Human factors and ergonomics

Nowadays, the technological maturity achieved by Extended Reality (XR) technologies allows to develop applications that foster positive emotions and promote personal growth. The scope of this work is to present the first results about a research project (BIOXRE - BIOphilic design of Extended Reality working Environments) focused on the application and evaluation of the biophilia hypothesis of Positive Computing in virtual workplaces. In the literature, there is effectively evidence that biophilic design improves human wellbeing. However, only a few studies evaluated its effectiveness in improving operators' wellbeing in the real workplace. For the first time, BIOXRE aims to exploit the biophilic design to XR applications supporting procedural tasks. The goal is to improve the wellbeing of human operators without affecting their performance in accomplishing manual tasks supported by XR visualization. The outcome of the project will be to extract guidelines for the spatial augmentation of real workplaces, fostered by the biophilic design, that will derive from a well-structured user evaluation in the target industrial scenario.

Keywords: biophilic design , Industry 5.0 , Positive Computing

This study aims to explore the use of Augmented Reality (AR) visual assets to convey procedural instructions, specifically for conveying information about component orientation. We focused on assembly scenarios where no affordance is provided for orientation while maintaining a consistently high affordance for how components are mounted. This information is recurrent in tasks where users are familiar with components that fit together without needing specialized tools but lack knowledge of the specific orientations required for the assembly. A typical example is placing rubber gaskets that fit smoothly into grooves but where no markings indicate the correct sealing side. We evaluated six different AR presentation modes for conveying component orientation: image, video, static side-by-side product model, animated side-by-side product model, static in-situ product model, and animated in-situ product model. The literature provides no clear agreement on which is the most effective. To fill this gap, we conducted a user study with 36 participants, measuring completion time, accuracy, and cognitive load across the six AR presentation modes. We also analyzed how users interacted with each of them and collected user subjective feedback. Our findings revealed that the animated side-by-side product model ensures better completion time, demanding less cognitive load and being favored by users.

Keywords: assembly , Augmented Reality , authoring , instruction , orientation , visual asset

This paper proposes an AI-driven framework for designing multimodal, symbiotic human-machine interfaces (HMIs) in modern vehicles. As automotive systems evolve toward higher levels of automation and user-centric design, the integration of data fusion techniques with deep neural networks becomes essential for enhancing driver safety and comfort. The proposed framework continuously monitors a range of inputs—including driver gaze, facial expressions, posture, and vehicle telemetry—and processes these signals through a hierarchical data fusion methodology. This approach involves raw signal preprocessing, feature extraction and selection, and decision-level integration using advanced deep learning models. Experimental validation using a high-fidelity driving simulator across varied scenarios (high-stress, relaxed, and frustration-inducing) demonstrates the framework's capability to accurately detect driver distraction and drowsiness.

Keywords: Adaptive HMI , Data fusion , Deep Learning , DMS , Symbiotic HMI

Background: Occupational Health and Safety (OHS) training is a crucial component of workplace safety. In recent years, Cross-reality (XR) has emerged as a viable path for immersive and successful training experiences. This paper presents preliminary results from the SCISSOR project, a groundbreaking initiative exploring the potential of XR technologies in enhancing OHS training, especially in healthcare setting regarding ionization radiation exposure. Materials and methods: This quasi-experimental pilot study consisted of several stages from July 2021 to June 2022. Phase 1 consists of the realization of the virtual scenario, Phase 2 in XR setting up and laboratory simulation, and lastly, and Phase 3 based on the on field testing. Results: About 60 participants, including students and researchers, participated in Phases 1 and 2 preliminary testing and 9 healthcare workers (HCWs) in Phase 3. Phase 1 involved the 3D reconstruction of a real-life scenario using digital photogrammetry. These models were then utilized to recreate the operating room scenario and program interactions with the equipment. In Phase 2, the overall appreciation for the level of realism, comfort, and usability of the application was investigated. The test results demonstrated a general appreciation for the application and its effectiveness. Phase 3, concerning field tests, involved HCWs. The results suggest that XR technologies hold significant promise in revolutionizing OHS training, offering a more engaging, effective, and adaptable approach to workplace safety education.

Keywords: Cross-reality , Ionizing radiation , Occupational health and safety , Operating room , Training

In the Industry 5.0 paradigm, where human-centricity and sustainability are paramount, there is a pressing need for a comprehensive framework to evaluate the social sustainability of products through the lens of employee well-being. This paper proposes a human-centric product sustainability index specifically tailored for this purpose. The index incorporates a multidimensional assessment of product processes, utilizing various indicators based on standardized methods. Sensor-based approaches enhance the objectivity and accuracy of the analysis. By providing a quantitative and standardized measure, the index facilitates informed decision-making, promotes responsible product design and manufacturing, and reduces occupational health risks.

Keywords: Human-centric manufacturing , Industry 5.1 , Occupational health , Social sustainability

Musculoskeletal disorders pose a significant occupational health challenge, impairing worker well-being and operational efficiency, particularly in manufacturing. The risks of improper postures require effective identification and mitigation strategies. Current assessment methods rely on subjective, qualitative measures and lack robust numerical metrics, especially in hand pose estimation. This study introduces an advanced machine learning algorithm designed for the identification and classification of hand poses using inertial gloves. By tracking hand positions, the proposed approach facilitates objective risk assessments during work activities. Experimental findings validate the system's efficacy, indicating its potential to improve ergonomic analysis and enhance workplace safety.

Keywords: hand posture recognition , Industry 5.0 , inertial gloves , machine learning algorithm

The design of an assistive device for a student with disabilities to enable his inclusion into the University is based on ethical reasons. The device must be personalized starting from a participative and human centered approach. This work presents the design and development phases of a highly functional assistive device which enables a quadriplegic user, who is unable to speak, to autonomously communicate. The starting phase deals with the identification of the user needs, considering the residual mobility of a single arm and the needs of the caregivers. The motion study allows to define the requirements which enable the definition of the system architecture. A modular architecture is chosen, allowing damaged or obsolete parts to be easily replaced extending the useful life of the device, but reducing the resources consumption during the life cycle. Sustainable and recyclable materials are selected using Additive Manufacturing technologies.

Keywords: Human centered design , Inclusive design , Personalized assistive device

CAD assembly models are typically represented as a collection of components, each of which can share geometric interfaces with others. In the literature, geometric interfaces have been shown to play a fundamental role in assembly model analysis, component characterization, and classification. While these interfaces are not explicitly defined in CAD models, they can be inferred from the relative positioning of components. The resulting geometric interfaces can be categorized as either interference or contact. However, it is often unclear whether these interfaces stem from intentional design choices related to component shape and function, from consistently applied relative positioning, or from unintended errors. In industrial practice, the design of complex products often involves models sourced from public catalogs for third-party components. These catalog models frequently include shape simplifications, which can lead to unintended intersections or clearances with surrounding components — deviations that do not exist in the final physical product. This study aims to provide a comprehensive analysis and formalization of geometric interfaces, based on the complementary roles of CAD assembly modules and digital component catalogs, both widely used in industry as foundational resources for generating assembly models. The results are directly applicable to industrial CAD assembly models and can serve as a reference for CAD developers seeking to improve and extend assembly processing, as well as for researchers conducting assembly analysis. This work introduces a formalization of geometric interfaces, including contacts, interferences, and interface envelopes, which are essential for defining component mounting requirements. An analysis of geometric interface perturbations caused by repetition operators is performed, leading to the concept of an interface envelope to model specific interface repetitions.

Keywords: Assembly , Component digital catalog , Component representation , Geometric constraints , Geometric modeler tolerance , Interface envelope , Interfaces , Symmetry properties

Stress is a reaction that occurs when a person perceives, with or without awareness, an imbalance between requests and available resources. Relying on this definition, we have carried out an experiment in a Virtual Reality environment to elicit (light) stress in the user and analyze the emotional responses with electroencephalography (EEG). The virtual environment is divided in eight parts; in each of them a stressor has been put in action, meaning that in every part the participants perform a task, but a specific resource is missing (time, knowledge, control, salvation, no or too many alternatives, engagement, self-confidence). EEG is used to assess the emotional response with the aid of Valence/Arousal/Dominance/Stress indicators presented in previous literature. Nine indicators calculated for 87 participants, labeled according to self-assessment replies (post-experimental questionnaires), were classified with eXtreme Gradient Boosting, k-Nearest Neighbor, Support Vector Machine and Random Forest classifiers.

Keywords: Affective computing , EEG , Emotion assessment , Machine learning , Stress , Virtual reality

The Industry 5.0 approach moves towards a sustainable, human-centric and adaptable manufacturing within companies, with a philosophy focused on meeting human needs and supporting societal values through the entire production process. Human–Robot Collaboration (HRC) emerges as a pivotal aspect of Industry 5.0, aiming to integrate robots and artificial intelligence seamlessly into workplaces alongside human workers. This integration presents challenges and opportunities for redefining job roles and enhancing productivity, while ensuring safety and workers well-being. This study focuses on the adoption of innovative technologies, such as collaborative robots (cobots) and Augmented Reality (AR) systems, to improve flexibility and safety in manufacturing and logistics for ensuring safe Human–Robot Interaction. The aim is to evaluate how AR devices, e.g., Hololens 2, could facilitate the interaction between cobots and operators, enhancing human performance and safety during robot programming and task execution, regardless of experience levels. Furthermore, the paper proposes the introduction of additional features to the existing AR platform for HRC, such as the creation of virtual walls and real-time operator concentration monitoring using gaze tracking. These improvements aim to empower operators with greater control over collaborative spaces, and enhance safety by ensuring operator focus during critical task phases. Overall, this work contributes to advancing the adoption of Industry 5.0 principles by demonstrating practical application of HRC technologies and AR systems to improve workplaces safety, productivity and Human–Robot Interaction in manufacturing environments.

Keywords: Augmented reality , Head mounted display , Human robot interaction , Robotics , Safety

With the accelerating embrace of automated vehicles, lack of public trust has become a critical concern worldwide. Nonetheless, the literature shows a significant and thorough effort to increase trust in automated vehicles. Biophilia is one strategy being utilized to establish a trustworthy environment, although it has not yet been applied in the automotive industry, thus, a biophilic interior design was developed and applied in a static driving simulator using a mixed reality headset for investigating the effect on the driving experience. A gender-balanced sample (11 F,11 M) was asked to perform an in-lab experiment and experienced four different driving scenarios, with and without Biophilic interventions. The regression results showed that initial trust is the only effective parameter on affective trust in automated vehicles is (P=0.05), and no significant relationship was found between gender and affective trust. As conclusion, the study allowed to assess that initial trust in self-driving cars plays a more important role than gender or driving experience contrary to initial expectations.

Keywords: Affective Trust , Biophilic Automated Vehicles , Gender , User Experience

The introduction of automated vehicles has redefined the level of interaction between the driver and the vehicle, introducing new tasks and so impose different workloads. Existing tools such as NASA-TLX and DALI are still used to assess driving workload in automated vehicles, despite not accounting for new tasks. This study introduces AV-TLX, a specialized tool for measuring workload in Level 3 automated driving. The development process began with a narrative literature review to identify the primary factors influencing workload. This was followed by a series of qualitative sessions during which the dimensions—and later the questions—of the questionnaire were designed. The tool's validity was first assessed using CVR and CVI indices, and its reliability and convergent validity were evaluated using a dynamic driving simulator with medium fidelity. The final version of AV-TLX comprises 19 questions across 8 subscales, demonstrating excellent reliability (0.86) and validity (CVR > 0.78). An agreement scores between the results of AV-TLX and NASA-TLX in the simulation study was 0.6, which is considered acceptable for the consistency of two questionnaires. Furthermore, this questionnaire can be utilized in two ways: by reporting the overall workload and/or divided into 8 primary subscales, or by categorizing the questions into two groups— “takeover task” workload and “automated driving task” workload.

Keywords: Autonomous driving , Autonomous vehicle , DALI , Driver workload , Driving Load , Human Factors in Transportation , NASA-TLX , Workload Questionnaire

Work-related musculoskeletal disorders (WMSDs) are a leading concern in industrial ergonomics, often stemming from sustained non-neutral postures and repetitive tasks. This paper presents a vision-based framework for real-time, frame-level ergonomic risk classification using a lightweight Vision Transformer (ViT). The proposed system operates directly on raw RGB images without requiring skeleton reconstruction, joint angle estimation, or image segmentation. A single ViT model simultaneously classifies eight anatomical regions, enabling efficient multi-label posture assessment. Training is supervised using a multimodal dataset acquired from synchronized RGB video and full-body inertial motion capture, with ergonomic risk labels derived from RULA scores computed on joint kinematics. The system is validated on realistic, simulated industrial tasks that include common challenges such as occlusion and posture variability. Experimental results show that the ViT model achieves state-of-the-art performance, with F1-scores exceeding 0.99 and AUC values above 0.996 across all regions. Compared to previous CNN-based system, the proposed model improves classification accuracy and generalizability while reducing complexity and enabling real-time inference on edge devices. These findings demonstrate the model's potential for unobtrusive, scalable ergonomic risk monitoring in real-world manufacturing environments.

Keywords: attention mechanism , computer vision , deep learning , ergonomic risks , human-centered manufacturing , posture recognition , work-related musculoskeletal disorders

Highlights: What are the main findings? EOG signals from smart eyewear effectively detect stress in controlled tasks Random forest outperforms with 85.8% (2-class) and 82.4% (3-class) accuracy What is the implication of the main finding? Smart glasses may enable practical stress detection in real-world work settings Future work should explore real-world settings and multimodal signal inputs To tackle work-related stress in the evolving landscape of Industry 5.0, organizations need to prioritize employee well-being through a comprehensive strategy. While electrocardiograms (ECGs) and electrodermal activity (EDA) are widely adopted physiological measures for monitoring work-related stress, electrooculography (EOG) remains underexplored in this context. Although less extensively studied, EOG shows significant promise for comparable applications. Furthermore, the realm of human factors and ergonomics lacks sufficient research on the integration of wearable sensors, particularly in the evaluation of human work. This article aims to bridge these gaps by examining the potential of EOG signals, captured through smart eyewear, as indicators of stress. The study involved twelve subjects in a controlled environment, engaging in four stress-inducing tasks interspersed with two-minute relaxation intervals. Emotional responses were categorized both into two classes (relaxed and stressed) and three classes (relaxed, slightly stressed, and stressed). Employing supervised machine learning (ML) algorithms—Random Forest (RF), Logistic Regression (LR), Support Vector Machine (SVM), Decision Tree (DT), and K-Nearest Neighbors (KNN)—the analysis revealed accuracy rates exceeding 80%, with RF leading at 85.8% and 82.4% for two classes and three classes, respectively. The proposed wearable system shows promise in monitoring workers' well-being, especially during visual activities.

Keywords: electrooculography , human factors , random forest , smart eyewear , stress detection , supervised machine learning

This paper presents a Digital Twin-based framework to support the reconfiguration process of Cyber–Physical Production Systems (CPPSs) through human–robot collaboration and Industry 5.0 enabling technologies. The proposed approach integrates a Mixed Reality (MR) module into the digital twin architecture to enhance human–machine interaction, data visualisation, and robot programming. It also incorporates Physics-Informed Neural Networks (PINNs), a hybrid methodology that combines machine learning and physical modelling to improve prediction accuracy and physical consistency. A proof of concept implementation of the framework is carried out on the reconfiguration of a real-world production line within a research platform. The communication mechanism between system modules is tested and discussed in detail. Additionally, the use of PINNs for predicting the energy consumption of a mobile robotic system involved in the reconfiguration task is implemented and benchmarked. The mobile robotic system integrates an AMR (Autonomous Mobile Robot) and a Cobot (collaborative robotic arm).

Keywords: Cyber–physical production system , Digital twin , Human–robot collaboration , Physics-informed neural network , Reconfiguration

Industry 5.0 requires practical methods to translate Augmented Reality (AR) concepts into effective shop floor applications, demonstrating their value to operators. This study introduces a framework for implementing and validating Augmented Reality (AR)-based tools designed to enhance sustainability awareness and assist operators in energy management within industrial settings. The approach combines a reference software architecture for rapid AR deployment with a three-part user-experience assessment, measuring usability (System Usability Scale - SUS), technology acceptance (Technology Acceptance Model - TAM), and cognitive workload (NASA-TLX). To test this framework, an AR-based prototype tool was deployed on enterprise smartphones and evaluated in three scenarios: monitoring service-utility energy consumption, monitoring production equipment, and conducting on- and off-the-job training of the operators. Thirty shop floor professionals completed tasks and provided UX feedback. The results showed good usability (mean SUS 78.4/100), with perceived ease of use and contextual relevance driving technology acceptance. Moreover Nasa TLX analysis indicates mental demand as the predominant factor.

Keywords: Augmented Reality , Employee training , Energy monitoring , Human-centric design , Industry 5.0 , Sustainability , User experience

In the modern industrial and robotics landscape, the demand for customized, high-performance solutions is increasing, posing challenges in reducing design and development times while ensuring product quality. Traditional approaches relying on physical prototypes are expensive, time-consuming, and often inadequate for addressing the growing system complexity. Virtual Prototyping (VP) offers an alternative, enabling analysis, simulation, and optimization of product performance in a virtual environment before physical implementation. This paper focuses on collaborative robotics, where different tools can effectively simulate specific aspects of robotic system design. However, a lack of an integrated framework that seamlessly combines Computer-Aided Design, multibody modelling, and requirement verification still represents a gap. To address this, a Virtual Prototyping framework for robotic system design and verification is proposed. It begins with system requirements elicitation to ensure that the virtual models are fully aligned with the stakeholder needs. By focusing on the components and interactions required to verify the defined requirements, the approach allows consistency and avoids unnecessary complexity. CAD models support visualization and analysis of physical configurations, while multibody modelling enables simulations to verify functional requirements. A case study demonstrates the framework's effectiveness through the development of an Autonomous Mobile Cobot by integrating two pre-existing robotic systems. The approach ensures requirement coherence while significantly reducing development time and cost.

Keywords: CAD , Human-robot collaboration , Multibody modelling , Virtual Prototyping

The cornerstone of Industry 5.0 is the human, its well-being, development, and creativity at the center of the production process. To achieve this, recording emotional, psychological, physical, and cognitive states efficiently in real-time is crucial. In particular, monitoring a complex psychophysiological state such as stress requires obtaining information from reliable biological signals such as electrocardiogram (ECG), galvanic skin response (GSR), electroencephalogram (EEG), and facial expressions. In this study, we introduce a new dataset, SenseCobotFusion, which collects stress-related metrics derived from physiological signals recorded from operators engaged in Human-Robot Collaboration (HRC) tasks. Labeled with the subjective operator rating obtained with the NASA-TLX questionnaire tool, SenseCobotFusion is a new dataset available to the research community focused on stress and workload detection. SenseCobotFusion is structured to be flexible and compatible with other existing datasets and potential experimental scenarios. To achieve this, a thorough dataset encompassing all metrics and specific sub-datasets for each signal type was developed, enabling seamless adaptation to user needs. As a tentative example of the potential of SenseCobotFusion, machine learning models were trained on each of these datasets. The results align with findings in the literature: the stress response is highly subjective and influenced by numerous factors, both dependent and independent of the operator. Additionally, the signal processing pipeline codes used to extract the metrics of interest, specific for GSR, EEG, ECG, and Emotions data, was also provided, which can be used as a guideline to extract stress-related metrics from SenseCobot and similar datasets.

Keywords: Human-Robot Collaboration , Machine Learning , SenseCobot , Stress Evaluation

In this article, we address the development of a robotic rehabilitation system for the upper limbs based on collaborative end-effector solutions. The use of commercial collaborative robots offers significant advantages for this task, as they are optimized from an engineering perspective and ensure safe physical interaction with humans. However, they also come with noticeable drawbacks, such as the limited range of sizes available on the market and the standard control modes, which are primarily oriented toward industrial or service applications. To address these limitations, we propose an optimization-based design method to fully exploit the capability of the cobot in performing rehabilitation tasks. In addition, we introduce a novel control architecture based on an admittance-type virtual fixture method, which constrains the motion of the robot along a prescribed path. This approach allows for an intuitive definition of the task to be performed via programming by demonstration and enables the system to operate both passively and actively. In passive mode, the system supports the patient during task execution with additional force, while in active mode, it opposes the motion with a braking force. Experimental results demonstrate the effectiveness of the proposed method.

Keywords: Assist-as-needed control , collaborative robotics , human-robot interaction , learning by demonstration (LbD) , upper limb rehabilitation , virtual fixtures

The Internet of Things (IoT) paradigm has transformed the establishment of interconnected technology settings, paving the way for novel approaches to intelligent service management. Smart Cities and Smart Homes are just a few examples of how this technology might improve people's lives. This study proposes a novel way to Smart Home management that combines the Internet of Things with graphic representation and machine learning techniques. Our technology enables you to collect and analyze data from sensors and other devices, resulting in relevant information and autonomous actions that optimize energy and ventilation control.

Keywords: Energy Optimization , Environmental Monitoring , Internet of things , Smart Home , User-Centered Automation