilitation programs is usually not reached due to a lack in human resources (high costs) and patients fatigue. In this work a cost-effective soft exosuit to support the hand’s grasping function is presented. The system is based on tendon-like wires and all fingers except the little finger are actuated. Each of the remaining four fingers is bidirectionally controlled by an electrical motor. This allows a variety of gripping situations, e.g. a power or precision grip. Our prototype weighs 435 g, including the battery and can be worn on the upper arm. The force applicable for a power grip exceeds 20N with a maximum gripping frequency of 4 Hz. Furthermore, a force control is implemented, giving the wearer the opportunity to grab sensitive objects. All components used are available in different sizes, allowing a quick and individual preparation per patient. Therefore, our prototype can be used for rehabilitation while doing activities of daily living (ADL) starting on the day of the injury.

Keywords: Brain-machine interfaces, Human-machine interfaces and robotic applications, Wearable robotic systems
Abstract: The use of robotic devices to provide active motor support and sensory feedback of ongoing motor intention, by means of a Brain Computer Interface (BCI), has received growing support by recent literature, with particular focus on neurorehabilitation therapies. At the same time, performance in the use of the BCI has become a more critical factor, since it directly influences congruency and consistency of the provided sensory feedback. As motor imagery is the mental simulation of a given movement without depending on residual function, training of patients in the use of motor imagery BCI can be extended beyond each rehabilitation session, and practiced by using simpler devices than rehabilitation robots available in the hospital. In this work, we investigated the use of haptic stimulation provided by vibrating electromagnetic motors to enhance BCI system training. The BCI is based on motor imagery of hand grasping and designed to operate a hand exoskeleton. We investigated whether haptic stimulation at fingerpads proves to be more effective than stimulation at wrist, already experimented in literature, due to the higher density of mechano-receptors. Our results did not show significant differences between the two body locations in BCI performance, yet a wider and more stable event-relateddesynchronization appeared for the finger-located stimulation. Future investigations will put in relation training with haptic feedback at fingerpads with BCI performance using the handexoskel

Keywords: Body-machine interfaces, Robotic prostheses - neural interfaces
Abstract: Surface electromyography (sEMG) is widely used in various fields to analyze user intentions. Conventional sEMG-based classifications are electrode-dependent; thus, trained classifiers cannot be applied to other electrodes that have different parameters. This defect degrades the practicability of sEMG-based applications. In this study, we propose a virtual sEMG signal-assisted classification to achieve electrode-independent classification. The virtual signal for any electrode configuration can be generated using muscle activation signals obtained from the proposed model. The feasibility of the virtual signal is demonstrated with regard to i) classifications using fewer sEMG channels by a pre-trained classifier without re-training and ii) electrode-independent classifications. This study focuses on preliminary tests of virtual sEMG signal-assisted classification. Future studies should consider model improvement and experiments involving more subjects to achieve plug-and-play classification.

Keywords: Assistive robotics, Robot-aided living, Human-machine interfaces and robotic applications
Abstract: Assistive robotic manipulators have the potential to support the lives of people suffering from severe motor impairments. They can support individuals with disabilities to independently perform daily living activities, such as drinking, eating, manipulation tasks, and opening doors. An attractive solution is to enable motor impaired users to teach a robot by providing demonstrations of daily living tasks. The user controls the robot ‘manually’ with an intuitive human-robot interface to provide demonstration, which is followed by the robot learning of the performed task. However, the control of robotic manipulators by motor impaired individuals is a challenging topic. In this paper, a novel head gesture-based interface for hands-free robot control and a framework for robot learning from demonstration are presented. The head gesture-based interface consists of a camera mounted on the user’s hat, which records the changes in the viewed scene due to the head motion. The head gesture recognition is performed using the optical flow for feature extraction and support vector machine for gesture classification. The recognized head gestures are further mapped into robot control commands to perform object manipulation task. The robot learns the demonstrated task by generating the sequence of actions and Gaussian Mixture Model method is used to segment the demonstrated path of the robot’s end-effector. During the robotic reproduction of the task, the modified Gaussian Mixture Model and

Keywords: Technologies for neurodegenerative disorders, New technologies and methodologies in human movement analysis
Abstract: Proprioceptive deficits are frequent and disabling symptoms of neurological diseases such as Multiple Sclerosis (MS). These deficits are poorly understood partly because of the limited sensitivity and reproducibility of clinical measures. However, their assessment is crucial in planning and evaluating rehabilitative treatments. Therefore, we designed a device and a protocol for assessing proprioceptive deficits by evaluating the position and force control performance. We focused on bimanual tasks, as most daily life activities require the combined use of both hands while MS induces coordination problems and often affects the two arms differently. Specifically, without being able to see their arms, subjects had (1) to reach with their hands a target positions holding objects of equal or different weights; (2) to exert equal isometric forces with the two hands in upward direction against rigid constraints at the same or different heights. For a first proof of concept of the feasibility we enrolled seven MS subjects with different levels of upper limb impairment and seven sex and age matched controls. We found that the ability to exert symmetric forces with both arms was significantly altered in all MS subjects, while position control decreased only for higher level of impairment. These preliminary findings suggest that in people with MS the ability to exert bilaterally required levels of force might be affected earlier compared to the ability to control hand position.

Keywords: Neuromodulation, Assistive robotics, Robotic prostheses - modeling and simulation
Abstract: Individuals with paralyzed limbs due to spinal cord injuries lack the ability to perform the reaching motions necessary to every day life. Functional electrical stimulation (FES) is a promising technology for restoring reaching movements to these individuals by reanimating their paralyzed muscles. We have proposed using a quasi-static model-based control strategy to achieve reaching controlled by FES. This method uses a series of static positions to connect the starting wrist position to the goal. As a first step to implementing this controller, we have completed a simulation study using a MATLAB based dynamic model of the arm in order to determine the suitable parameters for the quasi-static controller. The selected distance between static positions in the path was 6 cm, and the amount of time between switching target positions was 1.3 s. The final controller can complete reaches of over 30 cm with a median accuracy of 6.8 cm.

Keywords: Body-machine interfaces, Exoskeletons, From lab to market - Usability evaluation
Abstract: Over the last decade, the use of wearable exoskeletons for human locomotion assistance has become more feasible. The VariLeg powered lower limb robotic exoskeleton is an example of such systems, potentially enabling paraplegic users to perform upright activities of daily living. The acceptance of this type of robotic assistive technologies is often still affected by limited usability, in particular regarding the physical interface between the exoskeleton and the user (here referred to as pilot). In this study, we proposed and evaluated a novel pilot attachment system (PAS), which was designed based on a user-centered design with experienced paraplegic exoskeleton users. Subjective assessments to compare usability aspects of the initial and the redesigned physical interfaces were conducted with two paraplegic and five healthy pilots. The redesigned PAS showed a 45% increase in the system usability scale (SUS), normalized to the PAS of a commercial exoskeleton assessed in the same manner. Pain rating scales assessed with healthy pilots indicated an increased comfort using the redesigned PAS while performing several activities of daily living. Overall, an improvement in usability relative to the initial PAS was achieved through intensified user evaluation and individual needs assessments. Hence, the user-centered design of physical body-machine interfaces has the potential to positively influence the usability and acceptance of lower limb exoskeletons for paraplegic users.

Keywords: Technologies for neurodegenerative disorders, New technologies and methodologies in human movement analysis, New technologies and methodologies in biomechanics
Abstract: Parkinson disease (PD) is a common neurodegenerative disorders characterized by motor and non-motor impairments. Since the quality of life of PD patients becomes poor while pathology develops, it is imperative to improve the identification of personalized rehabilitation and treatments approaches based on the level of the neurodegeneration process. Objective and precise assessment of the severity of the pathology is crucial to identify the most appropriate treatments. In this context, this paper proposes a wearable system able to measure the motor performance of PD subjects. Two inertial devices were used to capture the motion of the lower and upper limbs respectively, while performing six motor tasks. Forty-one kinematic features were extracted from the inertial signals to describe the performance of each subjects. Three unsupervised learning algorithms (k-Means, Self-organizing maps (SOM) and hierarchical clustering) were applied with a blind approach to group the motor performance. The results show that SOM was the best classifier since it reached accuracy equal to 0.950 to group the instances in two classes (mild vs advanced), and 0.817 considering three classes (mild vs moderate vs severe). Therefore, this system enabled objective assessment of the PD severity through motion analysis, allowing the evaluation of residual motor capabilities and fostering personalized paths for PD rehabilitation and assistance.

Keywords: Human-machine interfaces and robotic applications, Body-machine interfaces, Assistive robotics
Abstract: Early fall detection is an important issue during gait rehabilitation training. This paper proposes an approach for pre-impact fall detection during gait rehabilitation training based on a 3D convolutional neural network (CNN). Firstly, pre-training data is collected and used to pre-train the 3D CNN to differentiate between a normal walking and a fall based on their general spatio-temporal patterns. Secondly, fine-tuning data is created by combining the pre-training data with a 3-second normal walking sample collected from a new trainee whose falls are to be detected. The pre-trained 3D CNN is further fine-tuned by the fine-tuning data to learn the spatio-temporal patterns of the new trainee. Finally, a temporal sliding window is used to feed video snippets into the fine-tuned 3D CNN for fall detection. To the best of our knowledge, this is the first pre-impact fall detection approach based on a 3D CNN using RGB images. Moreover, the training strategy used to train the 3D CNN can alleviate the generalization issue of the 3D CNN when only limited training data is available in gait rehabilitation training. With 225 testing trials from 5 trainees, the proposed pre-impact fall detection approach achieves a detection accuracy of 100% within 0.5 second after falls start. Experiment results show that this approach is efficient, accurate, and practical in achieving intelligent fall detection during gait rehabilitation training.

Keywords: Design and development in rehabilitation robotics, Wearable robotic systems
Abstract: When walking, the trunk not only oscillates in the vertical direction, but also in the medial-lateral direction. We developed a novel backpack that uses the medial-lateral oscillations of the trunk as input motion to drive medial-lateral oscillations of weight carried in a modified backpack. We use a combination of spring and damping elements to control mass motion, resulting in the ability to prescribe a variety of mass oscillation amplitudes and phase angles. We propose the device as a platform that can be used to study medial-lateral stability during walking. In particular, if the body’s ability to predict medial-lateral centre-of-mass state is affected by an oscillating external mass. In this paper, we present the design, model, and model evaluation of our novel load carriage device. During testing, our model was able to predict the oscillation dynamics of the carried mass while walking: demonstrating its capability to create a variety of load carriage scenarios for the user.

Keywords: Modeling and identification of neural control using robots, Exoskeletons, New technologies and methodologies in human movement analysis
Abstract: Assistive technology for the lower extremities has shown great promise towards improving gait function in people following neuromuscular injuries. However, our previous work assisting knee flexion torque in post-stroke Stiff-Knee gait found that augmenting strength can induce secondary complications such as spasticity due to stretching of the rectus femoris. In this work, we explore whether we could have obtained improved knee flexion but avoided a spastic response by simulating combinations of hip and knee flexion torques using musculoskeletal modeling and simulation. We explore previously collected data on a case-by-case basis to determine individual-specific quadriceps reflex thresholds based on estimated rectus femoris muscle fiber stretch velocities. We then implemented a forward simulation framework to identify the subject-specific hip-knee assistance prescription to improve knee range of motion without initiating a spastic response. The obtained subject-specific assistive prescription informs the development of new gait assistance strategies for post-stroke gait and could be extended to other neuromuscular gait impairments.

Keywords: Neuromodulation, Brain-machine interfaces, Neural processes in rehabilitation
Abstract: Objective: Cerebellar Transcranial direct current stimulation (ctDCS) of cerebellar lobules is challenging due to the complexity of the cerebellar structure. Therefore, we present a freely available computational pipeline to determine the subject-specific lobule-specific electric field distribution during ctDCS. Methods: The computational pipeline isolates subject-specific cerebellar lobules based on a spatially unbiased atlas template (SUIT) for the cerebellum, and then calculates the lobule-specific electric field distribution during ctDCS. The computational pipeline was tested using the Colin27 Average Brain. The 5 cm × 5 cm anode was placed 3 cm lateral to inion, and the same sized cathode was placed on the contralateral supra-orbital area (called Manto montage) and buccinators muscle (called Celnik montage). A published 4x1 HD-ctDCS electrode montage was also implemented for a comparison using analysis of variance. Results: The electric field strength of both the Celnik and the Manto montages affected the lobules Crus II, VIIb, VIII, and IX of the targeted cerebellar hemispheres while Manto montage had a more bilateral effect. The HD-ctDCS montage primarily affected the lobules Crus I, Crus II, VIIb of the targeted cerebellar hemisphere. Discussion: Our freely available subject-specific computational modeling pipeline can be used to analyze lobule-specific electric field distribution to select an optimal ctDCS electrode montage.

Keywords: Assistive robotics, Wearable robotic systems, Design and development in rehabilitation robotics
Abstract: Soft exosuits have advantages over their rigid counterparts in terms of portability, transparency and ergonomics. Our previous work has shown that a soft, fabric-based exosuit, actuated by an electric motor and a Bowden cable, reduced the muscular effort of the user when flexing the elbow. This previous exosuit used a gravity compensation algorithm with the assumption that the shoulder was adducted at the trunk. In this investigation, the shoulder elevation angle was incorporated into the gravity compensation control via inertial measurement units (IMUs). We assessed our updated gravity compensation model with four healthy, male subjects (age: 26.2±1.19 years) who followed an elbow flexion reference trajectory which reached three amplitudes (25°, 50°, 75°) and was repeated at three shoulder angles (25°, 50°, 75°). To assess the performance of the exosuit; the smoothness, tracking accuracy and muscle activity were investigated during each motion. We found a reduction of biceps brachii activation (24.3%) in the powered condition compared to the unpowered condition. In addition, there was an improvement in kinematic smoothness (0.83%) and a reduction of tracking accuracy (26.5%) in the powered condition with respect to the unpowered condition. We can conclude that the updated gravity compensation algorithm has increased the number of supported movements by considering the shoulder elevation, which has improved the usability of the device.

Keywords: Robotic orthoses - design and development, Assistive robotics, Design and development in rehabilitation robotics
Abstract: This paper provides a proof of concept for an actuating system comprised of a linear actuator and a spring steel strip that enables bidirectional articulation of a finger by transmitting the force directly to the finger tip. This proposed design can be distinguished from other orthosis designs, which use rigid linkages or cables with DC motors or fluidic systems for force generation and transmission. We designed an experimental setup with a 3D-printed model finger to mimic a passive human finger on which the actuation system was mounted and tested. The finger was positioned such that it would curl upward to lift various masses when articulated by the actuating system to demonstrate the system's force generation capability. We tested two linear actuators and two steel strips, using a wide range of masses to determine which would be the most suitable components for our design. We analyzed motion profiles, joint angles, force generation, and actuator stroke velocities during various experimental trials. Our results demonstrate that our actuating system is capable of generating sufficient forces and motions with an adequate response time to be used in the design of a hand orthosis for grasping/releasing assistance. From our tests, a prototype was designed with three linear actuators positioned on the dorsal side of the hand and actuated the thumb, index, and middle fingers. Future work will include sensor integration and performance evaluation of the orthosis.

Keywords: Wearable robotic systems, Orthotics - modeling and simulation, Biomechanics and robotics in physical rehabilitation
Abstract: Significant impairments in upper extremity function are commonly observed after stroke. While the efficacy of robotic training has been demonstrated, the use of these devices is confined to the laboratory setting due to its complexity and power requirements. In this study, we developed a passive, portable device (Portable Elbow Movement Assistant; PEMA) providing assistance for elbow movements of stroke survivors. The geometric properties of the device were designed to allow morphological changes in the elastic components during movements, so that the assistance produced by the elastic component counteract the angle-dependent flexor hypertonia of stroke survivors. A mathematical model for the proposed design was first developed to characterize the assistance provided by the device. The device performance was then tested in a pilot testing with four healthy subjects, for whom a custom device to simulate elbow flexor hypertonia (providing an increased resistance for the extended posture) was implemented. The proposed device was found to effectively counteract the angle-dependent flexion moment, produced by the hypertonia simulator, as a significant decrease was observed in the slope of the angle-activation relationship (movement) and activation level (stationary) of the triceps brachii muscle. The assistance did not affect the activation of the antagonist muscle, indicating an independent modulation of the agonist/antagonist muscles was resulted.

Keywords: Robotic prostheses - design and development, Wearable robotic systems, Body-machine interfaces
Abstract: Lower-limb amputees demonstrate decreased performance in stair ambulation compared to their intact-limb counterparts. An estimated 21% of amputees can navigate stairs without a handrail; almost 33% do not use stairs at all. The absence of tactile sensation on the bottom of the foot, creating uncertainty in foot placement, may be supervened by integrating sensory feedback into prosthesis design. Here we describe the design and evaluation of a haptic feedback system worn on the thigh to provide vibrotactile cues of foot placement with respect to stair steps. Tactor discrimination and foot placement awareness tests were performed to analyze system efficacy. Control participants wearing ski boots (N=10) and below-knee amputees (N=2) could discriminate individual tactor vibrations with 95.4% and 90.1% accuracy, respectively. The use of vibrotactile feedback increased accuracy in reporting foot placement by 15% and 17.5%, respectively. These results suggest that using vibrotactile arrays for sensory feedback may improve stair descent performance in lower-limb amputees.

Keywords: Modeling and identification of neural control using robots, Robotic platforms in neuroscience
Abstract: In a stable bimanual trajectory tracing task with interlimb spatial and temporal synchrony, blocking the visual information from one hand may alter the performance of either hand. In this paper, we investigate the effect of visual information on motor behaviour of dominant and non-dominant hands during a bimanual task, with a focus on motor lateralization theory’s anticipation for a more pronounced distortion on one hand due to visual information withdrawal. To address this question, four bimanual circle tracing experiments were designed with two rehabilitation robotic arms with real time visual feedback. Two experiments were conducted under the free-visual condition whereas the visual feedback from one hand was blocked for the other two. The in-depth analysis of the metrics extracted from 685 circles, drawn by 6 participants, revealed that non- dominant hand, when visible, generally performs worse than the dominant hand, for instance it exhibits less circularity. In their invisible modes, the performance of the dominant and non-dominant hands displayed inconsistent difference across the participants. Moreover, both hands showed a higher pace when partial visual information was available. Our findings using this robotic framework as a systematic tool on developing new paradigms are discussed.

Keywords: Robotic platforms in neuroscience, New technologies and methodologies in biomechanics, New technologies and methodologies in human movement analysis
Abstract: In the development of a robotic therapy system, tests must be first run to guarantee safety and performance of the system before actual human trials. Lower-limb robotic therapy system has an inherit injury risk and a human–like stunt robot is desirable. This study proposes such an alternative: anthropomorphic legs with a bio–inspired control method affording a human–like test bench for the robotic therapy system. Electromyography (EMG) of a mildly hemiparetic stroke patient was measured during body–weight–supported treadmill walking. The motor strategy of the hemiparetic gait was extracted from the EMG data and applied to the control of the anthropomorphic legs. We employed the concept of equilibrium point (EP) to extract motor synergies and strategy. The EP-based synergies expressed by the composites of muscle mechanical impedance clarified motor strategy including aspects related to the impedance and virtual trajectory. Results show that the EP–based synergies were able to characterize neuromuscular patterns of pathological gait. The anthropomorphic legs were able to reproduce patient’s gait by mimicking the EP–based synergies.

Keywords: Neural processes in rehabilitation, Human-machine interfaces and robotic applications, Design and development in rehabilitation robotics
Abstract: Performance of lower limb prostheses is related not only to the mechanical design and the control scheme, but also to the feedback provided to the user. Proprioceptive feedback, which is the sense of position and movement of one’s body parts, can improve the utility as well as facilitate the embodiment of the prosthetic device. Recent studies have shown that kinesthetic sense can be elicited when vibrating a muscle tendon proximal to the targeted joint. However, consistency and quality of the elicited sensation depend on several parameters and muscle tendons after lower limb amputation may not always be accessible. Here, we developed an experimental protocol to quantitatively and qualitatively assess the elicited kinesthetic illusion when vibrating a muscle belly. Furthermore, we explored ways to improve consistency and strength of the illusion by integrating another non-invasive feedback method, namely skin stretch. Our preliminary results from tests conducted with a person with below knee amputation show that stretching skin while vibrating a muscle belly on the residual limb provided a stronger and more consistent kinesthetic illusion (90%) than only vibrating the muscle (50%). In addition, we found that stretching skin enhances the range (1.5 times) and speed (3.5 times) of the illusory movement triggered by muscle vibration. These findings may enable the development of mechanisms for controlling feedback parameters for closing the control loop for various walking routine

Keywords: Human-machine interfaces and robotic applications, Control strategies in rehabilitation robotics, Body-machine interfaces
Abstract: Learning to get by without an arm or hand can be very challenging, and existing prostheses do not yet fill the needs of individuals with amputations. One promising solution is to improve the feedback from the device to the user. Towards this end, we present a simple machine learning interface to supplement the control of a robotic limb with feedback to the user about what the limb will be experiencing in the near future. A real-time prediction learner was implemented to predict impact-related electrical load experienced by a robot limb; the learning system's predictions were then communicated to the device's user to aid in their interactions with a workspace. We tested this system with five able-bodied subjects. Each subject manipulated the robot arm while receiving different forms of vibrotactile feedback regarding the arm's contact with its workspace. Our trials showed that using machine-learned predictions as a basis for feedback led to a statistically significant improvement in task performance when compared to purely reactive feedback from the device. Our study therefore contributes initial evidence that prediction learning and machine intelligence can benefit not just control, but also feedback from an artificial limb. We expect that a greater level of acceptance and ownership can be achieved if the prosthesis itself takes an active role in transmitting learned knowledge about its state and its situation of use.

Keywords: Robotic platforms in neuroscience, Design and development in rehabilitation robotics, Modeling and identification of neural control using robots
Abstract: Increased reticulospinal (RS) function has been observed to cause both positive and negative outcomes in the recovery of motor function after corticospinal lesions such as stroke. Current knowledge of RS function is limited by the lack of accurate, noninvasive methods for measuring RS function. Recent studies suggest that the RS tract may be involved in processing and generating Long Latency Responses (LLRs). As such, LLRs, elicited by applying precisely controlled perturbations, can thus act as a reliable stimulus to measure brainstem function using fMRI with high signal-to-noise ratio.

In this paper, we present StretchfMRI, a novel technique that enables simultaneous recording of neural and muscular activity during motor responses conditioned by robotic perturbations, which allows direct investigation of the neural correlates of LLRs.

Via preliminary validation experiments, we demonstrate that our technique can reliably elicit and identify LLRs in two wrist muscles–Flexor Carpi Radialis and Extensor Carpi Ulnaris. Moreover, via a single-subject pilot experiment, we show that the occurrence of an LLR in a flexor and extensor muscles modulates neural activity in distinct regions of the brainstem. The observed somatotopic organization is in agreement with the double reciprocal model of RS function observed in animal models, in which the right medullary and left pontine reticular formation are responsible for control of the motor activity in flexors and extensors, respectivel

Keywords: Neural processes in rehabilitation, Neuromodulation
Abstract: Movement is associated with power changes over sensory-motor areas in different frequency ranges, including beta (15-30 Hz). In fact, beta power starts decreasing before the movement onset (event-related desynchronization, ERD) and rebounds after its end (event-related synchronization, ERS). There is increasing evidence that beta modulation depth (measured as ERD-ERS difference) increases with practice in a planar reaching task, suggesting that this measure may reflect plasticity processes. In the present work, we analyzed beta ERD, ERS and modulation depth in healthy subjects to determine their changes over three regions of interest (ROIs): right and left sensorimotor and frontal areas, during a reaching task with the right arm. We found that ERD, ERS and modulation depth increased with practice with lower values over the right sensorymotor area. Timing of peak ERD and ERS were similar across ROIs, with ERS peak occurring earlier in later sets. Finally, we found that beta ERS of the frontal ROI involved higher beta range (23-29 Hz) than the sensory-motor ROIs (15-18 Hz). Altogether these results suggest that practice in a reaching task is associated with modification of beta power and timing. Additionally, beta ERS may have different functional meaning in the three ROIs, as suggested by the involvement of upper and lower beta bands in the frontal and sensorimotor ROIs, respectively.

Keywords: Neural processes in rehabilitation, Neuromodulation
Abstract: Movement is accompanied by modulation of oscillatory activity in different ranges over the sensorimotor areas. This increase is more evident in normal subjects and less in patients with Parkinson’s Disease (PD), a disorder associated with deficits in the formation of new motor skills. Here, we investigated whether such EEG changes improved in a group of PD patients, after two different treatments and whether this relates to performance. Subjects underwent either a session of 5 Hz repetitive Transcranial Magnetic Stimulation (rTMS) over the right posterior parietal cortex or a 4-week Multidisciplinary Intensive Rehabilitation Treatment (MIRT). We used a reaching task with visuo-motor adaptation to a rotated display in incremental 10° steps up to 60°. Retention of the learned rotation was tested before and after either intervention over two consecutive days. High-density EEG was recorded throughout the testing. We found that patients adapted their movements to the rotated display similarly to controls, although retention was poorer. Both rTMS and MIRT lead to improvement in retention of the learned rotation. Mean beta modulation levels changed significantly after MIRT and not after rTMS. These results suggest that rTMS produced local improvement reflected in enhanced short-term skill retention; on the other hand, MIRT determined changes across the contralateral sensorimotor area, reflected in beta EEG changes.

Keywords: Wearable robotic systems, Robotic prostheses - modeling and simulation, Body-machine interfaces
Abstract: Amputees often find wearing a prosthetic limb for a long period of time uncomfortable. Prosthetic sockets that adjust the socket's fit automatically, or adaptive sockets, would encourage amputees to wear their prosthesis more frequently. In this work, we simulate the control system design of a Multiple-Input, Multiple-Output (MIMO) adaptive socket using principles of optimal control and robust control. A data-driven model of the socket-limb interface is first obtained by applying regression to open-loop recordings of the socket interacting with the limb during a simulated grasping task. A MIMO controller is then designed to maintain a desired uniform socket fit. An H-Infinity controller, obtained from loop shaping synthesis using the Glover-McFarlane method, is shown to perform comparably to a Linear Quadratic Gaussian (LQG) controller while maintaining robustness to uncertainties in the socket-limb interface model. This work then outlines a potential procedure on how to develop the control system for a real adaptive prosthetic socket with multiple sensors and actuators.

Keywords: Human-machine interfaces and robotic applications, Control strategies in rehabilitation robotics, Robotic prostheses - neural interfaces
Abstract: Nowadays, electric-powered hand prostheses do not provide adequate sensory instrumentation and artificial feedback to allow users voluntarily and finely modulate the grasp strength applied to the objects. In this work, the design of a control architecture for a myocontrol-based regulation of the grasp strength for a robotic hand equipped with contact force sensors is presented. The goal of the study was to provide the user with the capability of modulating the grasping force according to target required levels by exploiting a vibrotactile feedback. In particular, the whole human-robot control system is concerned (i.e. myocontrol, robotic hand controller, vibrotactile feedback.) In order to evaluate the intuitiveness and force tracking performance provided by the proposed control architecture, an experiment was carried out involving four na"ive able-bodied subjects in a grasping strength regulation task with a myocontrolled robotic hand (the University of Bologna Hand), requiring for grasping different objects with specific target force levels. The reported results show that the control architecture successfully allowed all subjects to achieve all grasping strength levels exploiting the vibrotactile feedback information. This preliminary demonstrates that, potentially, the proposed control interface can be profitably exploited in upper-limb prosthetic applications, as well as for non-rehabilitation uses, e.g. in ultra-light teleoperation for grasping devices.

Keywords: Wearable robotic systems, Design and development in rehabilitation robotics, Exoskeletons
Abstract: In the author's previous study, an origami-structured compliant actuator, named OSCA, was invented for human-robot interaction where sufficient moment output is required under a limited power supply. In this study, the compressibility of airflow within the actuator that essentially provides elasticity as well as backdriveability is in the spotlight. Elasticity can contribute to increasing the actuator's power output in the same way as does a series elastic actuator (SEA). This study demonstrates that an OSCA operated using a servo valve is equivalent to an SEA with a typical spring. In other words, power output amplification can be achieved with appropriate stiffness of the air inside.