Higher Institute on Territorial Systems for Innovation
PORTO@iris (Publications Open Repository TOrino - Politecnico di Torino)Not a member yet
146173 research outputs found
Sort by
Real-Time TEM Observation of the Microstructural Evolution in Silver Nanowires under Heating and Electrical Biasing
Get PDFSilver nanowires (Ag NWs) are of interest for a variety of emerging technologies, such as transparent electrodes, nanoscale heaters, and neuromorphic devices, thanks to their excellent electrical conductivity, flexibility, and tunable nanoscale properties. However, the current understanding of the phenomena underpinning their behavior under electrical stimulation and heating, including failure and reconfiguration effects, is largely based on lab-scale device measurements, offering only indirect insights into the underlying mechanisms. In this work, in situ biasing and heating transmission electron microscopy imaging are performed on individual Ag NWs to directly investigate their morphological and structural evolution under controlled electrical and thermal stress in a vacuum. The results indicate that electrical NW breakdown is dominated by electromigration and localized Joule heating, leading to nanogap formation primarily at the cathode, while thermal decomposition proceeds more gradually along the crystallographic planes. They also provide direct evidence of rewiring phenomena, i.e., the electrically induced reconnection of a previously broken NW, highlighting the self-healing, adaptive, and memristive behavior of the NW under the action of an applied electrical stimulation. Altogether, this work offers fundamental insights into failure and reconfiguration mechanisms at the single NW level, informing the design of Ag NW-based components for flexible electronics, sensors, and neuromorphic systems
VAE-Semantic-Mapping: Analisi Latente e Clusterizzazione Spontanea di Corpus Testuali tramite Variational Autoencoders
Get PDFQuesto repository contiene l'implementazione di un modello di Intelligenza Artificiale basato su Variational Autoencoder (VAE) progettato per l'analisi semantica e la clusterizzazione di dataset testuali. Il progetto esplora la capacità di un'architettura generativa di far emergere strutture logiche (cluster) partendo da un embedding semplificato, misurando la "purezza" della classificazione attraverso il calcolo dei baricentri nello spazio latente
Prototipo di banco prova per esoscheletri industriali
No full textNEGLI ULTIMI ANNI VI È STATA UN’EVOLUZIONE RAPIDA
DELLE TECNOLOGIE ASSISTIVE, IN PARTICOLARE DEGLI
ESOSCHELETRI. LO STUDIO QUI PRESENTATO RIGUARDA
L’OTTIMIZZAZIONE DI UN BANCO PROVA PER ESOSCHELETRI
INDUSTRIALI, CONTROLLATO TRAMITE TRE CILINDRI
PNEUMATICI A DOPPIO EFFETTO E CON POSSIBILI VARI
CIRCUITI ALTERNATIVI UTILIZZANTI ELETTROVALVOL
L'Emergenza dell'Asse Chimico nel Gesso: Un Approccio basato su Variational Autoencoder per la Definizione dello Pseudospettro e della Polarità tra Fasi Idratate e Anidre
No full textIl presente lavoro analizza l'applicazione di architetture neurali di tipo Variational Autoencoder (VAE) per l'elaborazione di dati spettroscopici Raman relativi al sistema gesso-anidrite. Ispirandosi alle dinamiche di apprendimento delle intelligenze artificiali linguistiche — dove concetti astratti come la "polarità emotiva" emergono spontaneamente come dimensioni nello spazio dei vettori — abbiamo dimostrato come un modello VAE sia in grado di isolare autonomamente un asse chimico dell'idratazione. Attraverso la compressione dei dati in uno spazio latente bidimensionale, il modello organizza gli spettri non in modo caotico, ma secondo un gradiente fisico coerente. Questo processo permette l'estrazione dello pseudospettro, un archetipo vibrazionale pulito che funge da centroide statistico per l'identificazione minerale. I risultati mostrano una separazione netta tra i poli della fase idratata (gesso) e quella anidra (anidrite), validando l'uso delle strutture latenti come strumento predittivo e interpretativo per la mineralogia moderna
Enhanced thermoelectricity in nanowires with inhomogeneous helical states
No full textSemiconductor nanowires (NWs) with strong Rashba spin-orbit coupling (RSOC), when exposed to a suitably applied Zeeman field, exhibit one-dimensional helical channels with a spin orientation locked to the propagation direction within the magnetic energy gap. Here, by adopting a scattering-matrix approach applied to a tight-binding model of the NW, we demonstrate that the thermoelectric (TE) properties can be widely controlled by tuning the misalignment angle phi between the spin-orbit directions of two NW segments. In particular, when the RSOC vectors are antiparallel (Dirac-paradox configuration), we predict a significant violation of the Wiedemann-Franz law, and a strong enhancement of the Seebeck coefficient and the ZT figure of merit. We also show that the Zeeman gap determines the optimal energy window for doping and temperatures. These results suggest that controlling the spin-orbit field direction, which can be achieved with suitably applied wrap gates, is a promising alternative for tuning and optimizing the TE response in quantum-coherent semiconducting NW devices
Giovanni Muzio e il progetto per il "Villaggio SAFFA" a Pontenuovo di Magenta (MI), 1954-1962
No full textBetween the 1950s and the 1960s the Società Italiana Fabbriche Fiammiferi e Affini (SAFFA) entrusted the Milanese architect Giovanni Muzio (1893-1982) with the design of some public buildings intended for the working community of the industrial settlement of Pontenuovo di Magenta, in the province of Milan. The initiative, materialized in a small company town, offered Muzio the opportunity to look at similar episodes in Northern Europe, making the village of Pontenuovo an emblematic case of reception of the design culture of Germany, Scandinavia, and the Baltic regions in the Italian context. Through the analysis of archival sources and unpublished documents, the contribution focuses on the genesis of the settlement and on Muzio’s architectures, highlighting the links with possible Nordic and Germanic models and drawing attention to the condition of decay and abandonment in which the settlement has been lying since 2005, when SAFFA ceased its activity
Human-flood systems
No full textThis chapter reviews human-flood systems through a sociohydrological lens, highlighting the long-term dynamics and feedbacks that shape flood risk and its management. Flood risk management is reframed from short-term hazard reduction toward adaptive and resilient strategies supported by dynamic risk scenarios that capture phenomena such as levee and adaptation effects. Models that link individual behaviours with system-level processes, which represent explicitly the continuous feedbacks between floods and society, are reviewed. Progress in this field depends on interdisciplinary cooperation and methodological pluralism: integrating insights from critical social sciences enriches sociohydrological modelling with considerations of justice, power, and political economy, while combining system dynamics and agent-based models enables a holistic understanding of complex trajectories. Recognizing social heterogeneity is essential to diagnose vulnerability and ensure equitable adaptation. Sociohydrologic approaches also support participatory risk communication and the co-design of locally tuned strategies that are more effective than uniform, top-down measures. Finally, by incorporating human behaviour into flood forecasting and evacuation planning, sociohydrology demonstrates both immediate and long-term value for advancing dynamic, inclusive, and effective flood risk management
Characterization of the transient response of oscillating DBD plasma actuators for turbulent skin-friction control
No full textMachine Learning and Deep Learning for Cultural Heritage Conservation: A Bibliometric and Task-Oriented Review
Get PDFWith the rapid advancement of Artificial Intelligence (AI) technologies, Machine Learning (ML) and Deep Learning (DL) have become pivotal methods for driving the digital documentation, restoration, preservation, and preventive conservation of Cultural Heritage (CH). This paper constructs an integrated data + technology + task framework tailored for CH scenarios. It employs a combination of bibliometric analysis and systematic content study based on relevant literature published between 2011 and 2025. First, publication trends, sources of publication, global collaboration networks, and topic modeling reveal the overall landscape and evolutionary path of research on the digitization and intelligent transformation of CH. Subsequently, beginning with ML and DL systems, it summarizes classic workflows and outlines their applications in CH conservation. Concurrently, integrating topic modeling, existing research is categorized into three themes based on task attributes: Recognition, Reconstruction and Virtual Restoration, and Monitoring and Prediction. Representative literature, typical tasks, and technological trends within each theme are systematically outlined. Distinct from existing reviews, this study introduces a unified data technology task framework that explicitly links AI model paradigms to heritage specific constraints. Moving forward, by constructing high-quality heritage datasets, enhancing model interpretability, and exploring cross-model fusion approaches, AI technologies hold promise to play a more reliable and sustainable role in CH conservation, risk management, and digital dissemination
Air pollutant concentration fluctuations in an industrial site: A wind tunnel study
No full textThe accidental release of hazardous airborne pollutants on industrial sites creates risks associated with the exceedance of toxicity or explosivity limits. Capturing these risks requires predicting higher-order statistics of concentration fluctuations at various distances from the source. This challenge, already complex in atmospheric boundary layers, is further complicated by the typical built environment of industrial sites. To address this, we conducted wind-tunnel experiments on the dispersion of a passive scalar from a localized ground-level source within a reduced-scale model of an industrial site. The experiments measured the velocity and concentration fields, while varying the geometry of an upstream building simulating typical complex industrial structures. A key focus of our investigation is the one-point passive scalar concentration PDF, whose experimental realizations were systematically compared to three analytical models: the gamma, two-parameter Weibull and lognormal distributions. The gamma distribution generally provides the best predictions, although the lognormal model performs better within the building wake near the source. While the main discrepancies between theoretical distributions and experimental data consistently occur at low concentration values, all three distributions accurately predict the 95th and 99th concentration percentiles. Thus, peak and hazardous concentration levels can be reliably estimated even without fully capturing the complete concentration distribution