Recognizing cultural concepts: Joyce, Woolf, Mann and Musil

The purpose of this thesis is to establish a direct relationship between literature and fields of knowledge such as science and technology, by focusing on some concepts that were fundamental for both science and the humanities at the beginning of the 20th century. The concepts are those of simultaneity, multiple points of view, map, relativity and acausality. In the spirit of several recent ideas, for example Katherine Hayles’ isomorphism notion, the dissertation shows how writers such as James Joyce, Virginia Woolf, Thomas Mann and Robert Musil developed the mentioned concepts within their narratives. The working hypothesis is that those concepts were at a crossroad of human activities, and that those authors used them extensively within their narratives. It is further argued that those same concepts – as developed by Joyce in Ulysses, Woolf’s shorts stories and novels from the end of the 1910’s until the end of the1920’s, Mann’s Der Zauberberg (The Magic Mountain), and Musil’s Der Mann ohne Eigenschaften (The Man Without Qualities) — are still fundamental for our conception of time and space today. The thesis is divided into two parts. The first two chapters will analyse the concepts of simultaneity and multiple points of view and their relationship to cartography as developed within English literature and culture. The next two chapters will address the concepts of relativity and acausality, as developed within German literature and culture.

Analyzing urban green adaptation opportunities: concepts, approaches, & strategies for existing neighborhoods

Over the last decades, the growing evidence of human-caused climate change has raised awareness of the consequences of exceeding global temperature by 2˚C. This awareness has led to a contemporary approach to the conceptualization and management of green adaptation policies in spatial planning. This thesis aims to develop a comprehensive methodology for assessing the adaptability of existing neighborhoods to green strategies. The reliability of the proposed method is examined in the cities of Bologna and Imola and proved to be applicable in other geoghraphical locations. This thesis integrates three key themes of conceptual and implementation principles for urban green adaptation. This thesis initially defines methods for narrowing uncertainties in urban planning energy forecasting modeling by exploring the roles of integrated energy planning. The second is by exploring green retrofitting strategies in building, this thesis examines the effects of various energy-saving factors in roofing scenarios including a green roof, rooftop greenhouse, and insolated roof. Lastly, this thesis analyzes green strategies in urban spaces to enhance thermal comfort through facing urban heat exposure related to urban heat island effects. The roles of integrated energy policies and green strategic thinking are discussed to highlight various aspects of green adaptation on the neighborhood scale. This thesis develops approaches by which cities can face the challenges of current green urban planning and connect the conceptual and practical aspects of green spatial planning. Another point that this thesis highlight is that due to the interdependency of individuals and places, it is difficult to assure whether all the adaptation policies on a large scale are enhancing the resiliency of the neighborhood or they are simply shuffling the vulnerability through the individuals and places. Besides, it asserts that neglecting to reflect on these reallocations of the effects generates serious complications, and will result in long-term dysfunctional consequences.

Towards eco-friendly batteries: concepts for lithium and sodium ion batteries

Several possibilities are arising aiming the development of “greener”, more sustainable energy storage systems. One point is the completely water-based processing of battery electrodes, thus being able to renounce the use of toxic solvents in the preparation process. Despite its advantage of lower cost and eco-friendlyness, there is the need of similar mechanical and electrochemichal behavior for boosting this preparation mode. Another point – accompanying the water-based processing - is the replacement of solvent-based polymer binders by water-based ones. These binders can be based on fluorinated, crude-oil based polymers on the one side, but also on naturally abundant and economic friendly biopolymers. The most common anode materials, graphite and lithium titanate (LTO), have been subjected a water-based preparation route with different binder systems. LTO is a promising anode material for lithium ion batteries (LIBs), as it shows excellent safety characteristics, does not form a significant SEI and its volume change upon intercalation of lithium ions is negligible. Unfortunately, this material suffers from a rather low electric conductivity - that is why an intensive study on improved current collector surfaces for LTO electrodes was performed. In order to go one step ahead towards sustainable energy storage, anode and cathode active materials for a sodium ion battery were synthesized. Anode active material resulted in a successful product which was then subjected to further electrochemical tests. In this PhD work the development of “greener” energy storage possibilities is tested under several aspects. The ecological impact of raw materials and required battery components is examined in detail.

Variety, flexibility, and use of abstract concepts. A multiple grounded perspective.

The nature of concepts is a matter of intense debate in cognitive sciences. While traditional views claim that conceptual knowledge is represented in a unitary symbolic system, recent Embodied and Grounded Cognition theories (EGC) submit the idea that conceptual system is couched in our body and influenced by the environment (Barsalou, 2008). One of the major challenges for EGC is constituted by abstract concepts (ACs), like fantasy. Recently, some EGC proposals addressed this criticism, arguing that the ACs comprise multifaced exemplars that rely on different grounding sources beyond sensorimotor one, including interoception, emotions, language, and sociality (Borghi et al., 2018). However, little is known about how ACs representation varies as a function of life experiences and their use in communication. The theoretical arguments and empirical studies comprised in this dissertation aim to provide evidence on multiple grounding of ACs taking into account their varieties and flexibility. Study I analyzed multiple ratings on a large sample of ACs and identified four distinct subclusters. Study II validated this classification with an interference paradigm involving motor/manual, interoceptive, and linguistic systems during a difficulty rating task. Results confirm that different grounding sources are activated depending on ACs kind. Study III-IV investigate the variability of institutional concepts, showing that the higher the law expertise level, the stronger the concrete/emotional determinants in their representation. Study V introduced a novel interactive task in which abstract and concrete sentences serve as cues to simulate conversation. Analysis of language production revealed that the uncertainty and interactive exchanges increase with abstractness, leading to generating more questions/requests for clarifications with abstract than concrete sentences. Overall, results confirm that ACs are multidimensional, heterogeneous, and flexible constructs and that social and linguistic interactions are crucial to shaping their meanings. Investigating ACs in real-time dialogues may be a promising direction for future research.

Concepts and methods for efficient decentralized learning in federated settings

Deep Neural Networks (DNNs) have revolutionized a wide range of applications beyond traditional machine learning and artificial intelligence fields, e.g., computer vision, healthcare, natural language processing and others. At the same time, edge devices have become central in our society, generating an unprecedented amount of data which could be used to train data-hungry models such as DNNs. However, the potentially sensitive or confidential nature of gathered data poses privacy concerns when storing and processing them in centralized locations. To this purpose, decentralized learning decouples model training from the need of directly accessing raw data, by alternating on-device training and periodic communications. The ability of distilling knowledge from decentralized data, however, comes at the cost of facing more challenging learning settings, such as coping with heterogeneous hardware and network connectivity, statistical diversity of data, and ensuring verifiable privacy guarantees. This Thesis proposes an extensive overview of decentralized learning literature, including a novel taxonomy and a detailed description of the most relevant system-level contributions in the related literature for privacy, communication efficiency, data and system heterogeneity, and poisoning defense. Next, this Thesis presents the design of an original solution to tackle communication efficiency and system heterogeneity, and empirically evaluates it on federated settings. For communication efficiency, an original method, specifically designed for Convolutional Neural Networks, is also described and evaluated against the state-of-the-art. Furthermore, this Thesis provides an in-depth review of recently proposed methods to tackle the performance degradation introduced by data heterogeneity, followed by empirical evaluations on challenging data distributions, highlighting strengths and possible weaknesses of the considered solutions. Finally, this Thesis presents a novel perspective on the usage of Knowledge Distillation as a mean for optimizing decentralized learning systems in settings characterized by data heterogeneity or system heterogeneity. Our vision on relevant future research directions close the manuscript.

Two stroke cycle, novel combustion concepts and electrification for a new generation of internal combustion engines

The pursuit of decarbonization and increased efficiency in internal combustion engines (ICE) is crucial for reducing pollution in the mobility sector. While electrification is a long-term goal, ICE still has a role to play if coupled with innovative technologies. This research project explores various solutions to enhance ICE efficiency and reduce emissions, including Low Temperature Combustion (LTC), Dual fuel combustion with diesel and natural gas, and hydrogen integration. LTC methods like Dual fuel and Reactivity Controlled Compression Ignition (RCCI) show promise in lowering emissions such as NOx, soot, and CO2. Dual fuel Diesel-Natural Gas with hydrogen addition demonstrates improved efficiency, especially at low loads. RCCI Diesel-Gasoline engines offer increased Brake Thermal Efficiency (BTE) compared to standard diesel engines while reducing specific NOx emissions. The study compares 2-Stroke and 4-Stroke engine layouts, optimizing scavenging systems for both aircraft and vehicle applications. CFD analysis enhances specific power output while addressing injection challenges to prevent exhaust short circuits. Additionally, piston bowl shape optimization in Diesel engines running on Dual fuel (Diesel-Biogas) aims to reduce NOx emissions and enhance thermal efficiency. Unconventional 2-Stroke architectures, such as reverse loop scavenged with valves for high-performance cars, opposed piston engines for electricity generation, and small loop scavenged engines for scooters, are also explored. These innovations, alongside ultra-lean hydrogen combustion, offer diverse pathways toward achieving climate neutrality in the transport sector.