In this paper, we introduce our attempt to develop an assistive robot system which can contribute Brain-Machine Interface (BMI) rehabilitation. For the BMI rehabilitation, construct a Electroencephalogram(EEG)-Exoskeleton system, where exoskeleton is connected EEG so that users control by using their brain activities. We use classification method considers covariance matrices of measured signals as inputs decode The decoded activities are used movements. study, consider assisting stand-up...

In this study, we propose a human movement model both for myoelectric assistive robot control and biosignal-sensor-failure detection. We particularly consider an application to upper extremity exoskeleton control. When using electromyography (EMG)-based control, EMG electrodes can be easily disconnected or detached from skin surfaces because the body is always in contact with robot. If multiple are used estimate joint movements, probability of sensor electrode misplacement increases due...

Abstract Sports trainers often grasp and move trainees’ limbs to give instructions on desired movements, a merit of this passive training is the transferring via proprioceptive information. However, it remains unclear how affects system improves learning. This study examined changes in acuity due understand underlying mechanisms upper extremity training. Participants passively learned trajectory elbow-joint movement as per single-arm exoskeleton robot, performance target were assessed before...

This study proposes the design of electromyography (EMG)-based force feedback controller which explicitly considers human-robot interaction for exoskeletal assistive robot. Conventional approaches have been only consider one-directional mapping from EMG to control input robot control. However, and generated by interfere each other, e.g., amplitude decreases if limb movements are assisted In our proposed method, we first derive nonlinear signal muscle estimating human joint torque, convert it...

We propose to tackle in this paper the problem of controlling whole-body humanoid robot behavior through non-invasive brain-machine interfacing (BMI), motivated by perspective mapping human motor control strategies human-like mechanical avatar. Our solution is based on adequate reduction controllable dimensionality a high-DOF motion line with state-of-the-art possibilities BMI technologies, leaving complement subspace part be planned and executed an autonomous planning framework. The results...

In this paper, we introduce our newly developed biosignal-based vertical weight support system that is composed of pneumatic artificial muscles (PAMs) and an electromyography (EMG) measurement device. By using system, assist force can be varied based on measured muscle activities; most existing systems only generate constant forces. estimated knee ankle joint torques from EMGs floating base inverse dynamics. Knee are converted to forces by the kinematic model a subject. The used as inputs...

In this paper, we propose an estimation method of human joint movements from measured EMG signals for assistive robot control. We focus on how to estimate using multiple electrodes even under sensor failure situations. real world applications, might become disconnected or detached skin surfaces. If consider EMG-based control robots, such failures lead significant errors in the user movements. To cope with these failures, a state model that takes uncertain observations into account. Sensor...

Exoskeleton robots need to always actively assist the user's movements otherwise robot just becomes a heavy load for user. However, estimating diversified movement intentions in daily life is not easy and no algorithm so far has achieved that level of estimation. In this study, we rather focus on assisting limited number selected by using an EMG-based classification newly developed lightweight exoskeleton robot. Our knee composed carbon fiber frame highly backdrivable joint driven pneumatic...

Abstract A physical trainer often physically guides a learner’s limbs to teach an ideal movement, giving the learner proprioceptive information about movement be reproduced later. This instruction requires perceive kinesthetic and store instructed temporarily. Therefore, (1) acuity accurately taught kinesthetics (2) short-term memory perceived are two critical functions for reproducing movement. While importance of has been suggested active motor learning, little is known passive learning....

This paper proposes a fault tolerant framework for biosignal-based robot control with multiple sensor electrodes. In this approach, to cope faults, reliable joint torque estimation model is selected from group of models based on failure classifiers. The correlation among the electromyography (EMG) signal streams used as input feature vectors detection. To validate our proposed method, we artificially disconnect an EMG electrode or detach one side probe skin surface during elbow-joint...

Since exoskeletons show potential for rehabilitation therapy, many scientists have been designing upper extremity exoskeletons. Unfortunately, few successfully provided a shoulder exoskeleton severe impairment. Toward Brain-Machine-Interface (BMI) robot therapies impairment, this paper introduces with modular joint and an off-board actuator. We applied Modular Exoskeletal Joint (MEJ) to that was driven by Pneumatic Artificial Muscles (PAMs) transmitted Bowden cable. Our objective is...

In this study, we propose a novel human-in-the-loop optimization approach for exoskeleton robot control. We develop method to optimize widely-used Electromyography (EMG)-based assistive strategies. If use multiple EMG channels control multi-DoF robots, process becomes complex and requires large amount of data. To make the tractable, exploit synergies both human muscles artificial robots reduce number parameters show that can extract not only from user's muscle activities but pneumatic (PAMs)...

In assistive control strategies, we must estimate the user's movement intentions. previous studies, such intended motions were inferred by linearly converting muscle activities to joint torques of an robot or classifying identify most likely from pre-designed motion classes. However, performances these approaches are limited in terms accuracy and flexibility. this study, propose optimal strategy that uses estimated user intentions as terminal cost function not only for generating movements...

Many elderly people suffer from declined motor ability, and it is important to develop an assistive device support their daily life. Some have been developed previously help human sit-to-stand motion, but they generally body extension slowly thus the user of could not utilize own muscles. Here, we a chair type assist which lift up seat push hip joint for supporting rising extension. Our employed zip chain actuator realizes fast strong follow user's movement provide assistance. Evaluation...

With the recent advance in robot technologies and aging of society, a variety sit-to-stand (StS) assisting systems have been proposed. In this research, we preliminary measured performance chair assistance that enables StS motions faster movements than conventional assistive chairs. The has two actuators for sliding forward lifting up seat, enough power speed enabling fast StS. We analyzed users' coordination muscle activities terms synergy through experiments. results show users can perform...

This study introduces a body-weight-support (BWS) robot actuated by two pneumatic artificial muscles (PAMs). Conventional BWS devices typically use springs or single actuator, whereas our has split force-controlled (SF-BWS), in which actuators independently support the left and right sides of user’s body. To reduce experience weight, vertical unweighting forces are transferred directly to hips through newly designed harness with an open space around shoulder upper chest area allow freedom...

In this study, we propose a databasedriven torque estimation approach for EMG-based robot control. For conventional controllers, models need to be carefully calibrated control robots that have multiple degrees of freedom. However, such calibration procedure requires significant effort and restricts the applications methods practical situations. To cope with issue, use large-scale data acquired from other users avoid process collaborative filtering estimate joint new user by exploiting...

Previous works in the literature have claimed that characteristics of electromyography (EMG) signals depend on each person, and thus, EMG interfaces need to be carefully calibrated for user myoelectric control. In this study, we show interface used estimate joint torques a can constructed simply by incorporating other users' data without typical calibration process. To achieve plug-and-play capability, introduce concept collaborative filtering torque novel exploiting preidentified...

In this paper, the shoulder glenohumeral displacement during movement of upper arm is studied. Four modeling approaches were examined and compared to estimate humeral head elevation (vertical displacement) translation (horizontal displacement). A biomechanics-inspired method was used firstly model in which a least squares implemented for parameter identification. Then, three Gaussian process regression models following variable sets employed: i) adduction/abduction angle, ii) combination...

In this study, we propose an optimal assistive control strategy that uses estimated user's movement intention as the terminal cost function. We estimate by observing human joint angle, angluar velocity, and muscle activities for very short period of time. A task-related low-dimensional feature space is extracted from observed data. assume discrete number laws associated to different target tasks are pre-computed. Then, policy derived blending pre-computed based on linear Bellman combination...