1,720,969 research outputs found
Design and testing of (A)MICO: a multimodal feedback system to facilitate the interaction between cobot and human operator
Full text linkThe present work describes the design, development and testing of a multimodal feedback system, named (A)MICO, with visual and acoustic feedback designed to facilitate the interaction of workers with collaborative robots (cobots) in production lines. The feedback is designed to make the human operator more aware of the cobot’s ongoing and future activities, and therefore gain more control over the situation. The ultimate goal is to obtain a new intuitive mode for transferring information through the combination of lights and sounds, not only to facilitate the flow of communication from the cobot to the operator, but also to make the interaction more accessible to neurodivergent groups, such as people with autism spectrum disorders. The design process focused on the evaluation of the human–robot interaction to select the situations where additional information is needed, and which is the best way to transfer messages as intuitively as possible. Potential end-users were actively involved during all stages of the design and development process. Five volunteers with high functioning autism participated in a preliminary co-design to identify the issues related to the interaction with the cobot and the logic of the multimodal signals. Then, to assess the system’s adaptability to several needs and the level of usability in providing information, validation tests were carried out involving a wider group of participants with ASD. The results suggest that the adoption of a multimodal communication strategy can be useful for making the workplace accessible and improving the well-being of all workers
Biomechanical Assessments of the Upper Limb for Determining Fatigue, Strain and Effort from the Laboratory to the Industrial Working Place: A Systematic Review
Full text linkRecent human-centered developments in the industrial field (Industry 5.0) lead companies and stakeholders to ensure the wellbeing of their workers with assessments of upper limb performance in the workplace, with the aim of reducing work-related diseases and improving awareness of the physical status of workers, by assessing motor performance, fatigue, strain and effort. Such approaches are usually developed in laboratories and only at times they are translated to on-field applications; few studies summarized common practices for the assessments. Therefore, our aim is to review the current state-of-the-art approaches used for the assessment of fatigue, strain and effort in working scenarios and to analyze in detail the differences between studies that take place in the laboratory and in the workplace, in order to give insights on future trends and directions. A systematic review of the studies aimed at evaluating the motor performance, fatigue, strain and effort of the upper limb targeting working scenarios is presented. A total of 1375 articles were found in scientific databases and 288 were analyzed. About half of the scientific articles are focused on laboratory pilot studies investigating effort and fatigue in laboratories, while the other half are set in working places. Our results showed that assessing upper limb biomechanics is quite common in the field, but it is mostly performed with instrumental assessments in laboratory studies, while questionnaires and scales are preferred in working places. Future directions may be oriented towards multi-domain approaches able to exploit the potential of combined analyses, exploitation of instrumental approaches in workplace, targeting a wider range of people and implementing more structured trials to translate pilot studies to real practice
A Dataset on Human-Cobot Collaboration for Action Recognition in Manufacturing Assembly
No full textThis paper introduces a dataset on Human-cobot collaboration for Action Recognition in Manufacturing Assembly (HARMA3). It is a collection of RGB frames, Depth maps, RGB-to-depth-Aligned (RGB-A) frames and Skeleton data relative to actions performed by different subjects in collaboration with a cobot for building an Epicyclic Gear Train (EGT). In particular, 27 subjects executed several trials of the assembly task, which consisted of 7 actions. Data were collected in a laboratory scenario using two Microsoft® Azure Kinect cameras positioned in frontal and lateral positions. The dataset represents a good foundation for developing and testing advanced action recognition as well as action segmentation systems with far-reaching implications beyond human-cobot collaboration. Further potential applications include Computer Vision, Machine Learning, and Smart Manufacturing. Preliminary experiments for action segmentation by applying a state-of-the-art method on features extracted from RGB and skeletal data are presented in this paper, showing high-performance rates
The effects of robotic assistance on upper limb spatial muscle synergies in healthy people during planar upper-limb training
Full text linkBACKGROUND: Robotic rehabilitation is a commonly adopted technique used to restore motor functionality of neurological patients. However, despite promising results were achieved, the effects of human-robot interaction on human motor control and the recovery mechanisms induced with robot assistance can be further investigated even on healthy subjects before translating to clinical practice. In this study, we adopt a standard paradigm for upper-limb rehabilitation (a planar device with assistive control) with linear and challenging curvilinear trajectories to investigate the effect of the assistance in human-robot interaction in healthy people. METHODS: Ten healthy subjects were instructed to perform a large set of radial and curvilinear movements in two interaction modes: 1) free movement (subjects hold the robot handle with no assistance) and 2) assisted movement (with a force tunnel assistance paradigm). Kinematics and EMGs from representative upper-limb muscles were recorded to extract phasic muscle synergies. The free and assisted interaction modes were compared assessing the level of assistance, error, and muscle synergy comparison between the two interaction modes. RESULTS: It was found that in free movement error magnitude is higher than with assistance, proving that task complexity required assistance also on healthy controls. Moreover, curvilinear tasks require more assistance than standard radial paths and error is higher. Interestingly, while assistance improved task performance, we found only a slight modification of phasic synergies when comparing assisted and free movement. CONCLUSIONS: We found that on healthy people, the effect of assistance was significant on task performance, but limited on muscle synergies. The findings of this study can find applications for assessing human-robot interaction and to design training to maximize motor recovery
A framework for human–robot collaboration enhanced by preference learning and ergonomics
Full text linkIndustry 5.0 aims to prioritize human operators, focusing on their well-being and capabilities, while promoting collaboration between humans and robots to enhance efficiency and productivity. The integration of collaborative robots must ensure the health and well-being of human operators. Indeed, this paper addresses the need for a human -centered framework proposing a preference -based optimization algorithm in a human- robot collaboration (HRC) scenario with an ergonomics assessment to improve working conditions. The HRC application consists of optimizing a collaborative robot end -effector pose during an object -handling task. The following approach (AmPL-RULA) utilizes an Active multi -Preference Learning (AmPL) algorithm, a preferencebased optimization method, where the user is requested to iteratively provide qualitative feedback by expressing pairwise preferences between a couple of candidates. To address physical well-being, an ergonomic performance index, Rapid Upper Limb Assessment (RULA), is combined with the user's pairwise preferences, so that the optimal setting can be computed. Experimental tests have been conducted to validate the method, involving collaborative assembly during the object handling performed by the robot. Results illustrate that the proposed method can improve the physical workload of the operator while easing the collaborative task
Modeling perception of human-robot interaction: toward natural and social HRI experiences
Full text linkThe increasing integration of robotic systems into various domains such as industrial, medical, and social environments highlights the growing importance of user-centered design to ensure these technologies enhance human well-being and inclusivity. This thesis investigates the development of a generalized human-driven control architecture aimed at enabling natural and socially-aware human-robot interaction. The proposed framework draws on insights from biomechanics, physiology, psychology, and social science to construct a comprehensive model of the user’s experience. This model enables robotic systems to dynamically adapt their behavior, fostering interactions that are both personalized and empathetic. The research focuses on two domains where user experience is central: collaborative industrial robotics and robotic neurorehabilitation. In industrial settings, cobots are designed to complement human capabilities, improving ergonomics, efficiency, and operator well-being. In neurorehabilitation, robotic systems aim to enhance therapeutic practices and patient engagement through socially-responsive behaviors. These two application areas serve as testbeds for implementing and validating the proposed framework. The methodology consists of three phases: a thorough literature review to identify key user experience factors in HRI; the design and implementation of experimental setups in the selected domains; and validation through empirical studies with diverse participant groups, including neurodivergent individuals. The approach emphasizes ethical considerations and non-invasive data collection, ensuring usability and privacy compliance. Experimental results demonstrate the framework’s capacity to integrate heterogeneous data—biomechanical, physiological, psychological, and social—into actionable insights for real-time robotic adaptation. By enabling smoother, more inclusive interactions, the framework supports a user-centered approach to robotics. Based on these results, the thesis outlines practical guidelines for replicating and extending the architecture across different scenarios, emphasizing its potential to improve usability and social acceptability in HRI. The findings underscore the relevance of interdisciplinary approaches in designing robotic systems that prioritize human experience. Future work will focus on refining adaptability and extending applications to broader contexts.La crescente integrazione di sistemi robotici in ambiti industriali, medici e sociali richiede una sempre maggiore attenzione a un design centrato sull’utente, per garantire che tali tecnologie promuovano benessere e inclusività. Questa tesi presenta lo sviluppo di un’architettura di controllo generalizzata incentrata sull’esperienza dell’utente, al fine di favorire interazioni uomo-robot più naturali e sociali. Il framework proposto integra contributi da biomeccanica, fisiologia, psicologia e scienze sociali per offrire un modello completo dell’esperienza dell’utente, consentendo l’adattamento dinamico dei comportamenti robotici in modo personalizzato ed empatico. Le attività di ricerca si focalizzano su due contesti applicativi principali: la robotica collaborativa industriale e quella neuroriabilitativa. Nel primo caso, i cobot sono impiegati per migliorare ergonomia, efficienza e benessere degli operatori, coniugando la flessibilità umana con la precisione robotica. Nel secondo, i robot supportano le pratiche terapeutiche favorendo il coinvolgimento dei pazienti e l’efficacia riabilitativa tramite comportamenti adattivi e socialmente responsivi. Entrambi gli ambiti fungono da banco di prova per l’implementazione e validazione dell’architettura proposta. La metodologia si articola in tre fasi: revisione della letteratura per individuare i fattori chiave dell’esperienza utente in HRI; progettazione e realizzazione di setup sperimentali nei due ambiti selezionati; validazione empirica tramite studi con partecipanti eterogenei, inclusi soggetti neurodivergenti. L’approccio sperimentale è guidato da principi etici e privilegia la raccolta di dati non invasiva, nel rispetto degli standard di privacy. I risultati evidenziano la capacità dell’architettura di integrare segnali biomeccanici, fisiologici e sociali, traducendoli in strategie adattive in tempo reale. Le soluzioni proposte supportano un controllo robot centrato sull’utente, e la tesi fornisce linee guida pratiche per la sua estensione in contesti diversi, promuovendo usabilità e accettabilità sociale
Emotion based machine learning algorithm for an assistive rehabilitation controller
No full textLAUREA MAGISTRALEI pazienti che necessitano di riabilitazione degli arti superiori stanno diventando sempre più comuni. Questo problema può essere risolto con le tecnologie robotiche. I dispositivi robotici sono altamente avanzati e destinati principalmente ad applicazioni mediche. Il meccanismo in esame era un dispositivo a cinque barre parallele con due gradi di libertà. Un morsetto è stato utilizzato per legare il prototipo già progettato a un bordo del tavolo. L'obiettivo di questo progetto di tesi era assistere il paziente nell'esecuzione di esercizi con il braccio umano. Una telecamera è stata fissata davanti al viso del paziente. Ha catturato le emozioni facciali del paziente e sulla base di queste emozioni, l'algoritmo di apprendimento automatico ha proposto una serie di parametri del controller del tunnel per aiutare il paziente a eseguire l'esercizio con un'esperienza più piacevole.
Il metodo utilizzato sono le reti neurali artificiali (ANN) ispirate alla rete neurale biologica umana, che è lo stato dell'arte delle reti neurali e dell'RNN LSTM. Un modello di algoritmo di apprendimento automatico è stato progettato per la riabilitazione assistiva per prevedere i valori del controller del tunnel per regolare il comportamento del sistema robotico in base al soggetto, per rendere più piacevole l'esperienza di interazione per il paziente. Gli algoritmi che sono stati implementati in questa ricerca erano il modello K-fold ANN, MLP e RNN LSTM. Tra questi tre algoritmi, il modello di previsione delle serie temporali LSTM basato su RNN ha mostrato i risultati migliori. Il set di dati includeva input che erano stati emotivi, ad esempio valenza, eccitazione e dolore. Dopo aver addestrato i dati dal set di dati, l'algoritmo di apprendimento automatico ha proposto una serie di parametri di controllo Guadagno proporzionale Kp, Guadagno integrale Ki, Larghezza del controller del tunnel e rigidità virtuale per il controller del tunnel come uscite per assistere il paziente. Gli output previsti per il controller del tunnel erano molto vicini agli output effettivi del set di dati. Il ricercatore stesso ha sottoposto per la parte sperimentale ad esaminare ed esplorare le prestazioni dell'architettura, mentre il lavoro futuro consisterà in un'ampia campagna sperimentale.Patients who need upper limb rehabilitation are becoming more and more common. This problem can be solved by robotic technologies. Robotic devices are highly advanced and mostly intended for medical applications. The mechanism under examination was a parallel five bar device with two degrees of freedom. A clamp was used to bind the already designed prototype to one edge of the table. The goal of this thesis project was to assist the patient to perform exercise with human arm. A camera was fixed in front of patient's face. It captured patient's facial emotions and based on these emotions, machine learning algorithm proposed a set of tunnel controller parameters to assist the patient to perform exercise with more enjoyable experience.
The method used was artificial neural networks (ANNs) inspired from the human biological neural network, which is the state of the art of neural networks and RNN LSTM. A model of machine learning algorithm was designed for assistive rehabilitation to predict the values of tunnel controller to tune the behaviour of robotic system according to the subject, to make experience of interaction for the patient more enjoyable. The algorithms which were implemented in this research were K-fold ANN, MLP and RNN LSTM model. Among these three algorithms the RNN based LSTM time series forecasting model has shown the best results. Data set included inputs which were emotional states i.e. valence, arousal and pain. After training the data from the data set, the machine learning algorithm proposed a set of control parameters Kp proportional gain, Ki Integral gain, Width of tunnel controller and virtual stiffness for tunnel controller as the outputs to assist the patient. The predicted outputs for tunnel controller were very close to the actual outputs of data set. The researcher itself subjected for the experimental part to examine and explore the performance of the architecture, whereas the future work will consist of extensive experimental campaign
From collaborative robot to collaborative space : intelligent human robot collaboration for smart factory
Full text linkLAUREA MAGISTRALESistemi in grado di raccogliere dati, comunicare e rispondere in tempo reale sono
componenti critiche per la realizzazione di una collaborazione uomo-robot sicura ed efficace
per applicazioni manufatturiere in future “smart factories”. È quindi di fondamentale
importanza una perfetta integrazione di strumenti cognitivi, di rilevamento e previsione
all’interno della struttura di controllo del robot. Inoltre, il robot è inserito in un contesto
manufatturiero eterogeneo, caratterizzato dalla presenza sia di operatori che di attrezzatura.
L’obiettivo di questa tesi è, quindi, lo sviluppo di una cosiddetta Proactive Adaptive Collaboration
Intelligence (PACI) e di una logica di switch, componenti centrali dell’architettura
di controllo proposta. Grazie a questo approccio, gli autori intendono conferire al robot la
capacità di adattare i propri movimenti dinamicamente sulla base delle traiettorie definite
offline e del comportamento rilevato dell’operatore. La sfida sta quindi nello sviluppo
di capacità decisionali avanzate finalizzate ad ottenere un sistema robotico innovativo,
in grado di comprendere la specifica situazione in cui si trova e reagire di conseguenza,
rispettando i requisiti di sicurezza e garantendo alti livelli di produttività.
Gli autori hanno scelto di utilizzare ROS Melodic Morenia (sistema operativo: Ubuntu
18.04 Bionic Beaver) in quanto rappresenta uno standard per la ricerca robotica e assicura
grande scalabilità e manutenibilità del sistema. La struttura di controllo, sviluppata
all’interno di questa piattaforma, presenta le seguenti caratteristiche: flessibilità (adatta
ad un ampio spettro di applicazioni), accessibilità (interfaccia user friendly), modularità
(tecniche di offline planning, soluzioni di controllo e comportamento selezionabili e facilmente
espandibili), sicurezza e produttività. I comportamenti reattivi del robot sono stati ottenuti
tramite una logica di switch che sfrutta funzioni di costo per attivare in tempo reale il
controller più adatto alla situazione. Il “costo di attivazione” è valutato sulla base di
dati relativi a sicurezza (distanza dall’operatore o altri ostacoli) e produttività (ritardi
di produzione registrati). Sfruttando librerie open-source (MoveIt!), i moduli di controllo
sviluppati hanno dimostrato alti livelli di modularità e flessibilità. Un banco di prova
con hardware-in-the-loop (e.DO robot) e percezione dell’operatore emulata è stato, infine,
sfruttato per validare le performance del sistema soggetto a diversi livelli di interazione
uomo-robot.
Questa tesi si inserisce in un contesto di collaborazione internazionale ed è stata
svolta presso University of Florida (Gainesville, FL, USA). I risultati di questo lavoro
rappresentano il punto di partenza per un progetto di ampio respiro chiamato “Intelligent
Human-Robot Collaboration for Smart Factory” e finanziato dal programma NSF-NRI.To enable safe and effective human-robot collaboration (HRC) in smart manufacturing,
seamless integration of sensing, cognition and prediction into the robot controller is critical
for real-time awareness, response and communication. Further complicating matters, the
robot is co-operating within a heterogeneous manufacturing environment (robots, humans,
equipment). Therefore, the specific research objective of this thesis is to provide the robot
Proactive Adaptive Collaboration Intelligence (PACI) and switching logic within its control
architecture. That is, give the robot the ability to optimally and dynamically adapt its
motions given a priori knowledge and predefined execution plans for its assigned tasks,
and detected human actions. The challenge lies in augmenting the robot’s decision-making
process to have a greater situation awareness and to yield robot behaviors/reactions subject
to different human-robot actions while simultaneously maintaining safety and production
efficiency.
The work was carried out using ROS Melodic Morenia (running on Ubuntu 18.04
Bionic Beaver) since it is today’s standard platform for robotic research and ensures great
scalability and maintainability of the system. Inside this framework, a control architecture
was developed to have features: flexibility (suitable for a large range of applications),
accessibility (user friendly interface), modularity (selective and expandable path planning
techniques, high-level controllers, behavior definitions), safety and productivity. Robot
reactive behaviors were achieved via cost function-based switching logic activating the best
suited high-level controller. The cost is a function of safety (e.g., obstacle/human proximity)
and productivity (e.g., induced time delays). Leveraging the availability of numerous path
planning and robot controllers in existing open-source robot libraries (MoveIt!), the PACI’s
underlying segmentation and switching logic framework was demonstrated to yield a high
degree of modularity and flexibility. Using a hardware-in-the-loop testbed setup, the
performance of the developed control architecture subjected to different levels of humanrobot
interactions was validated in the University of Florida e.DO robot testbed, simulating
perception of the human operator.
This research has been carried out at University of Florida (Gainesville, FL, USA),
member of a multi-university/industry international collaboration. It represents the starting
point for a long-term project funded by NSF-NRI and called “Intelligent Human-Robot
Collaboration for Smart Factory”
Design of a multimodal device to improve well-being of autistic workers interacting with collaborative robots
No full text- …
