6949 research outputs found
Sort by
Prospective perception through cognitive emulation for robot manipulation tasks: "Perceiving like humans do"
No full textThis thesis argues that bottom-up theories of perception suffer from the high semantic entropy arising from the severe spatial, temporal, and informational limitations of sensory input during everyday manipulation tasks. However, by emulating the “dark matter” of perception — including intent, functionality, utility, causality, and physis — and integrating it with sparse sensory data, robotic perception can achieve a causal, transparent, and computationally efficient ability to anticipate and explain relevant events in such tasks.
To this end, the thesis introduces Probabilistic Embodied Scene Grammars (PESGs) to formalize this perceptual “dark matter.” It also presents a generator and a parser to respectively anticipate and explain event-centric scenes. The approach is demonstrated in complex real-world scenarios, including household tasks such as pancake making in kitchen environments, shopping tasks in supermarkets, and sterility testing tasks in medical laboratories
Effekte pyrolytischer Pflanzenkohle auf Bodeneigenschaften, Collembolen, Milben und Keimlingsentwicklung
No full textAbstract: Worldwide soils are degrading, reducing the suitable habitat for soil organisms. The loss of soil biodiversity in response adversely affects soil ecosystem functions and services which humans depend on. Pyrolytic biochar is produced by pyrolysis under oxygen-limited conditions, thereby converting biomass into a more degradation-resistant carbon. Its application to soils is intended to support environmental and soil management, while mitigating climate change through carbon sequestration. The present dissertation investigated the effects of a bio-activated, wood-based, pyrolytic biochar (BCa) on soil properties, soil mesofauna and early seed performance of plants. Therefore, I conducted one field experiment and three in the laboratory using soil and mesofauna from the fiele experiment test area (a temperate grassland in Germany). Key aspects and findings of this thesis – that mesofauna are an abundant and essential part of soil ecosystems, that they can affect soil properties and seed performance (e.g., increase the seed germination rate), that they can be affected by BCa, but themselves can also influence the effects of BCa (although their influence may depend on other factors such as soil moisture, plant species or the soil additive) - suggest a high relevance of the effects observed in this thesis for grassland ecosystems where BCa is applied, with consequences from the species level up to the community level. This thesis has also shown that the characteristics of the observed effects (presence, magnitude, direction) may depend on several other factors, that there is a potential for beneficial effects, but also for detrimental effects, and that effects of BCa on Collembola can change from positive to negative over time.
Zusammenfassung: Weltweit verschlechtert sich Bodenzustand, wodurch der für Bodenorganismen geeignete Lebensraum abnimmt. Der damit einhergehende Verlust der biologischen Vielfalt im Boden beeinträchtigt die Ökosystemfunktionen und -leistungen des Bodens, auf welche der Mensch angewiesen ist. Pyrolytische Pflanzenkohle wird unter Sauerstoff-limitierten Bedingungen durch Pyrolyse hergestellt, wobei Biomasse in einen abbauresistenteren Kohlenstoff umgewandelt wird. Ihre Anwendung im Boden soll das Umwelt- und Bodenmanagement unterstützen und zugleich den Klimawandel durch die Bindung von Kohlenstoff im Boden abschwächen. Die vorliegende Dissertation untersuchte die Effekte einer bio-aktivierten, holz-basierten, pyrolytischen Pflanzenkohle (BCa) auf Bodeneigenschaften, Bodenmesofauna und die Keimlingsentwicklung von Pflanzen. Hierzu führte ich ein Experiment im Feld und drei im Labor durch, mit dem Boden und der Mesofauna von dem Standort des Feldversuches (ein gemäßigtes Grünland in Deutschland).
Wichtige Aspekte und Erkenntnisse dieser Dissertation - dass die Mesofauna ein häufiger und wesentlicher Bestandteil von Bodenökosystemen ist, dass sie die Bodeneigenschaften und die Keimlingsentwicklung beeinflussen kann (z.B. Anstieg der Keimungsrate), dass sie zwar von BCa beeinflusst wird, aber auch selbst die Auswirkungen von BCa beeinflussen kann (wobei ihr Einfluss von anderen Faktoren wie Bodenfeuchtigkeit, Pflanzenart oder dem Bodenhilfsstoff abhängen kann) - deuten auf eine hohe Relevanz der in dieser Arbeit beobachteten Effekte für Grünlandökosysteme hin, in denen BCa eingesetzt wird; mit Konsequenzen von der Art- bis hin zur Gemeinschaftsebene. Diese Arbeit hat außerdem gezeigt, dass die Charakteristika der beobachteten Effekte (deren Vorhandensein, Größenordnung, Richtung) von mehreren anderen Faktoren abhängen können, dass es ein Potenzial für positive, aber auch für schädliche Wirkungen gibt und dass sich die Effekte von BCa auf Collembolen mit der Zeit von positiv zu negativ verändern können
Codebook of the Historical Database on Maternity Leave (HDML) (2nd updated and extended edition)
No full textThe Historical Dataset on Maternity Leave (HDML) provides harmonized, cross-national data on paid maternity leave policies across 153 independent nation-states with populations exceeding 500,000, covering the period from 1883 to 2020. Developed as part of the Global Welfare State Information System (WeSIS), the HDML captures comprehensive policy information on the existence, duration, benefit levels, eligibility criteria, coverage, and financing mechanisms of maternity leave programs over time. The dataset is particularly designed to enable historical and comparative analyses of maternity protection standards in light of the International Labour Organization’s Maternity Protection Conventions. Each observation corresponds to a country-year and includes both standardized and original-format variables. A consistent coding system for missing or inapplicable values ensures analytical transparency. This unique longitudinal resource facilitates research into the global development and diffusion of paid maternity leave policies.002
Employer branding in tourism: Generation Z's expectations of employers
No full textDiese Masterarbeit untersucht die spezifischen Erwartungen der Generation Z an Arbeitgeber in der Tourismusbranche und leitet daraus praxisorientierte Implikationen für das Employer Branding ab. Angesichts eines zunehmenden Fachkräftemangels sowie hoher Fluktuationsraten im Tourismussektor wird deutlich, dass traditionelle Ansätze der Arbeitgeberkommunikation nicht ausreichen, um junge Talente zu gewinnen und langfristig zu binden.
Auf Grundlage eines theoretischen Rahmens zu Markenführung, Employer Branding und den soziokulturellen Merkmalen der Generation Z wurde eine qualitative Studie mit leitfadengestützten Interviews durchgeführt. Die Ergebnisse zeigen, dass junge Beschäftigte insbesondere Flexibilität, transparente Kommunikation, flache Hierarchien sowie eine werteorientierte Unternehmenskultur erwarten. Ebenso gewinnen sinnstiftende Tätigkeiten, Entwicklungsmöglichkeiten sowie Authentizität entlang der gesamten Candidate Journey zunehmend an Bedeutung.
Die Arbeit identifiziert zentrale Handlungsfelder für Unternehmen im Tourismus: Eine glaubwürdige Positionierung der Arbeitgebermarke, gezielte Kommunikation über digitale Kanäle und eine konsistente Umsetzung von Employer Branding-Maßnahmen in Rekrutierung, Onboarding und Mitarbeiterbindung. Best-Practice-Beispiele veranschaulichen, wie erfolgreiche Konzepte strategisch und operativ umgesetzt werden können.
Die Arbeit schließt mit Empfehlungen zur Ausrichtung des Employer Brandings entlang der Erwartungen der Generation Z und liefert Impulse für eine nachhaltige Positionierung touristischer Arbeitgeber im Wettbewerb um junge Talente
Uncertain Election Polls, Uncertain News Coverage: Two-Sided Heterogeneity in News Outlet–Pollster Relations
No full textElection polls and their news coverage function as a form of chained gatekeeping, shaping voters’ perceptions of fellow citizens’ preferences. The recent proliferation of polling firms equipped with low-cost, novel methods has not only multiplied competing results—confusing the perception of public opinion—but also compelled news outlets to selectively rely on trust-ed pollsters. To examine how outlets make these organizational choices, this study concep-tualizes poll coverage as an interorganizational relationship between outlets and polling firms. By evaluating poll diversity at the outlet level, we find that highly institutionalized outlets adhere to journalistic norms by diversifying their coverage, yet still exhibit a prefer-ence for legacy pollsters. Further, under conditions of dual heterogeneity, we show that emergent outlets tend to increase the visibility of less credible pollsters, thereby contributing to a more fragmented news ecosystem.5
"Da musst du dann irgendwo Prioritäten setzen" : Theorie pflegerischer Verteilungsentscheidungen in der stationären Langzeitpflege
No full textDie Theorie pflegerischer Verteilungsentscheidungen in der stationären Langzeitpflege wurde auf Basis einer qualitativen Untersuchung mit Hilfe der Methodologie der Grounded Theory (GTM) entwickelt. Sie stellt die komplexen Ursachen, Bedingungen, Kontexte, Umsetzungsstrategien und Wirkungen von Verteilungsentscheidungen Pflegender dar. Die daraus entwickelte Typologie des Entscheidungshandelns Pflegender enthält eine Beschreibung und Bewertung der Deutungs- und Handlungsmuster von Pflegenden in der stationären Langzeitpflege in herausfordernden beruflichen Verteilungssituationen. Mit dem Typ des flexibel-reflexiven Entscheidens wird dabei auch ein Entscheidungstyp vorgestellt, der normativ als anzustrebender Typ professionellen Pflegehandelns in Verteilungssituationen verstanden werden kann und zu guten Verteilungsentscheidungen in der stationären Langzeitpflege beiträgt
Ice Sheet evolution in Patagonia and Northern Europe during the last ice age - Insights from numerical modelling
No full textDuring the Last Glacial Maximum (LGM, ~23–19 ka), large ice sheets covered much of the Northern Hemisphere and southern South America. Geochronological data indicate contrasting behaviours between the Patagonian Ice Sheet (PIS) and the Eurasian Ice Sheet Complex (EISC): the PIS reached its maximum during Marine Isotope Stage (MIS) 3, while the EISC expanded to its maximum during the LGM. This dissertation employs numerical ice sheet modelling (SICOPOLIS) and surface mass balance (SMB) analysis to explore the climatic conditions influencing these ice sheets across the last glacial cycle. Results show the PIS advanced during MIS4 and late MIS3–MIS2, driven by variations in integrated summer energy at mid-latitudes, with millennial-scale fluctuations linked to Dansgaard-Oeschger events. The study reveals that climate model limitations, such as low resolution and oversimplified topography, lead to inaccurate PIS simulations, often misrepresenting ice growth patterns. Offshore SST records from the southeastern Pacific more accurately capture the MIS3 expansion than Antarctic core-derived indices. For the EISC, SMB simulations highlight the dominant role of summer temperature in shaping maximum extent. Regional analysis shows strong correlations between SMB in different ice sheet sectors and climate conditions over the Greenland-Iceland-Norwegian and Labrador Seas. These findings underscore the need for high-resolution climate models and robust proxy integration to refine reconstructions of past ice sheet dynamics
Deep learning for temporal reconstruction of FESOM-derived sea surface temperature and 3D ocean variables
No full textIn oceanography, the growing volume of data collection raises the challenge of efficient data storage and computational efficiency for ocean simulations. To address this, we adopted an interpolation-based approach for reducing data storage requirements while ensuring accurate reconstruction of sea surface temperature (SST) and 3D ocean variables, including temperature and horizontal velocities (u and v) derived from the Finite-volumE Sea ice–Ocean Model (FESOM2).
Our research explores three distinct strategies: single-step interpolation, temporal-window interpolation, and multi-step interpolation. To achieve these goals within a supervised learning framework, we propose a hybrid CNN-BiLSTM deep learning model for both accurate interpolation and data storage optimization. This model extracts features from time-series data by combining the strengths of convolutional neural networks (CNNs) for capturing spatial patterns and bidirectional long short-term memory (BiLSTM) networks for capturing temporal dependencies in time-series data. Additionally, our approach uses a multi-input multi-output (MIMO) strategy to improve computational efficiency and prevent error accumulation, resulting in high prediction accuracy. By incorporating neighborhood information and preprocessing data to consider spatial and temporal dependencies, our approach effectively prepares the data for training. We use subsampling to reduce memory usage and introduce diverse, uncorrelated samples.
The results show that our model outperforms traditional methods, including linear interpolation (LI) and linear regression (LR), achieving lower mean squared error (MSE) and higher correlation coefficients when reconstructing ocean variables. Single-step interpolation achieved a 75% improvement, temporal-window interpolation demonstrated an overall 46.80% improvement, and multi-step interpolation achieved an overall 59.51% improvement in predictive accuracy. Remarkably, the model predicts multiple time steps for 3D ocean fields without introducing artifacts, achieving this within a single model framework.
After training on diverse datasets, the model generalizes well to unseen data without requiring additional computational resources, making it a scalable and sustainable solution for addressing the growing need for data storage efficiency in oceanographic research. To improve model performance, key techniques were applied, including overfitting prevention, balanced batching, multi-GPU training, adaptive learning rate schedules such as polynomial decay and cyclic learning, and fine-tuning.
We also demonstrate the practical application of our model by computing ocean heat transport in zonal and meridional directions. The results confirm that our model predicts ocean heat content more accurately than traditional methods. However, the computation of ocean heat transport using mixed-sign velocity values presents challenges, despite individual velocity predictions being highly accurate. This work highlights the potential of advanced neural networks in reconstructing temperature and velocity data, accurately matching observed values, and enhancing our understanding of ocean heat dynamics
Bio-inspired load-bearing strategies in engineering design
No full textThis dissertation investigates the enhancement of load-bearing engineering designs through the application of nature's structural principles. The study begins by exploring stiffness optimization methods inspired by the complex structures of diatoms, leveraging these organisms' high strength-to-weight ratios to produce highly performant engineering parts. It then progresses to examine the effectiveness of naturally prevalent TPMS (Triply Periodic Minimal Surface) lattices in stiffness optimization through homogenization. Benchmarking these lattices against other types reveals their superior homogeneity, load-path efficiency, and adaptability to varying boundary conditions, leading to enhanced performance in practical design scenarios. The research documents the effectiveness of the biological design principles studied and outlines strategies for their application in engineering, underscoring the significant potential of incorporating more natural principles into structural design to develop efficient and optimized components
Rechtliche Dimension Virtueller und Erweiterter Realität
No full textThese project and conference proceedings are based on the research project “LEG-ART-CHIP“ of the Department of Law at the University of Genoa and the Hugo Grotius gGmbH (2019-2025) as well as on the workshop “Legal Dimension of Virtual and Augmented Reality“ held on 30 June 2022 at the University of Genoa, and have been supplemented by additional contributions from various authors on this topic