Search

Over the past few decades, there have been many studies of human-human physical interaction to better understand why humans physically interact so effectively and how dyads outperform individuals in certain motor tasks. Because of the different methodologies and experimental setups in these studies, however, it is difficult to draw general conclusions as to the reasons for this improved performance. In this study, we propose an open-source experimental system for the systematic study of the effect of human-human interaction, as mediated by robots, at the ankle joint. We also propose a new framework to study various interactive behaviors (i.e., collaborative, cooperative, and competitive tasks) that can be emulated using a virtual spring connecting human pairs. To validate the proposed experimental framework, we perform a transparency analysis, which is closely related to haptic rendering performance. We compare muscle EMG and ankle motion data while subjects are barefoot, attached to the unpowered robot, and attached to the powered robot implementing transparency control. We also validate the performance in rendering virtual springs covering a range of stiffness values (5–50 Nm/rad) while the subjects track several desired trajectories (sine waves at frequencies between 0.1 and 1.1 Hz). Finally, we demonstrate the feasibility of the system in studying human-human interaction under different interactive behaviors.

Robot rehabilitation is now recognized as an important method for the efficient recovery. In European Project FP7 BioMot, we have discussed the potential of the robot rehabilitation and proposed the suitable process for it. In this paper, we describe the proposed rehabilitation process and create the theoretical basis for the robot rehabilitation through designing the control system and the patient model. To design the patient model, we describe the source of paralysis and motion controller separately and define the recovery function from the paralysis. In the theoretical analysis of the control system, we show that the robot motions are first adapted to the patient abnormal motions and gradually drive the patient motions to the better ones by the motion support. The singular perturbation analysis proves that the stabilities of the two different process, adaptation to the patient motions and the motion support to the better ones, as a slow motion subsystem and a fast motion subsystem, respectively. The simulation results show that the proposed control system can drive the patients to the better state depending on the patient conditions such as recovery speed and recovery potential. The proposed system can be tuned to fit to the variety of the real patient conditions when we apply it to the real applications., This study was funded by a grant from the European Commission within its Seventh Framework Programme (IFP7-ICT-2013-10-611695: BioMot – Smart Wearable Robots with Bioinspired Sensory-Motor Skills).

Interest in the investigation of novel software and control algorithms for wearable robotics is growing. However, entry into this field requires a significant investment in a testing platform. This work introduces CANopen Robot Controller (CORC)—an open source software stack designed to accelerate the development of robot software and control algorithms—justifying its choice of platform, describing its overall structure, and demonstrating its viability on two distinct platforms.

A flexible simulation platform for human gait with exoskeletons is presented to support the analysis of the human-robot interaction and performance assessment of control strategies design, e.g., assist-as-needed. The platform includes the dynamic modeling of the lower limb’s exoskeleton-human with six degrees of freedom in sagittal plane, actuated by servomotors, where the movement equations are explicitly obtained through the Euler-Lagrange approach. The full model is hybrid and it contains four sub-models that are switched between themselves according to the gait cycle phases; also, it comprises the joint actuators dynamics and the low level control system, this modeling was assembled in Simulink-Matlab. It is shown that the proposed platform is a useful tool for assistance control strategies in the gait rehabilitation. Furthermore, simulation results for the Exo-H3 exoskeleton show the successful tracking of angular trajectories, the level of human participation, and the realistic torques.

Objectives. This paper aims to investigate the feasibility and the validity of applying deep convolutional neural networks (CNN) to identify motor unit (MU) spike trains and estimate the neural drive to muscles from high-density electromyography (HD-EMG) signals in real time. Two distinct deep CNNs are compared with the convolution kernel compensation (CKC) algorithm using simulated and experimentally recorded signals. The effects of window size and step size of the input HD-EMG signals are also investigated. Approach. The MU spike trains were first identified with the CKC algorithm. The HD-EMG signals and spike trains were used to train the deep CNN. Then, the deep CNN decomposed the HD-EMG signals into MU discharge times in real time. Two CNN approaches are compared with the CKC: (a) multiple single-output deep CNN (SO-DCNN) with one MU decomposed per network, and (b) one multiple-output deep CNN (MO-DCNN) to decompose all MUs (up to 23) with one network. Main results. The MO-DCNN outperformed the SO-DCNN in terms of training time (3.2–21.4 s epoch−1 vs 6.5–47.8 s epoch−1, respectively) and prediction time (0.04 vs 0.27 s sample−1, respectively). The optimal window size and step size for MO-DCNN were 120 and 20 data points, respectively. It results in sensitivity of 98% and 85% with simulated and experimentally recorded HD-EMG signals, respectively. There is a high cross-correlation coefficient between the neural drive estimated with CKC and that estimated with MO-DCNN (range of r-value across conditions: 0.88–0.95). Significance. We demonstrate the feasibility and the validity of using deep CNN to accurately identify MU activity from HD-EMG with a latency lower than 80 ms, which falls within the lower bound of the human electromechanical delay. This method opens many opportunities for using the neural drive to interface humans with assistive devices.

The hypothesis of modular control, which stands on the existence of muscle synergies as building blocks of muscle coordination, has been investigated in a great variety of motor tasks and species. Yet, its role during learning processes is still largely unexplored. To what extent is such modular control flexible, in terms of spatial structure and temporal activation, to externally or internally induced adaptations, is a debated issue. To address this question, we designed a biofeedback experiment to induce changes in the timing of muscle activations during leg cycling movements. The protocol consisted in delaying the peak of activation of one target muscle and using its electromyography (EMG) envelope as visual biofeedback. For each of the 10 healthy participants, the protocol was repeated for three different target muscles: Tibialis Anterioris (TA), Gastrocnemius Medialis (GM), and Vastus Lateralis (VL). To explore the effects of the conditioning protocol, we analyzed changes in the activity of eight lower limb muscles by applying different models of modular motor control [i.e., fixed spatial components (FSC) and fixed temporal components (FTC)]. Our results confirm the hypothesis that visual EMG biofeedback is able to induce changes in muscle coordination. Subjects were able to shift the peak of activation of the target muscle, with a delay of (49 ± 27°) across subjects and conditions. This time shift generated a reorganization of all the other muscles in terms of timing and amplitude. By using different models of modular motor control, we demonstrated that neither spatially invariant nor temporally invariant muscle synergies alone were able to account for these changes in muscle coordination after learning, while temporally invariant muscle synergies with adjustments in timing could capture most of muscle activity adaptations observed after the conditioning protocol. These results suggest that short-term learning in rhythmic tasks is built upon synergistic temporal commands that are robust to changes in the task demands., This research has been supported by Spanish project ASSOCIATE “A comprehensive and wearable robotics based approach to the rehabilitation and assistance of people with stroke and SCI” (DPI2014-58431-C4-1-R), and the H2020 project EUROBENCH “European Robotic Framework for Bipedal Locomotion Benchmarking” (Grant Agreement No. 779963).

The human being in his blessed curiosity has always wondered how to make machines feel, and, at the same time how a machine can detect emotions. Perhaps some of the tasks that cannot be replaced by machines are the ability of human beings to feel emotions. In the last year, this hypothesis is increasingly questioned by scientists who have done work that seeks to understand the phenomena of brain functioning using the state of the art in instrumentation, sensors, and signal processing. Today, the world scientists have powerful machine learning methods developed to challenge this issue.The field of emotion detection is gaining significance as the technology advances, and particularly due to the current developments in machine learning, the Internet of Things, industry 4.0 and Autonomous Vehicles. Machines will need to be equipped with the capacity to monitor the state of the human user and to change their behaviour in response. Machine learning offers a route to this and should be able to make use of data collected from questionnaires, facial expression scans, and physiological signals such as electroencephalograms (EEG), electrocardiograms, and galvanic skin response. In this study, an approach was proposed to identify the emotional state of a subject from the collected data in the elicited emotion experiments. An algorithm using EEG data was developed, using the power spectral density of the frequency cerebral bands (alpha, beta, theta, and gamma) as features for classifier training. A K Nearest Neighbors algorithm using Euclidian distance was used to predict the emotional state of the subject. This article proposes a novel approach for emotion recognition that not only depends on images of the face, as in the previous literature, but also on the physiological data. The algorithm was able to recognize nine different emotions (Neutral, Anger, Disgust, Fear, Joy, Sadness, Surprise, Amusement, and Anxiety), nine valence positions, and nine positions on arousal axes. Using the data from only 14 EEG electrodes, an accuracy of approximately 97% was achieved. An approach has been developed for evaluating the state of mind of an driver in the context of a semi-autonomous vehicle context, for example. However, the system has a much wider range of potential applications, from the design of products to the evaluation of the user experience.

In order to prepare humanoid robots for real-world applications, it is necessary to assess the robots’ locomotion performance and to determine their readiness level. We have carried out an extensive statistical literature review of 204 publications evaluating what kind of benchmarking scenarios and performance indicators are used and if they are sufficient to take a robot from prototype to production. We report the results of our analysis and discuss the most relevant findings such as the overall increase in the number of publications and motion tasks considered. Previous benchmarking efforts have been devoted to the functional assessment of the robotic system itself. A trend towards goal level oriented performance indicators such as robustness to external disturbance and dependability can be observed. We also identify a deficiency in benchmarking of non-functional aspects like safety or human-robot interaction. Based on our results, we outline the creation of a benchmarking framework with respect to current benchmarking approaches also introducing testbeds considering currently neglected motion tasks with a focus on a high degree of standardization.

Comunicació presentada a: International Conference on NeuroRehabilitation (ICNR) celebrat del 13 al 16 d'octubre de 2020 de manera virtual. Wireless active implantable medical devices (AIMDs) can be an alternative for overcoming the drawbacks faced with superficial and percutaneous technologies. However, current AIMDs require bulky and rigid components for powering, hampering their miniaturization. AIMDs based on power transfer by volume conduction do not need these voluminous parts, allowing the development of thread-like devices that could be used for distributed stimulation and sensing of the neuromuscular system. In this paper, we present an in vitro evaluation of a protocol and an architecture for bidirectional communications in networks of injectable wireless implants powered and controlled by volume conduction. The wireless prototypes were successfully addressed from the external systems, and end-to-end bidirectional communication was performed at 256 kbps with a success rate of 87%. This work was funded by the European Union’s Horizon 2020 Research and Innovation Programme (Project EXTEND—Bidirectional HyperConnected Neural System) under grant agreement No 779982. Antoni Ivorra gratefully acknowledges the financial support by ICREA under the ICREA Academia programme.

Wearable robots, such as exoskeletons and soft exosuits, can augment normal function or serve as prostheses for missing limbs. In both cases, they extend, complement, substitute, or enhance human functions and capability and can empower or replace human limbs. Cognitive and physical interactions between human and robot are key for these robots to seamlessly deliver assistance when required. The physical interaction between a robot and its wearer generates forces to overcome the wearer's physical limits, and cognitive interactions allow the wearer to guide and control the robot at all times. On page 668 of this issue, Kim et al. (1) report on a soft exosuit that switches assistance profiles for different physical interaction strategies—in this case, walking versus running—through a versatile cognitive interaction in which algorithms accurately determine and detect the wearer's gait.

Exoskeleton technology has made significant advances during the last decade, resulting in a considerable variety of solutions for gait assistance and rehabilitation. The mechanical design of these devices is a crucial aspect that affects the efficiency and effectiveness of their interaction with the user. Recent developments have pointed towards compliant mechanisms and structures, due to their promising potential in terms of adaptability, safety, efficiency, and comfort. However, there still remain challenges to be solved before compliant lower limb exoskeletons can be deployed in real scenarios. In this review, we analysed 52 lower limb wearable exoskeletons, focusing on three main aspects of compliance: actuation, structure, and interface attachment components. We highlighted the drawbacks and advantages of the different solutions, and suggested a number of promising research lines. We also created and made available a set of data sheets that contain the technical characteristics of the reviewed devices, with the aim of providing researchers and end-users with an updated overview on the existing solutions., This work is supported by the project EUROBENCH (European ROBotic framework for bipedal locomotion bENCHmarking) funded by H2020 Topic ICT 27–2017: “System abilities, SME & benchmarking actions, safety certification” (reference: 779963). We acknowledge support of the publication fee by the CSIC Open Access Publication Support Initiative through its Unit of Information Resources for Research (URICI).

Walking rehabilitation processes include many repetitions of the same physical movements in order to replicate, as close as possible, the normal gait trajectories, and kinematics of all leg joints. In these conventional therapies, the therapist's ability to discover patient's limitations-and gradually reduce them-is key to the success of the therapy. Lower-limb robotic exoskeletons have strong deficiencies in this respect as compared to an experienced therapist. Most of the currently available robotic solutions are not able to properly adapt their trajectories to the biomechanical limitations of the patient. With this in mind, much research and development is still required in order to improve artificial human-like walking patterns sufficiently for valuable clinical use. The work herein reported develops and presents a method to acquire and saliently analyze subject-specific gait data while the subject dons a passive lower-limb exoskeleton. Furthermore, the method can generate adjustable, yet subject-specific, kinematic gait trajectories useful in programming controllers for future robotic rehabilitation protocols. A human-user study with ten healthy subjects provides the experimental setup to validate the proposed method. The experimental protocol consists in capturing kinematic data while subjects walk, with the donned H2 lower-limb exoskeleton, across three experimental conditions: walking with three different pre-determined step lengths marked on a lane. The captured ankle trajectories in the sagittal plane were found by normalizing all trials of each test from one heel strike to the next heel strike independent of the specific gait features of each individual. Prior literature suggests analyzing gait in phases. A preliminary data analysis, however, suggests that there exist six key events of the gait cycle, events that can adequately characterize gait for the purposes required of robotic rehabilitation including gait analysis and reference trajectory generation. Defining the ankle as an end effector and the hip as the origin of the coordinate frame and basing the linear regression calculations only on the six key events, i.e., Heel Strike, Toe Off, Pre-Swing, Initial Swing, Mid-Swing, and Terminal Swing, it is possible to generate a new calculated ankle trajectory with an arbitrary step length. The Leave-One-Out Cross Validation algorithm was used to estimate the fitting error of the calculated trajectory vs. the characteristic captured trajectory per subject, showing a fidelity average value of 95.2, 96.1, and 97.2%, respectively, for each step-length trial including all subjects. This research presents method to capture ankle trajectories from subjects and generate human-like ankle trajectories that could be scaled and computed on-line, could be adjusted to different gait scenarios, and could be used not only to generate reference trajectories for gait controllers, but also as an accurate and salient benchmark to test the human likeness of gait trajectories employed by existing robotic exoskeletal devices., The research was partially funded by CONACyT Doctoral program, Tecnolgico de Monterrey/Robotics Lab and research chair program for distinguished professors in Biomechatronics.

Robots are entering our everyday life at an exponential pace. Benchmarking can help researchers and developers to improve their systems, as well as providing end-user with easy-to-understand “performance scores” able to identify the best solution for their needs. Unfortunately, a benchmarking methodology for robotics is still not available. The EUROBENCH project wants to provide rigorous tools at both software and hardware level, to allow companies, researchers and users to test robotic systems under multiple facets. The project, initially focused on bipedal robotics technologies (exoskeletons, prostheses, humanoids), will offer funding opportunities to third parties to participate in the development and validation of the different components of the framework.

The difficulty of defining standard benchmarks for human likeness is a well-know problem in bipedal robotics. This chapter reviews methods and criteria for the assessment of the sensorimotor mechanisms involved in human walking and posture. We focused on the potential of the reviewed methods to be used as benchmarks for human-like locomotion of bipedal robots. For walking conditions, several criteria and methods related to dynamic similarity, passivity and dynamicity, static stability, and energy consumption have been identified. As for standing functions, we identified the most relevant features characterizing the human postural sensorimotor mechanisms, and presented the experimental protocols currently used to evaluate the human-like robotic performance. Furthermore, we discussed how the current robotic competitions such as RoboCup and DARPA Robotics Challenges can contribute to the identification of relevant benchmarks. Finally, we speculated about the importance of international consensus on the quantitative definition of human likeness, and suggested some future actions for improving collaboration and standardization within the scientific community., This research activity has been founded by the European Seventh Framework Programme FP7-ICT-2011-9, under the grant agreement no 60069 - H2R “Integrative Approach for the Emergence of Human-like Robot Locomotion”.

This paper reports a preliminary overview on the existing metrics and benchmarks for the assessment of lower limb wearable robotics performance. This analysis was conducted to identify the current necessities, preferences, and deficiencies on this topic within the scientific robotics community. Based on these results, we aim to establish a robust set design criteria of a comprehensive benchmarking scheme for wearable robotic technology considering real life scenarios.

Robotic exoskeletons that induce leg movement have proven effective for lower body rehabilitation, but current solutions offer limited gait patterns, lack stabilization, and do not properly stimulate the proprioceptive and balance systems (since the patient remains in place). Partial body weight support (PBWS) systems unload part of the patient’s body weight during rehabilitation, improving the locomotive capabilities and minimizing the muscular effort. HYBRID is a complete system that combines a 6DoF lower body exoskeleton (H1) with a PBWS system (REMOVI) to produce a solution apt for clinical practice that offers improves on existing devices, moves with the patient, offers a gait cycle extracted from the kinematic analysis of healthy users, records the session data, and can easily transfer the patient from a wheelchair to standing position. This system was developed with input from therapists, and its response times have been measured to ensure it works swiftly and without a perceptible delay.

Physical interfaces have an important role in achieving efficient, safe and comfortable transmission of forces between the exoskeleton and the human body. They are normally composed of different compliant elements disposed in series between the skin of the user and the exoskeleton frame. Modelling how the compliant properties of physical interface will affect the transmission of forces may be useful to improve the design process towards more effective, safe and user-specific exoskeletal devices. As a first step in this direction, we propose a simplified 2-dimensional model representing the interaction of a single-actuated-joint exoskeleton with the human limb through a compliant element. We studied the effects of stiffness value associated to the tissues in the behavior of the whole system with simulated results of the model.

The relative motion between human and exoskeleton is a crucial factor that has remarkable consequences on the efficiency, reliability and safety of human-robot interaction. Unfortunately, its quantitative assessment has been largely overlooked in the literature. Here, we present a methodology that allows to predict the motion of the human joints from the knowledge of the angular motion of the exoskeleton frame. Our method combines a subject-specific skeletal model with a kinematic model of a lower limb exoskeleton (H2, Technaid), imposing specific kinematic constraints between them. To calibrate the model and validate its ability to predict the relative motion in a subject-specific way, we performed experiments on seven healthy subjects during treadmill walking tasks. We demonstrate a predictions accuracy lower than 3.5 degrees globally, and around 1.5 degrees at the hip level, which represent an improvement up to 66% compared to the traditional approach assuming no relative motion between the user and the exoskeleton.

Rehabilitation of walking ability is one of the most important objectives after a spinal cord injury. Robotic and neuroprosthetic technologies hold a considerable potential for driving walking rehabilitation therapies. However, new developments are needed in order to improve the walking rehabilitation interventions based in these technologies.We recently presented a cooperative control strategy of Kinesis, a lower limb exoskeleton for providing hybrid therapy of walking (Del-Ama, 2014). Its design aimed to actively manage muscle fatigue caused by surface electrical stimulation, and to implement the assist-as-needed control paradigm in which both stimulation and robotic controller cooperate with the residual functionality of the user. In this article we present three case studies for investigating the feasibility of the hybrid therapy of walking delivered with Kinesis in patients with incomplete spinal cord injury. Besides, the adaptability features of Kinesis stimulation-robot cooperative control are assessed, characterizing the behavior of the cooperative controller while providing hybrid therapy of walking.Patients with incomplete spinal cord injury participated in the experiments. The protocol consisted of walking with Kinesis during 6?min. Three configurations of the cooperative controller were tested for each patient in separate sessions in order to investigate its adaptability features. The immediate impact of the hybrid therapy of walking was assessed through several variables that represent the physiological impact, user-exoskeleton physical interaction, stimulation intensity and user subjective perception of the hybrid therapy of walking.Results show that the cooperative controller of Kinesis adapted to patient functional deficits and voluntary actions during walking, modulating stimulation and robotic assistance, which was the aim of the controller design. Nevertheless, no noticeable differences were observed in the comparison between compliant and trajectory exoskeleton control. Further work is envisioned regarding several aspects of hybrid walking control: stimulation control based on muscle activation estimate, improved semi-automatic control of walking, and improved muscle fatigue monitoring. The hybrid walking therapy was tolerated by the patients without adverse effects, along with a tolerable physical demand. This shows a potential for walking rehabilitation in motor incomplete SCI patients, guaranteeing further research on this topic. We investigate the feasibility of providing the hybrid therapy of walking with Kinesis hybrid exoskeleton.We compare and investigate the role of the stimulation and the compliant robot control.Three case studies are presented. The cooperative controller of Kinesis adapted to patients and voluntary actions.No noticeable differences were observed between compliant and trajectory control.

Background Gait training for individuals with neurological disorders is challenging in providing the suitable assistance and more adaptive behaviour towards user needs. The user specific adaptation can be defined based on the user interaction with the orthosis and by monitoring the user intentions. In this paper, an adaptive control model, commanded by the user intention, is evaluated using a lower limb exoskeleton with incomplete spinal cord injury individuals (SCI). Methods A user intention based adaptive control model has been developed and evaluated with 4 incomplete SCI individuals across 3 sessions of training per individual. The adaptive control model modifies the joint impedance properties of the exoskeleton as a function of the human-orthosis interaction torques and the joint trajectory evolution along the gait sequence, in real time. The volitional input of the user is identified by monitoring the neural signals, pertaining to the user’s motor activity. These volitional inputs are used as a trigger to initiate the gait movement, allowing the user to control the initialization of the exoskeleton movement, independently. A Finite-state machine based control model is used in this set-up which helps in combining the volitional orders with the gait adaptation. Results The exoskeleton demonstrated an adaptive assistance depending on the patients’ performance without guiding them to follow an imposed trajectory. The exoskeleton initiated the trajectory based on the user intention command received from the brain machine interface, demonstrating it as a reliable trigger. The exoskeleton maintained the equilibrium by providing suitable assistance throughout the experiments. A progressive change in the maximum flexion of the knee joint was observed at the end of each session which shows improvement in the patient performance. Results of the adaptive impedance were evaluated by comparing with the application of a constant impedance value. Participants reported that the movement of the exoskeleton was flexible and the walking patterns were similar to their own distinct patterns. Conclusions This study demonstrates that user specific adaptive control can be applied on a wearable robot based on the human-orthosis interaction torques and modifying the joints’ impedance properties. The patients perceived no external or impulsive force and felt comfortable with the assistance provided by the exoskeleton. The main goal of such a user dependent control is to assist the patients’ needs and adapt to their characteristics, thus maximizing their engagement in the therapy and avoiding slacking. In addition, the initiation directly controlled by the brain allows synchronizing the user’s intention with the afferent stimulus provided by the movement of the exoskeleton, which maximizes the potentiality of the system in neuro-rehabilitative therapies.

Powered lower-limb exoskeletons are emerging as a suitable robot-aided therapy for the human-impaired locomotion. These devices have been designed to safely restore the human motor skills through active muscle activation delivered throughout task-oriented and repetitive gait therapy. In this chapter, we revise the current status and the clinical effectiveness, safety, and usability exhibited by the most advanced wearable exoskeletons proposed for the gait rehabilitation. We also give an overview of the traditional actuation technologies applied in wearable exoskeletons and the alternative actuation mechanisms that are being applied based on compliant actuation or lightweight mechanisms. Moreover, we describe the promising directions with regard to the control strategies to guide the patient according to his needs and encourage his active participation in gait therapy. Novel horizons involving the human-exoskeleton symbiotic interaction via interfaces with neural structures are also introduced.

Recent projects highlight how motor learning and a high level of attention control can potentially improve submaximal force production during recovery after stroke. This study focuses on the assessment of detailed metrics of force production and position control -healthy subjects- and their correlation with submaximal force production control learning during a new task consisting in maintaining the position for early rehabilitation after stroke.

The motor system may rely on a modular organization (muscle synergies activated in time) to execute different tasks. We investigated the common control features of walking and cycling in healthy humans from the perspective of muscle synergies. Three hypotheses were tested: 1) muscle synergies extracted from walking trials are similar to those extracted during cycling; 2) muscle synergies extracted from one of these motor tasks can be used to mathematically reconstruct the electromyographic (EMG) patterns of the other task; 3) muscle synergies of cycling can result from merging synergies of walking. A secondary objective was to identify the speed (and cadence) at which higher similarities emerged. EMG activity from eight muscles of the dominant leg was recorded in eight healthy subjects during walking and cycling at four matched cadences. A factorization technique [nonnegative matrix factorization (NNMF)] was applied to extract individual muscle synergy vectors and the respective activation coefficients behind the global muscular activity of each condition. Results corroborated hypotheses 2 and 3, showing that 1) four synergies from walking and cycling can successfully explain most of the EMG variability of cycling and walking, respectively, and 2) two of four synergies from walking appear to merge together to reconstruct one individual synergy of cycling, with best reconstruction values found for higher speeds. Direct comparison of the muscle synergy vectors of walking and the muscle synergy vectors of cycling ( hypothesis 1) produced moderated values of similarity. This study provides supporting evidence for the hypothesis that cycling and walking share common neuromuscular mechanisms.

In this paper a method to detect velocity-dependant spasticity is shown. The implementation and test were performed using the ankle joints of a six degrees of freedom exoskeleton. In this first stage of the project all subjects are healthy and have no difficulty in walking. The subjects left the ankle joint relaxed while sat down and did not rest the heel on the floor. Force sensors are used in order to measure the interaction forces between the exoskeleton and de human limb. The sensors were located in the foot sole under the subject’s forefoot and in the instep brace. A variable oscillator is used to generate the angle reference signal and to vary the frequency thereof. The presence of spasticity was detected based on the interaction force information and the joint movement was decreased in speed.

Synthesizing human movement in computational neuro-mechanical models is a complex problem. In this study, we describe a predictive framework that combines computational models of neural, sensory and mechanical processes. Our proposed formulation facilitates the transition towards the design of a new class of bio-mimetic assistive devices that do not rely on explicit representations of task-specific motor control models.

Coupling motor intentions decoded from cortical activities with coherent proprioceptive feedback is of interest for the motor rehabilitation of neurological patients with lesions in the central nervous system. For these interventions to be effective, repeated sessions need to be carried out to achieve functional long-lasting plastic changes of cortical circuits. Electroencephalography-based Brain-Computer Interfaces typically show significant decreases in accuracy when used across multiple sessions with fixed parameters. Therefore, it is important to look for optimal strategies to recalibrate these classifiers. Here we compare different recalibration strategies for systems decoding motor intentions based on electroencephalographic data of neurological patients.

In this abstract we present the potentialities of muscle synergy analysis, according to the most recent and debated functional interpretations. We also provide some ideas on the use of the synergy approach beyond the classical descriptive modalities, in order to enhance the biologically inspiration of current neurorehabilitation techniques.

The combined use of the functional electrical stimulation (FES) with robotic devices known as Hybrid Robotic System emerges as a promising solution to improve rehabilitation therapies after neurological injuries (e.g. stroke). This work presents a first step towards the implementation of a Hybrid Robotic System simulation platform that could allow exploring several control strategies to improve its performance. The results show the feasibility of the platform to deploy several control strategies that combines FES and robotic devices.

In this study we explore a way of controlling a lower limb exoskeleton based on the detection of the user intention by recording and classifying information from force sensors placed on both knees and hips. The classifier is based on Linear Discriminant Analysis and has been tested offline in 5 healthy subjects, obtaining an average accuracy of 91.11 % for the sit-to-stand transition, 72.5 % for the stand-to-walk transition and 70 % for the stand-to-sit transition.

© The Authors., Hybrid robotic systems represent a novel research field, where functional electrical stimulation (FES) is combined with a robotic device for rehabilitation of motor impairment. Under this approach, the design of robust FES controllers still remains an open challenge. In this work, we aimed at developing a learning FES controller to assist in the performance of reaching movements in a simple hybrid robotic system setting. We implemented a Feedback Error Learning (FEL) control strategy consisting of a feedback PID controller and a feedforward controller based on a neural network. A passive exoskeleton complemented the FES controller by compensating the effects of gravity. We carried out experiments with healthy subjects to validate the performance of the system. Results show that the FEL control strategy is able to adjust the FES intensity to track the desired trajectory accurately without the need of a previous mathematical model., This work has been done with the financial support of the Ministry of Science and Innovation of Spain, project HYPER (CSD 2009-00067 Hybrid Neuroprosthetic and Neurorobotic Devices for Functional Compensation and Rehabilitation of Motor Disorders).

In this research, we describe an actuation and control system designed for geared electromagnetic motors, which is characterized by its simple implementation, fast response to external input loads, reliable human-machine interaction features, no need for previous set-up or calibration from user to user and high portability due to the reduction of weight and space used. By the implementation of the proposed system, an electromagnetic motor can become a multitasking, wearable actuation system capable of: detecting the user's intentions regarding motion, tracking the limbs with minimal force interaction within a wide bandwidth and also providing controllable assistance and resistance forces to the user's movements, without the use of any biological signal. Validation of the proposed approach is shown by the construction of a powered orthosis for the knee, used to test the system's performance under real human motion conditions. The proposed system was tested on one healthy subject by measuring electromyographic levels both with and without the orthosis, under controlled flexion and extension cycles. Experimental results demonstrate the effectiveness of the proposed approach in detecting the user's intentions regarding motion, reducing and increasing muscular activity when configured for assistance and resistance, respectively, and also increasing the transparency of the actuation system when perfect tracking of the limbs is needed.

[ES] En este trabajo se presentan en primer lugar unas consideraciones generales sobre la aplicación de la robótica en un sentido amplio a funciones de asistencia personal atendiendo de un modo particular a las necesidades de las personas con discapacidad física o sensorial y cognitiva. El núcleo del trabajo consiste en la exposición de la configuración de una serie de sensores ultrasónicos para la medida de la distancia relativa entre los pies del usuario y el andador y a partir de esta información obtener parámetros descritivos de la marcha humana. Estos parámetros han sido utilizados para la elaboración de una formulación analítica de la evolución de los pies presentado también en este trabajo. El sistema ultrasónico está integrado en un andador avanzado que consiste en una plataforma multisensorial para el estudio de la interacción y cooperación hombre-máquina., [EN] In this work, general considerations are placed regarding the application of robotics in personal assistance aiming in particular the necessities of people with physical, sensory or cognitive impairments. The main focus of this work is the configuration of a series of ultrasonic transducers for the measurement of the relative position between the user’s feet and a robotic walker during gait. From such measurements, descriptive parameters of the human gait are obtained. Such parameters were used to elaborate a novel analytical formulation describing the feet evolution in time. Such ultrasonic system is integrated in an advanced walker that consists of a multisensory platform to study the human-machine interaction and cooperation in assisted gait., Este trabajo se ha desarollado en el marco del Proyecto SIMBIOSIS (DPI2005-07417)

Quantification of pathological tremor amplitude and frequency is fundamental for clinical diagnosis of neurological disorders and for robotics based-tremor suppression. In this work, a novel approach for real-time estimation of tremor amplitude and frequency using inertial sensors is presented. First, a miniaturized MEMS sensor is attached to the patient’s hand and calibrated with respect to it. This procedure permits to directly obtain wrist flexion/extension. Afterwards, a two stage algorithm for tremor estimation is implemented. It consists of a tracking algorithm that isolates tremorous movement, followed by an adaptive filter that estimates tremor parameters. The proposed technique has inherent zero phase, providing a tremor amplitude error of 0.3766 rad/s.

The extent to which the electroencephalographic activity allows the characterization of movements with the upper limb is an open question. This paper describes the design and validation of a classifier of upper-limb analytical movements based on electroencephalographic activity extracted from intervals preceding self-initiated movement tasks. Features selected for the classification are subject specific and associated with the movement tasks. Further tests are performed to reject the hypothesis that other information different from the task-related cortical activity is being used by the classifiers. Six healthy subjects were measured performing self-initiated upper-limb analytical movements. A Bayesian classifier was used to classify among seven different kinds of movements. Features considered covered the alpha and beta bands. A genetic algorithm was used to optimally select a subset of features for the classification. An average accuracy of 62.9 ± 7.5 % was reached, which was above the baseline level observed with the proposed methodology (30.2 ± 4.3 %). The study shows how the electroencephalography carries information about the type of analytical movement performed with the upper limb and how it can be decoded before the movement begins. In neurorehabilitation environments, this information could be used for monitoring and assisting purposes., This work has been funded by grant from the Spanish Ministry of Science and Innovation CONSOLIDER INGENIO, project HYPER (Hybrid NeuroProsthetic and NeuroRobotic Devices for Functional Compensation and Rehabilitation of Motor Disorders, CSD2009-00067), from Proyectos Cero of FGCSIC, Obra Social la Caixa, CSIC, and from the project PIE 201050E087.

The modulation of joint stiffness is a crucial mechanism that allows humans, and other vertebrates, to perform stable and efficient locomotion. Compliant actuation is an emerging branch of robotics, which aims to mimic the biological elastic properties of muscle fibers and series-elastic tendon structures. This paper presents the mechanical design and real prototype of a compliant 2-Degrees of Freedom (DoF) ankle-foot system for application in a humanoid biped robot. The proposed design centralizes the two DoFs in one single joint, while allowing independent control of the equilibrium position and stiffness in both DoFs. The presented prototype achieves higher ranges of motion in both sagittal and frontal plane than those expected in normal walking conditions, while at the same time increases the versatility of the ankle due to its compliant properties.

Great progress has been achieved in the last few years in Neurorehabilitation and Neural Engineering. Thanks to the parallel development of medical research, the chances to survive a neural injury are growing, so it is necessary to develop technologies that can be used both for rehabilitation and to improve daily activities and social life. New robotic and prosthetic devices and technologies such as Functional Electrical Stimulation, Brain-Computer Interfaces and Virtual Reality are slowly becoming part of clinical rehabilitation setting, but they are far from being part of everyday life for the patients. As technology improves, expectations keep rising and new problems emerge: from cost reduction to the diffusion in the medical environment, from the enlargement of the number of patients who may benefit from these technologies to transferring rehabilitation to the patient’s home. All these requests lead to great challenges that researchers have to face to enhance their contribution towards the improvement of rehabilitation and life conditions of patients with neural impairment.

Lower limb wearable robots are entering an exciting era. An increasing number of solutions are moving out of the lab, approaching the everyday rehabilitation practice and home-based assistive scenarios. In this context, the quantitative assessment of the technology is crucial for its correct inclusion in the market. Nevertheless, the tool normally used to support this process, i.e. benchmarking, hasn't been formulated yet in the wearable robotics field. Within the European projects H2R and Biomot we are developing a scheme for the definition of benchmarks specifically designed for wearable robot devices. This scheme takes into account two main perspectives, named "functional" and "interaction". The functional perspective aims at evaluating the stability, efficiency and correctness of motion of the global system constituted by the patient wearing the robot. The interaction perspective aims at evaluating the symbiotic interplay between the user and the device, under the physical, cognitive, and psychophysiological standpoints. In addition, we also identify the critical role of the technology in the process of bringing benchmarks into the everyday laboratory practice, with a specific focus on the emerging measurement and perturbing solutions.

A general need for Wearable Robots (WRs) is to perform optimal selection of information for control and monitoring during daily assistance with an autonomous systems. A good feature selection algorithm is key to perform automated estimation of the mode of locomotion under environmental variations. Ambulatory body weight support (BWS) systems can be combined with WRs to provide safe ambulation and support in overground walking in cases of lower limb paralysis. This study aimed to develop a support vector machine (SVM) model for binary and multiclass classification that performs gait pattern recognition for different values of partial BWS during overground robot-aided walking. The principal component analysis (PCA) and kernel-based PCA (kPCA) were applied to improve the classification performance. As a result, the combination of temporal and kinematic features showed to improve the accuracy in the discrimination of gait patterns in healthy patients (88%). In SVM multiclass classification the “one-against-one” approach showed to have a more stable performance (true positive and true negative rate are consistent) than “one-against-all” approach and also lower computational cost both for training and SVM's decision making.

This paper presents the design and implementation of a cost-effective and small inertial measurement unit (IMU) for application on leg smart orthotics and prosthetics. IMU design based on biomechanical considerations for lower leg devices is presented. Methods for calculation of a number of biomechanical parameters related to gait based on the unit are discussed, including calibration and offset correction procedures. An approach for electronic knee joint control of orthosis during cyclic walking based on IMU signals is discussed. Finally, experiments are conducted for a subject walking on a flat surface wearing a mechanically driven orthosis with the proposed sensor. Experimental results demonstrate the IMU suitability and feasibility for real-time embedded control of wearable assistive devices for walking restoration and monitoring.

Miniaturization is not a logical trend in actuator systems. Unlike actuators, sensors intrinsically perform more efficiently upon miniaturization. This is a logical consequence of the exchange of energy in the transduction process when applying sensors: measurement ideally should not influence the system being measured, thus the minimum exchange of energy is necessary and this intrinsically leads to miniaturization. In actuators, a transduction process is likewise established but the aim is to impose a mechanical state on a system. It is of particular interest not having this state influenced by perturbations, thus there are strong requirements on power delivered by the actuator. In view of current trends towards miniaturization, it is worth inquiring how the performance of piezoelectric actuators is affected by reducing their size. We are not concerned here with the domain of micro-actuators, i.e. actuators with sizes in the micrometer range. The analysis in this paper focuses on studying how four useful parameters for describing the performance of actuators are influenced by miniaturization: resonance frequency, force density, response time (bandwidth), stroke and energy density per cycle. In so doing, the analysis is restricted to non resonant piezoelectric actuators, i.e. stack, multimorph and inchworm actuators, but reference to other piezoelectric, emerging and traditional actuators is included for comparison.

Informatics is becoming essential in our daily activities. The use of the human machine interface requires fine movements from the user. When human movements are distorted, for instance, by tremor, performance could be improved by digitally filtering the intermediate signal before it reaches the controlled interface. This paper presents a novel device able to filter out tremorous movements from the cursor of a mouse in the screen of the computer. This task requires real-time discrimination between voluntary and tremorous movement. To address this problem, a learning algorithm is introduced as well as the evaluation and validation with users of the prototype developed.

This paper introduces a comparative analysis of Piezoelectric (PZ) and Magnetostrictive (MS) actuators as components in smart structures. There is an increasing interest in functional structures which are able to adapt to external or internal perturbations, i.e. changes in loading conditions or ageing. Actuator technologies must perform concomitantly as sensors and actuators to be applicable in smart structures. In this paper we will comparatively analyze the possibility of using PZ and MS actuators in smart structures and in so doing their capability to act concomitantly as sensors and of modifying their material characteristics. We will also focus on the analysis of how them can be integrated in structures and on the analysis of the most appropriate structures for each actuator. The operational performance of PZ (Stacks) and MS actuators will be compared and eventually some conclusions will be drawn.

Dextrous artificial hand design and manipulation is an active research topic. A very interesting practical application is the field of upper limb prosthetics. This paper presents the mechanical design and manipulation aspects of the MANUS-HAND project to develop a multifunctional upper limb prosthesis. The kinematics of our design makes use of the so-called underactuated principle and leads to an innovative design that triples the performance of currently existing commercial hand prosthesis. In addition, the thumb design allows its positioning both in flexion and opposition. As a consequence, up to four grasping modes (cylindrical, precision, hook and lateral) are available with just two actuators. The proposed impedance control approach allows the fingers to behave as virtual springs. Given the difficulty of including the user in the control loop, this approach is based on an autonomous coordination and control of the grasp. As a consequence, the requirements on the Human Machine interface are reduced. At the end of the paper, we briefly describe the clinical trials that were set up for evaluation purposes.

Joint stiffness of the exoskeleton robot is one of the most important factors to support bipedal walking. In this paper, we discuss the robot joint stiffness tuning algorithm using the bio-mimetic learning method called tacit learning. We experimentally showed that the proposed controller can tune the joint stiffness of the exoskeleton robot by tuning the integral gain in the controller. The walking experiment wearing the exoskeleton robot suggest that the stiffness tuning is applicable to control the walking speed.

Nowadays wearable rehabilitation exoskeletons are one of the most studied gait rehabilitation tool from a technological point of view. Current devices use prerecorded healthy gait patterns. This leads to potentially non-natural imposed gait patterns, and to solve this issue, we propose the use of regression-based methods to reconstruct speed dependent angular trajectories. Results suggest that the proposed method can lead to a more natural gait. Consequently, a naive user may more easily learn to walk under the presence of a robot guidance.

Wearable robots are expected to expand the use of robotics in rehabilitation since they can widen the assistance application context. An important aspect of a rehabilitation therapy, in terms of lower extremity assistance, is balance control. In this article, we propose and evaluate an adaptive control strategy for robotic rehabilitation therapies to guarantee static stability using a wearable robot. Postural balance control can be implemented either acting on the hip, on the ankle joint or on both, depending on the kind of perturbation acting on the subject: internal or external. Internal perturbations can be produced by any voluntary movement of the body, such as bending the trunk. External perturbations, in the form of an impact force, are applied by the exoskeleton without any prior notice to observe the proactive response of the subject. We have used a 6 degree of freedom planar lower limb exoskeleton, H1, to perform this analysis. The developed control strategy has been designed to provide the necessary assistance, related to balance recovery and postural stability, under the “Assist-as-needed” paradigm. The interaction forces between orthosis and subject are monitored, as they play a relevant role in the definition of assistive and resistive movements to be applied to the joints. The proposed method has been tested with 5 healthy subjects in presence of internal and external disturbances. The results demonstrate that knowing the stability limit of each subject, in combination with a therapeutically selected scaling factor, the proposed adaptive control helps in providing an effective assistance in therapy. This method is efficient in handling the individual and combined effect of external perturbations acting on any joint movements., This work was supported by the project HYPER (Hybrid Neuroprosthetic and Neurorobotic devices for functional compensation and rehabilitation of motor disorders), Grant CSD2009-00067 CONSOLIDER INGENIO 2010 from MINECO (Spanish Ministry for Science and Education).

This paper presents an analysis of the main sources of error in a 3D-positioning system using ultrasonic waves, coming to different technical improvements. We suggest a new processing algorithm that will overcome the main sources of error encountered in practice. Comparing with existing processing methods, the proposed technique shows an error reduction by a factor of 20, making the system especially robust against outliers measurements.

Upper limb tremor is the most prevalent movement disorder and, unfortunately, it is not effectively managed in a large proportion of the patients. Neuroprostheses that stimulate the sensorimotor pathways are one of the most promising alternatives although they are still under development. To enrich the interpretation of data recorded during long-term tremor monitoring and to increase the intelligence of tremor suppression neuroprostheses we need to be aware of the context. Context awareness is a major challenge for neuroprostheses and would allow these devices to react more quickly and appropriately to the changing demands of the user and/or task. Traditionally kinematic features are used to extract context information, with most recently the use of joint angles as highly potential features. In this paper we present two algorithms that enable the robust extraction of joint angle and related features to enable long-term continuous monitoring of tremor with context awareness. First, we describe a novel relative sensor placement identification technique based on orientation data. We focus on relative rather than absolute sensor location, because in many medical applications magnetic and inertial measurement units (MIMU) are used in a chain stretching over adjacent segments, or are always placed on a fixed set of locations. Subsequently we demonstrate how tremor parameters can be extracted from orientation data using an adaptive estimation algorithm. Relative sensor location was detected with an accuracy of 94.12% for the 4 MIMU configuration, and 100% for the 3 MIMU configurations. Kinematic tracking error values with an average deviation of 8% demonstrate our ability to estimate tremor from orientation data. The methods presented in this study constitute an important step toward more user-friendly and context-aware neuroprostheses for tremor suppression and monitoring. © 2014 Lambrecht, Gallego, Rocon and Pons., This work has been funded by the European project NeuroTremor(ICT-2011.5.1-287739)andt he Spanish Consolider project HYPER (CSD2009-00067).