University of Modena and Reggio Emilia
Archivio istituzionale della ricerca - Università di Modena e Reggio EmiliaNot a member yet
116400 research outputs found
Sort by
Intuitive Control of a Social Robot Using Natural Language with a Large Language Model and Error Correction Capabilities
Get PDFIn this paper, a novel framework for interacting with the social robot NAO using natural language commands is proposed. The motivation for this work arises from the needs of operators at the PASCIA Center (Heart Failure Care Program, Childhood Heart Diseases, and Those at Risk), a specialized department at the Polyclinic of Modena focused on childhood heart diseases. They aim to use the NAO robot during cardiology visits with young patients affected by autism spectrum disorders. To overcome the barrier that prevents unskilled operators from independently controlling the robot, a new interface has been developed. By leveraging the capabilities of Large Language Models, particularly ChatGPT, operators’ requests expressed in natural language can be transformed into concrete actions executed by the robot. Additionally, a mechanism has been implemented to assist doctors in adjusting and correcting ChatGPT’s responses based on a feedback loop. The results have been evaluated using accuracy metrics to compare successful and failed tasks performed by the robot. The findings demonstrate the framework’s efficiency and usability
Realising human-robot collaboration in manufacturing? A journey towards industry 5.0 amid organisational paradoxical tensions
Get PDFHuman-robot collaboration is envisioned as a cornerstone of the future ‘ideal’ industry (Industry 5.0)—resilient,
sustainable, and human-centred. While this goal has not yet been fully realised, advancements in collaborative
robotic technology are expected to accelerate progress. Central to this vision is a workforce equipped with the
skills necessary to collaborate effectively with robots in ‘ideal’ hybrid teams. Existing literature widely supports
this optimistic outlook, suggesting that with the right technological developments, workforce reskilling and
upskilling, and the resolution of key ethical and social concerns, human-robot collaboration in manufacturing
will eventually become a reality. In this paper, we draw on a one-year field study that engaged with 39 representatives
from industry, research, and other key stakeholders in both the technical and human factors of
collaborative applications. Using constructive grounded theory and abductive reasoning, we challenge the
assumption that the trajectory towards human-robot collaboration is straightforward or can be resolved through
a one-time solution. Instead, our results reveal a journey marked by a series of paradoxical tensions, providing a
fresh perspective on the complexities and unexpected empirical ‘surprises’ that define the transition towards
Industry 5.0. We employ Paradox Theory to examine and elucidate this evolving journey, where paradoxes—
such as automation vs augmentation, technical efficiency vs human wellbeing, and exploration vs exploitation—
emerge, shift, and are managed in unexpected ways, revealing interdependencies between different types
of responses across micro, meso, and macro levels of analysis. Extending beyond current theorisations on the
implementation of Industry 5.0, our study contributes substantively and theoretically to understanding the
evolving socio-technical complexities that shape this transition, highlighting the interplay between technological
advancements, organisational dynamics, and workforce adaptation
Structural Investigations on Yokeless Electrically-Excited Segmented Armature Axial Flux Motor
No full textIMMERSIVE STORYTELLING IN JOURNALISM. How Newsgames Are Changing the News Landscape
No full textJournalism is constantly undergoing profound transformations, and now it is also intersecting and colliding with the gamification of digital spaces. Driven by interactivity, multimedia, and global access to information, these changes are contextualized within the broader discourse on immersive digital technologies, including virtual and augmented reality, which have transformed storytelling into a participatory and multisensory experience. Among these innovations, newsgames have emerged as a groundbreaking format that combines game mechanics with journalistic
storytelling. This article explores how newsgames challenge traditional reporting by integrating interactive simulations, branching narratives, and procedural rhetoric, allowing users to navigate diverse perspectives and develop empathy while engaging with real-world issues. Through international case studies – such as September 12th, which critiques the “war on terror”, Syrian Journey by BBC News, and The Waiting Game by ProPublica, as well as examples from recent wars (such as The Flashback) – this paper demonstrates how newsgames deepen public awareness of global conflicts and humanitarian crises. By examining these examples, this article seeks to contribute to the expanding discourse on gamified storytelling, highlighting the importance of
multidisciplinary approaches in addressing the cultural, ethical, and cognitive implications of this evolving medium
Navigating Multiple Virtual Teams: How Variety in Communication Rules Affects Knowledge Sharing
Get PDFMolecular-Level Mechanistic Insights into PETase-Catalyzed Plastics Hydrolysis from Accurate QM/MM Free Energy Calculations
Get PDFThe enzyme PETase is capable of depolymerizing plastics such as polyethylene terephthalate (PET) at moderate temperatures, and demonstrated even higher activity toward polyethylene-2,5-furan dicarboxylate (PEF), opening promising routes for the efficient upcycling of plastic wastes. To fully exploit the potential of these biocatalytic systems, an understanding of the mechanism of their activity at the atomistic level is pivotal. To this end, this study investigates two fundamental stages of the catalytic cycle of PET and PEF hydrolysis by PETase-acylation and deacylation-using hybrid QM/MM enhanced sampling molecular dynamics simulations to capture all relevant dynamic effects. Well-tempered metadynamics simulations at the DFTB3 level are performed along collective variables optimized via linear discriminant analysis, a supervised learning-assisted approach that accounts for the contributions of each potentially relevant degree of freedom. The free energy (FE) profiles indicate that the acylation stage is the rate-limiting step for both PET and PEF degradation, with barriers approximate to 8 kcal/mol and approximate to 4 kcal/mol higher than those obtained for the deacylation step, respectively. Remarkably, substantial mechanistic differences are found. While PET acylation occurs in a concerted manner, with a single energy barrier of approximate to 21 kcal/mol, PEF acylation follows a two-step mechanism where after the first barrier, approximate to 10 kcal/mol high, a metastable intermediate state is formed, which then evolves toward the product once a second barrier of approximate to 2 kcal/mol is overcome. This mechanistic description is consistent with the FE profiles obtained at higher levels of theory (PBE, B3LYP, RI-MP2) via FE perturbation, thus validating the key insights elucidated by metadynamics simulations. Finally, both global and local reactivity descriptors derived from conceptual density functional theory suggest that PEF is more electrophilic and susceptible to nucleophilic attack than PET. The results obtained by means of the robust computational protocol adopted here offer thermodynamic and mechanistic insights into PET and PEF hydrolysis by PETase at the molecular level, corroborating the experimentally observed enhanced activity of this enzyme toward PEF. The distinctive hallmarks of PETase depolymerization uncovered in this work provide valuable foundations for enzyme engineering efforts aimed at developing universal biocatalysts for semiaromatic plastic recycling, ultimately paving the way for efficient application in industrial settings
MininetGym: A modular SDN-based simulation environment for reinforcement learning in cybersecurity
No full textUterine atony and postpartum haemorrhage: predisposing genetic factors and postmortem findings
Get PDFBoronate-Based Inhibitors of Penicillin-Binding Proteins: An Underestimated Avenue for Antibiotic Discovery?
Get PDFPenicillin-binding proteins (PBPs) are essential enzymes involved in bacterial cell wall biosynthesis and represent the primary targets of β-lactam antibiotics. However, the efficacy of these agents is threatened by β-lactamase production and PBP alterations, prompting the search for alternative strategies. In this context, boronic acids, long established as potent inhibitors of serine β-lactamases (SBLs), have been proposed as scaffolds for PBP inhibition based on the shared structural and mechanistic features of these enzyme families. This perspective provides a literature-based survey with structural analysis to evaluate emerging evidence on the potential role of boronic acids as PBP-targeting agents, with a particular focus on peptidomimetic boronic acids, repurposed β-lactamase inhibitors, and novel scaffold architectures. While early work showed limited activity against low-molecular-mass PBPs, more recent compounds, particularly certain bicyclic boronates, have demonstrated potent binding and, in some cases, antibacterial activity. Structural analyses reveal diverse binding modes and underscore the role of conformational dynamics in modulating affinity. Despite these advances, significant challenges remain, including target selectivity, membrane permeability, and species-specific differences. Nevertheless, the direct inhibition of PBPs by boronic acids, while still in early development, may offer a viable complement or alternative to β-lactam therapy, warranting further exploration through structure-guided design and comprehensive biological evaluation. Here, we analyze the potential of boronic acid inhibitors (BAIs) to target PBP enzymes, considering their promise as non-β-lactam antimicrobial agents with possible clinical relevance
La prevenzione del rischio di discriminazione nel management algoritmico nel diritto del lavoro europeo
No full textNumerose evidenze empiriche mostrano che la diffusione dei sistemi di gestione algoritmica delle risorse umane solleva preoccupazioni di carattere etico e acuisce il rischio di discriminazioni sistemiche.
Le discriminazioni che ne risultano, qualificate in dottrina come “discriminazioni algoritmiche”, hanno generalmente carattere collettivo e indiretto. Tali elementi, se considerati unitamente alla tipica opacità, manifestano l’inidoneità del modello rimediale predisposto dal diritto antidiscriminatorio europeo. A tal proposito, il presente studio ha l’obiettivo di analizzare gli strumenti legislativi che aspirano ad una tutela collettiva, partecipativa e preventiva avverso le discriminazioni, in quanto l’approccio che ne deriva risulta particolarmente adeguato a contrastare i rischi della digitalizzazione del lavoro. L’indagine si concentra prima sugli strumenti di attivazione predisposti dal diritto antidiscriminatorio europeo e, in seguito, si espande fino a considerare le connessioni tra la disciplina antidiscriminatoria e la normativa prevenzionistica sulla sicurezza e sulla salute sul lavoro