Добірка наукової літератури з теми "Human-exoskeleton interaction"

The design of a wearable robotic exoskeleton needs to consider the interaction, either physical or cognitive, between the human user and the robotic device. This paper presents a method to analyse the interaction between the human user and a unilateral, wearable lower-limb exoskeleton. The lower-limb exoskeleton function was to compensate for muscle weakness around the knee joint. It is shown that the cognitive interaction is bidirectional; on the one hand, the robot gathered information from the sensors in order to detect human actions, such as the gait phases, but the subjects also modified their gait patterns to obtain the desired responses from the exoskeleton. The results of the two-phase evaluation of learning with healthy subjects and experiments with a patient case are presented, regarding the analysis of the interaction, assessed in terms of kinematics, kinetics and/or muscle recruitment. Human-driven response of the exoskeleton after training revealed the improvements in the use of the device, while particular modifications of motion patterns were observed in healthy subjects. Also, endurance (mechanical) tests provided criteria to perform experiments with one post-polio patient. The results with the post-polio patient demonstrate the feasibility of providing gait compensation by means of the presented wearable exoskeleton, designed with a testing procedure that involves the human users to assess the human-robot interaction.

The advantages of exoskeletons based on the Bowden cable include being lightweight and flexible, thus being convenient in assisting humans. However, the performance of an exoskeleton is limited by the structure and human–exoskeleton interaction, which is analyzed from the established mathematical model of the human–exoskeleton system. In order to improve the auxiliary accuracy, corresponding control methods are proposed. The disturbance observer is designed to compensate for disturbances and parameter perturbations in the inner loop. The human–exoskeleton interaction feedforward model is integrated into the admittance control, which overcomes the limitation of the force loading caused by the friction of the Bowden cable and the change in stiffness of the human–exoskeleton interaction. Furthermore, an angle prediction method using the encoder as the signal source is designed to reduce the disturbance of the force loading caused by human motion. Finally, the effectiveness of the design method proposed in this paper is verified through experiments.

In this article, a method of multi-connection load compensation and load information calculation for an upper-limb exoskeleton is proposed based on a six-axis force/torque sensor installed between the exoskeleton and the end effector. The proposed load compensation method uses a mounted sensor to measure the force and torque between the exoskeleton and load of different connections and adds a compensator to the controller to compensate the component caused by the load in the human–robot interaction force, so that the human–robot interaction force is only used to operate the exoskeleton. Therefore, the operator can manipulate the exoskeleton with the same interaction force to lift loads of different weights with a passive or fixed connection, and the human–robot interaction force is minimized. Moreover, the proposed load information calculation method can calculate the weight of the load and the position of its center of gravity relative to the exoskeleton and end effector accurately, which is necessary for acquiring the upper-limb exoskeleton center of gravity and stability control of whole-body exoskeleton. In order to verify the effectiveness of the proposed method, we performed load handling and operational stability experiments. The experimental results showed that the proposed method realized the expected function.

The upper-limb exoskeleton is capable of enhancing human arm strength beyond normal levels, whereas deriving the operator’s desired action straightforward turns out to be one of the significant difficulties facing human-robot interaction research. In the study, the human-robot interface was presented to regulate the exoskeleton tracking human elbow motion trajectory that employed the contact force signals between the exoskeleton and its operator as the primary means of information transportation. The signals were recorded by adopting the novel soft skin sensors attached to the bracket on the exoskeleton linkage, which could reflect the human arm motion intention through the virtual admittance model and adaptive control. Subsequently, a 1-DOF upper-limb exoskeleton was designed to illustrate the performance of the proposed sensor and the interaction control method in the human-robot cooperation experiment.

Stroke and related complications such as hemiplegia and disability create huge burdens for human society in the 21st century, which leads to a great need for rehabilitation and daily life assistance. To address this issue, continuous efforts are devoted in human–machine interaction (HMI) technology, which aims to capture and recognize users’ intentions and fulfil their needs via physical response. Based on the physiological structure of the human hand, a dimension-adjustable linkage-driven hand exoskeleton with 10 active degrees of freedom (DoFs) and 3 passive DoFs is proposed in this study, which grants high-level synergy with the human hand. Considering the weight of the adopted linkage design, the hand exoskeleton can be mounted on the existing up-limb exoskeleton system, which greatly diminishes the burden for users. Three rehabilitation/daily life assistance modes are developed (namely, robot-in-charge, therapist-in-charge, and patient-in-charge modes) to meet specific personal needs. To realize HMI, a thin-film force sensor matrix and Inertial Measurement Units (IMUs) are installed in both the hand exoskeleton and the corresponding controller. Outstanding sensor–machine synergy is confirmed by trigger rate evaluation, Kernel Density Estimation (KDE), and a confusion matrix. To recognize user intention, a genetic algorithm (GA) is applied to search for the optimal hyperparameters of a 1D Convolutional Neural Network (CNN), and the average intention-recognition accuracy for the eight actions/gestures examined reaches 97.1% (based on K-fold cross-validation). The hand exoskeleton system provides the possibility for people with limited exercise ability to conduct self-rehabilitation and complex daily activities.

Exoskeletons have been assessed by qualitative and quantitative features known as performance indicators. Within these, the ergonomic indicators have been isolated, creating a lack of methodologies to analyze and assess physical interfaces. In this sense, this work presents a three-dimensional relative motion assessment method. This method quantifies the difference of orientation between the user’s limb and the exoskeleton link, providing a deeper understanding of the Human–Robot interaction. To this end, the AGoRA exoskeleton was configured in a resistive mode and assessed using an optoelectronic system. The interaction quantified a difference of orientation considerably at a maximum value of 41.1 degrees along the sagittal plane. It extended the understanding of the Human–Robot Interaction throughout the three principal human planes. Furthermore, the proposed method establishes a performance indicator of the physical interfaces of an exoskeleton.

In this work, a general concept of the human-exoskeleton compatibility and interaction control is addressed. Rehabilitation, as applied to humans with motor control disorder, involves repetitive gait training in relation to lower limb extremity and repetitive task training in relation to upper limb extremity. It is in this regard that exoskeletal systems must be kinematically compatible with those of the subject in order to guarantee that the subject is being trained properly. The incompatibility between the wearable robotic device and the wearer results in joint misalignment, thus introducing interaction forces during movement. This, therefore, leads to the introduction of the need for interaction control in wearable robotic devices. Human-exoskeleton joint alignment is an uphill task; hence, measures to actualize this in order to guarantee the safety and comfort of humans are necessary. These measures depend on the types of joints involved in the rehabilitation or assistive process. Hence, several upper and lower extremity exoskeletons with concepts relating to interaction forces reduction are reviewed. The significant distinction in the modelling strategy of lower and upper limb exoskeletons is highlighted. Limitations of certain exoskeletal systems which may not allow the application of interaction control are also discussed.

Exoskeletons and exosuits have witnessed unprecedented growth in recent years, especially in the medical and industrial sectors. In order to be successfully integrated into the current society, these devices must comply with several commercialization rules and safety standards. Due to their intrinsic coupling with human limbs, one of the main challenges is to test and prove the quality of physical interaction with humans. However, the study of physical human–exoskeleton interactions (pHEI) has been poorly addressed in the literature. Understanding and identifying the technological ways to assess pHEI is necessary for the future acceptance and large-scale use of these devices. The harmonization of these evaluation processes represents a key factor in building a still missing accepted framework to inform human–device contact safety. In this review, we identify, analyze, and discuss the metrics, testing procedures, and measurement devices used to assess pHEI in the last ten years. Furthermore, we discuss the role of pHEI in safety contact evaluation. We found a very heterogeneous panorama in terms of sensors and testing methods, which are still far from considering realistic conditions and use-cases. We identified the main gaps and drawbacks of current approaches, pointing towards a number of promising research directions. This review aspires to help the wearable robotics community find agreements on interaction quality and safety assessment testing procedures.

With the increased use of exoskeletons in a variety of fields such as industry, military, and health care, there is a need for measurement standards to understand the effects of exoskeletons on human motion. Optical tracking systems (OTS) provide high accuracy human motion tracking, but are expensive, require markers, and constrain the tests to a specified area where the cameras can provide sufficient coverage. This study describes the feasibility of using lower cost, portable, markerless depth camera systems for measuring human and exoskeleton 3-dimensional (3D) joint positions and angles. A human performing a variety of industrial tasks while wearing three different exoskeletons was tracked by both an OTS with modified skeletal models and a depth camera body tracking system. A comparison of the acquired data was then used to facilitate discussions regarding the potential use of depth cameras for exoskeleton evaluation.

Control of systems requiring direct physical human-robot interaction (pHRI) requires special consideration of the motion, dynamics, and control of both the human and the robot. Humans actively change their dynamic characteristics during motion, and robots should be designed with this in mind. Both the case of humans trying to control haptic robots using physical contact and the case of using wearable robots that must work with human muscles are pHRI systems. Force feedback haptic devices require physical contact between the operator and the machine, which creates a coupled system. This human contact creates a situation in which the stiffness of the system changes based on how the operator modulates the stiffness of their arm. The natural human tendency is to increase arm stiffness to attempt to stabilize motion. However, this increases the overall stiffness of the system, making it more difficult to control and reducing stability. Instability poses a threat of injury or load damage for large assistive haptic devices with heavy loads. Controllers do not typically account for this, as operator stiffness is often not directly measurable. The common solution of using a controller with significantly increased controller damping has the disadvantage of slowing the device and decreasing operator efficiency. By expanding the information available to the controller, it can be designed to adjust a robot's motion based on the how the operator is interacting with it and allow for faster movement in low stiffness situations. This research explored the utility of a system that can estimate operator arm stiffness and compensate accordingly. By measuring muscle activity, a model of the human arm was utilized to estimate the stiffness level of the operator, and then adjust the gains of an impedance-based controller to stabilize the device. This achieved the goal of reducing oscillations and increasing device performance, as demonstrated through a series of user trials with the device. Through the design of this system, the effectiveness of a variety of operator models were analyzed and several different controllers were explored. The final device has the potential to increase the performance of operators and reduce fatigue due to usage, which in industrial settings could translate into better efficiency and higher productivity. Similarly, wearable robots must consider human muscle activity. Wearable robots, often called exoskeleton robots, are used for a variety of tasks, including force amplification, rehabilitation, and medical diagnosis. Force amplification exoskeletons operate much like haptic assist devices, and could leverage the same adaptive control system. The latter two types, however, are designed with the purpose of modulating human muscles, in which case the wearer's muscles must adapt to the way the robot moves, the reverse of the robot adapting to how the human moves. In this case, the robot controller must apply a force to the arm to cause the arm muscles to adapt and generate a specific muscle activity pattern. This related problem is explored and a muscle control algorithm is designed that allows a wearable robot to induce a specified muscle pattern in the wearer's arm. The two problems, in which the robot must adapt to the human's motion and in which the robot must induce the human to adapt its motion, are related critical problems that must be solved to enable simple and natural physical human robot interaction.

In the latest decades, the common perception about the role of robotic devices in the modern society dramatically changed. In the early stages of robotics, temporally located in the years of the economic boom, the development of new devices was driven by the industrial need of producing more while reducing production time and costs. The demand was, therefore, for robotic devices capable of substituting the humans in performing simple and repetitive activities. The execution of predefined basic activities in the shortest amount of time, inside carefully engineered and confined environments, was the mission of robotic devices. Beside the results obtained in the industrial sector, a progressive widening of the fields interested in robotics – such as rehabilitation, elderly care, and medicine – led to the current vision of the device role. Indeed, these challenging fields require the robot to be a partner, which works side-by-side with the human. Therefore, the device needs to be capable of actively and efficiently interacting with humans, to provide support and overcome their limits in the execution of shared activities, even in highly unpredictable everyday environments. Highly complex and advanced robots, such as surgical robots, rehabilitation devices, flexible manipulators, and service and companion robots, have been recently introduced into the market; despite their complexity, however, they are still tools to be used to perform, better or faster, very specific tasks. The current open challenge is, therefore, to develop a new generation of symbiotically cooperative robotic partners, adding to the devices the capability to detect, understand, and adapt to the real intentions, capabilities, and needs of the humans. To achieve this goal, a bidirectional information channel shall be built to connect the human and the device. In one direction, the device requires to be informed about the state of its user; in the other direction, the human needs to be informed about the state of the whole interacting system. This work reports the research activities that I conducted during my PhD studies in this research direction. Those activities led to the design, development, and assessment on a real application of an innovative multilevel framework to close the cooperation loop between a human and a robotic device, thus promoting and enhancing their symbiotic interaction. Three main levels have been identified as core elements to close this loop: the measure level, the model level, and the extract/synthesize level. The former aims at collecting experimental measures from the whole interacting system; the second aims at estimating and predicting its dynamic behavior; the last aims at providing quantitative information to both the human and the device about their performances and about how to modify their behavior to improve their interaction symbiosis. Within the measure level, the focus has been concentrated on investigating, critically comparing, and selecting the most suitable and advanced technologies to measure kinematics and dynamics quantities in a portable and minimally intrusive way. Particular attention has been paid to new emerging technologies; moreover, useful protocols and pipelines already recognized as de-facto in other fields have been successfully adapted to fit the needs of the man-machine interaction context. Finally, the design of a new sensor has been started to overcome the lack of tools capable of effectively measuring human-device interaction forces. To implement the model level, a common platform to perform integrated multilevel simulations – i.e. simulations where the device and the human are considered together as interacting entities – has been selected and extensively validated. Furthermore, critical aspects characterizing the modeling of the device, the human, and their interactions have been studied and possible solutions have been proposed. For example, modeling the mechanics and the control within the selected software platform allowed accurate estimations of their behavior. To estimate human behavior, new methodologies and approaches based on anatomical neuromusculoskeletal models have been developed, validated, and released as open-source tools for the community, to allow accurate estimates of both kinematics and dynamics at run-time – i.e. at the same time that the movements are performed. An inverse kinematics approach has been developed and validated to estimate human joint angles from the orientation measurements provided by wearable inertial systems. Additionally, a state of the art neuromusculoskeletal modeling toolbox has been improved and interfaced with the other tools of the multilevel framework, to accurately predict human muscle forces, joint moments, and muscle and joint stiffness from electromyographic and kinematic measures. To estimate and predict the interactions, contact models, parameters optimization procedures, and high-level cooperation strategies have been investigated, developed, and applied. Within the extract/synthesize level, the information provided by the other levels has been combined together to develop informative feedbacks for both the device and the human. In one direction, the device has been provided with control signals defining how to adjust the provided support to comply with the task goals and with the human current capabilities and needs. In the other direction, quantitative feedbacks have been developed to inform the human about task execution performances, task targets, and support provided by the device. This information has been provided to the user as visual feedbacks designed to be both exhaustively informative and minimally distractive, to prevent possible loss of focus. Moreover, additional feedbacks have been devised to help external observers – therapists in the rehabilitation contexts or task planners and ergonomists in the industrial field – in the design and refinement of effective personalized tasks and long-term goals. The integration of all the hardware and software tools of each level in a modular, flexible, and reliable software framework, based on a well known robotic middleware, has been fundamental to handle the communication and information exchange processes. The developed general framework has been finally specialized to face the specific needs of robotic-aided gait rehabilitation. In this context, indeed, the final aim of promoting the symbiotic cooperation is translatable in maximizing treatment effectiveness for the patients by actively supporting their changing needs and capabilities while keeping them engaged during the whole rehabilitation process. The proposed multilevel framework specialization has been successfully used, as valuable answer to those needs, within the context of the Biomot European project. It has been, indeed, fundamental to face the challenges of closing the informative loop between the user and the device, and providing valuable quantitative information to the external observers. Within this research project, we developed an innovative compliant wearable exoskeleton prototype for gait rehabilitation capable of adjusting, at run-time, the provided support according to different cooperation strategies and to user needs and capabilities. At the same time, the wearer is also engaged in the rehabilitation process by intuitive visual feedbacks about his performances in the achievement of the rehabilitation targets and about the exoskeleton support. Both researchers and clinical experts evaluating the final rehabilitation application of the multilevel framework provided enthusiastic feedbacks about the proposed solutions and the obtained results. To conclude, the modular and generic multilevel framework developed in this thesis has the potential to push forward the current state of the art in the applications where a symbiotic cooperation between robotic devices and humans is required. Indeed, it effectively endorses the development of a new generation of robotic devices capable to perform challenging cooperative tasks in highly unpredictable environments while complying with the current needs, intentions, and capabilities of the human.
Negli ultimi anni si è assistito a un radicale cambiamento negli obiettivi della ricerca robotica.
Agli albori della robotica moderna, storicamente collocati nel contesto del boom economico, lo sviluppo dei dispositivi robotici era guidato dall’esigenza industriale di ridurre tempi e costi di produzione per ottenere quantitativi sempre maggiori. Spesso questo coincideva con l’esigenza di sviluppare dispositivi robotici per sostituire gli uomini nello svolgimento di mansioni semplici e ripetitive. Questa esigenza portava poi alla progettazione di ambienti dedicati intorno ai sistemi robotici. Più recentemente vi è stato un progressivo interesse verso la robotica di nuovi settori quali la riabilitazione, l’assistenza agli anziani, la chirurgia. In questi ambiti il ruolo del dispositivo cambia radicalmente: non è più solo uno strumento da utilizzare, ma diventa un partner con cui lavorare fianco a fianco. Pertanto, il dispositivo deve essere capace di cooperare attivamente ed efficacemente con le persone, comprendendone le esigenze ed aiutandole al fine di ottenere obiettivi condivisi in ambienti non strutturati come quelli in cui quotidianamente ci muoviamo. Lo stato attuale del mercato vede robot utilizzati in diversi campi di applicazione, come robot chirurgici, dispositivi riabilitativi, manipolatori flessibili e robot di servizio e assistenziali ma essi sono ancora spesso semplici strumenti per svolgere specifici compiti. L’attuale sfida aperta è pertanto quella di sviluppare una nuova generazione di robot che sappiano invece essere partner, cooperando in simbiosi con l’uomo. In altre parole, l’obiettivo di ricerca è quello di fornire ai dispositivi robotici la capacità di rilevare, comprendere ed adattarsi alle reali intenzioni, capacità ed esigenze degli esseri umani. Questa cooperazione simbiotica richiede uno scambio bidirezionale di informazioni tra l’uomo e il dispositivo. Da un lato, il dispositivo necessita di essere informato circa le necessità, le capacità e le intenzioni dell’essere umano. Dall’altro lato, l’uomo deve essere informato circa il proprio stato e le intenzioni del dispositivo con cui sta cooperando. Da tali considerazioni, tuttavia, emerge chiaramente la necessità di attingere ed integrare i contributi forniti dalla ricerca della comunità biomeccanica. Questi obiettivi sono quelli che hanno guidato le attività condotte durante il periodo di studio del mio dottorato e che sono riportate, insieme ai risultati ottenuti, in questo elaborato. Tali attività hanno portato a progettare, sviluppare e realizzare un nuovo framework multilivello volto a chiudere l’anello di cooperazione tra essere umano e dispositivo robotico, di fatto promuovendo la loro interazione simbiotica. Al fine di raggiungere tale obbiettivo, sono stati identificati tre livelli principali all’interno del framework multilivello: il livello di misura, il livello di modellazione ed il livello di estrazione/sintesi delle informazioni. Il primo mira a raccogliere misure sperimentali dall’intero sistema cooperante; il secondo a stimare e prevedere il suo comportamento dinamico; l’ultimo a fornire informazioni quantitative sia all’uomo che al dispositivo in merito alle loro prestazioni e a come modificare il loro comportamento per migliorare la loro simbiosi. Nell’ambito del livello di misura, l’attenzione si è concentrata sull’analisi, sul confronto critico e sulla scelta di tecnologie indossabili e minimamente invasive per misurare al meglio la cinematica e la dinamica. Inoltre, protocolli e procedure già sviluppati e riconosciuti come standard de-facto in altri campi sono stati adattati con successo alle esigenze del contesto dell’interazione uomo-macchina. Infine, è stata avviata la progettazione di un nuovo sensore per colmare la mancanza di strumenti in grado di misurare efficacemente le forze emergenti dall’interazione dinamica tra uomo e dispositivo robotico indossabile. In tale contesto, infatti, gli attuali dispositivi di misura non risultano essere utilizzabili senza interferire con l’interazione stessa. Al fine di realizzare il livello di modellazione, è stata innanzitutto selezionata ed ampiamente validata una piattaforma software che fosse in grado di eseguire simulazioni integrate multilivello, cioè simulazioni in cui il dispositivo e l’uomo sono considerati contemporaneamente come entità interagenti. Inoltre, sono stati studiati gli aspetti critici che caratterizzano la modellazione del dispositivo, dell’umano e delle loro interazioni e sono state proposte possibili soluzioni per affrontarli. Ad esempio, la modellazione della meccanica e dei sistemi di controllo dei dispositivi, realizzata attraverso gli strumenti messi a disposizione dalla piattaforma software selezionata, ha permesso di ottenere stime accurate del loro comportamento dinamico. Per stimare il comportamento umano, invece, sono state sviluppate, validate e rilasciate come strumenti open-source alla comunità scientifica nuove metodologie e nuovi approcci basati su modelli anatomici neuromuscoloscheletrici. Tale lavoro ha consentito di ottenere stime accurate sia della cinematica che della dinamica in tempo reale, cioè nello stesso istante in cui i movimenti vengono eseguiti. Al fine di stimare la cinematica articolare dell’uomo, nel corso del mio dottorato ho sviluppato e convalidato un approccio di cinematica inversa basato su un modello muscoloscheletrico anatomicamente attendibile, che utilizza come input le misure di orientazione fornite dai sistemi inerziali indossabili. Inoltre, lo strumento di modellazione neuromuscoloscheletrica che rappresenta l’attuale stato dell’arte in ambito biomeccanico è stato migliorato ed interfacciato con gli altri strumenti del framework multilivello. Il lavoro svolto ha consentito di prevedere con precisione ed in tempo reale le forze muscolari, le coppie articolari, e la rigidità muscolare ed articolare dell’essere umano a partire da misure elettromiografiche e cinematiche. Per stimare e prevedere le interazioni, infine, sono stati studiati, sviluppati ed applicati modelli di contatto, procedure di ottimizzazione dei parametri e strategie di cooperazione ad alto livello volte ad incrementare la simbiosi tra essere umano e dispositivo robotico. Nell’ambito del livello di estrazione/sintesi delle informazioni, le misure e le stime ottenute attraverso gli strumenti realizzati negli altri livelli sono state combinate per ottenere accurati feedback quantitativi sia per il dispositivo che per le persone. Da un lato, al dispositivo sono stati forniti segnali di controllo volti a modulare il supporto al fine di soddisfare al meglio gli obiettivi dell’attività in corso di svolgimento, nel rispetto delle reali capacità ed esigenze umane. Dall’altro lato, sono stati sviluppati feedback quantitativi per informare l’utente sulle proprie prestazioni nell’esecuzione dei compiti, sugli obiettivi delle attività e sul supporto fornito dal dispositivo. Tali informazioni sono state fornite all’utente sotto forma di feedback visivi, concepiti per essere esaustivi senza però distrarre l’attenzione, al fine di evitare eventuali perdite di concentrazione e coinvolgimento. Inoltre, sono stati definiti ulteriori feedback volti ad aiutare gli osservatori esterni, quali terapisti in contesti riabilitativi o gestionali ed ergonomisti in campo industriale, nella progettazione e nel perfezionamento di attività personalizzate ed obiettivi a lungo termine. Tutti gli strumenti hardware e software appartenenti ai diversi livelli sono stati poi integrati sviluppando un framework software modulare, flessibile ed affidabile, basato su un noto middleware robotico, al fine di gestire i processi di comunicazione e scambio di informazioni. Infine, il framework sviluppato nel corso del mio dottorato è stato specializzato per realizzare un’applicazione di riabilitazione della camminata assistita da un dispositivo esoscheletrico. Questo contesto è stato scelto perché la cooperazione simbiotica è fondamentale per raggiungere l’obiettivo finale: massimizzare l’efficacia del percorso riabilitativo che deve essere dinamicamente adattato per seguire al meglio le mutevoli esigenze e capacità del paziente mantenendolo allo stesso tempo coinvolto e concentrato. La specializzazione del framework multilivello proposto è stata utilizzata con successo per realizzare gli obiettivi del progetto Europeo Biomot. All’interno di tale progetto, infatti, abbiamo sviluppato un innovativo prototipo di esoscheletro indossabile per la riabilitazione della camminata in grado di modulare in tempo reale il supporto fornito, seguendo diverse strategie di cooperazione ed in funzione delle esigenze e capacità dell’utente. Allo stesso tempo, l’utente risulta essere coinvolto attivamente nel proprio processo di riabilitazione attraverso accattivanti feedback visivi sulle sue prestazioni nel raggiungimento degli obiettivi di riabilitazione e sul sostegno fornitogli dell’esoscheletro. Il framework si è dimostrato fondamentale per chiudere l’anello di informazioni che collega utente e dispositivo e per fornire preziosi feedback quantitativi agli osservatori esterni. Sia i ricercatori che gli esperti clinici che hanno valutato l’applicazione riabilitativa del framework multilivello hanno fornito feedback entusiasti in merito alle soluzioni proposte e ai risultati ottenuti. Pertanto, si può affermare che il framework multilivello sviluppato in questa tesi ha le potenzialità di avanzare l’attuale stato dell’arte nell’ambito dell’interazione simbiotica uomo–macchina. Infatti, tale framework potrà supportare lo sviluppo di una nuova generazione di dispositivi robotici capaci di cooperare con l’uomo nell’esecuzione di compiti impegnativi in ambienti non strutturati, nel rispetto delle reali esigenze, intenzioni e capacità di quest’ultimo.

The purpose of this thesis is to analyse the Physical Human-Robot Interaction (PHRI) which is an important extension of traditional HRI work. This work of analysis helps in understanding the effects on the upper limb of the human musculoskeletal system when human user interacts with the robotic device. This is concerned for various applicational interests, like in the field of health care, industrial applications, military, sport science and many more. We developed a CAD model of an exoskeleton in SolidWorks to satisfy all the properties required. The designed upper limb exoskeleton has been implemented within the simulating software OpenSim via the platform Notepad++ using xml language. This framework has been used to simulate and analyse the effects at muscular level when the exoskeleton is coupled with the model of the upper limb of the human body for a desired elbow flexion and extension movements. Then the results i.e. force generated by muscles with and without exoskeleton contribution are plotted and compared. The results of the simulations show that, wearing the exoskeleton, the forces exerted by the muscles decrease significantly. This thesis is only the starting point of a wide range of possible future works. Aiming at the use of exact controller, optimization technique, cost estimation possibilities applying to real word model and reaching the people in need.

The control objective of exoskeleton for human performance augmentation is to minimize the human machine interaction force while carrying external loads and following human motion. This paper addresses the dynamics and the interaction force control of a 1-DOF hydraulically actuated joint exoskeleton. A spring with unknown stiffness is used to model the human-machine interface. A cascade force control method is adopted with high-level controller generating the reference position command while low level controller doing motion tracking. Adaptive robust control (ARC) algorithm is developed for both two controllers to deal with the effect of parametric uncertainties and uncertain nonlinearities of the system. The proposed adaptive robust cascade force controller can achieve small human-machine interaction force and good robust performance to model uncertainty which have been validated by experiment.