We report the application of fiber Bragg grating (FBG) arrays in polymer optical fibers for quasi-distributed displacement and torque sensing a series elastic actuator's spring. The were inscribed cyclic transparent through direct-write, plane-by-plane inscription method using femtosecond laser. Two with four FBGs each positioned on rotary spring, numerical analysis finite-element was performed to evaluate sensors' behavior under different loading conditions. Experimental tests results show...

Series elastic actuators (SEAs) can provide low output impedance, bandwidth close to the human movement, and direct measurement of torque through spring deflection. These advantages enable application SEAs in active orthosis exoskeletons. However, conventional technologies measure deflection are bulky, inhibit natural pattern movement or sensitive misalignments. This paper presents polymer optical fiber (POF) as a sensor overcome some issues technologies, since it is compact, lightweight,...

This paper presents the application of polymer optical fiber (POF) sensors in wearable devices for gait assistance. Intensity variation-based POF curvature sensor acquisition joint angles and a insole based to detect events were developed. Three employed practical validation developed sensors: an active orthosis with one degree freedom, robotic exoskeleton, functional electrical stimulation (FES) system is viscoelastic material, which does not have constant response stress or strain also can...

Robot-aided neuro-rehabilitation is a potential solution to help resolve the increasing demand for post-stroke therapy. In this letter, we propose an assistive-resistive approach promote active patient participation during robotic order automatically determine most appropriate level of task difficulty, recommend novel real-time adaptation assistance according instantaneous patient's and performance. First, estimated dynamic con tribution from robot's torque kinematic data. Second, adapted...

This paper develops procedures based on H/sub /spl infin// techniques to achieve position control underactuated manipulators. First, in a deterministic approach, we compare two nonlinear order verify the differences structure and robustness of controllers quasi-linear parameter varying (LPV) representation game theory. In this case, use brakes is necessary when fault occurs. Secondly, develop procedure Markov theory, where an Markovian controller used manipulators subject abrupt changes...

Real time identification of gait events is a mandatory condition for adaptive or patient-tailored control robotic devices during therapy. Despite most the studies in literature have reported high accuracy phases healthy subjects, them were not tested on impaired subjects and/or are suitable real-time implementations. In this paper, we evaluated feasibility some known algorithms events. We propose novel algorithm that exploits advantages different approaches used detection built wearable...

This paper develops a Markovian jump model to describe the fault occurrences in manipulator robot of three joints. includes changes operation points and probability that occurs an actuator. After fault, works as with free Based on developed model, comparative study among controllers, H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> , xmlns:xlink="http://www.w3.org/1999/xlink">infin</sub> mixed /H is presented, applied actual subject one...

The human-robot interaction has played an important role in rehabilitation robotics and impedance control been used the regulation of forces between robot actuator human limbs. Series elastic actuators (SEAs) have efficient solution design this kind robotic application. Standard implementations with SEAs require internal force loop for guaranteeing desired output. However, nonlinearities uncertainties hamper such a guarantee accurate level interaction. This paper addresses dependence...

This paper presents an assist-as-needed method for ankle rehabilitation based on adaptive impedance control strategies. First, the dynamic contribution of patient during motion is estimated from robot's torque and kinematic information. Then, two robot assistance strategies (complementary optimal) are proposed to calculate stiffness parameter controller. In addition, level adapted patient's performance in a video game. The were implemented using Anklebot evaluated four post-stroke patients...

This paper deals with the evaluation of an exoskeleton designed for assisting individuals to rehabilitate compromised lower limb movements resulting from stroke or incomplete spinal cord injury. The is composed lightweight tubular structures and six free joints that provide a modular feature system. allows be adapted assist movement one more patient joints. actuation also modular, can performed passively, by means springs dampers, actively through actuators. In addition, its telescopic...

Automatic detection of gait phases through supervised learning is a feasible approach that takes advantage the consistency cycle among healthy subjects. However, patterns subjects with impairments are not as consistent and most existing algorithms have limited performance detecting during impaired gait. In this paper, we proposed one algorithm used linear classifiers to detect in real-time transition between consecutive phases. Our generalization rule- threshold-based for event detection....

In this paper, we discuss a strategy for the adaptation of "difficulty level" in games intended to include motor planning during robotic rehabilitation. We consider concurrently motivation user and his/her performance Pong game. User is classified three levels (not motivated, well motivated overloaded). measured as combination knowledge results-achieved goals score points game - joint displacement, speed, aiming, work, etc. Initial results pilot test with unimpaired healthy young volunteers...

This article proposes a reinforcement learning approach to dynamically model the player skills in applications that integrate games and rehabilitation robotic. The aims match game difficulty skills, keeping proper motivation (flow) during process. traditional process involves repetitive exercises. Robots serious provide new means improve user commitment treatment. Each person shows different when facing challenges posed by computer games. Thus, level should be adjusted each skill level....

This paper summarizes our on-going efforts to design adaptive assist-as-needed impedance controllers for ankle rehabilitation. Two robot assistance control strategies were evaluated: the first one attempted normalize combined and patient via a complementary stiffness based on estimate of patient's second searched an optimal solution that minimized cost function relating rehabilitation goal interaction between robot. For both strategies, level was adapted performance distinct video games...

Automatic identification of gait events is an essential component the control scheme assistive robotic devices. Many available techniques suffer limitations for real-time implementations and in guaranteeing high performances when identifying subjects with impairments. Machine learning algorithms offer a solution by enabling training different models to represent patterns subjects. Here our aim twofold: remove need stages using unsupervised learning, modify parameters according changes within...