Virtual devices are a large attack surface of hypervisors. Vulnerabilities in virtual devices may enable attackers to jailbreak hypervisors or even endanger co-located virtual machines. While fuzzing has discovered vulnerabilities in virtual devices across both open-source and closed-source hypervisors, the efficiency of these virtual device fuzzers remains limited because they are unaware of the complex behaviors of virtual devices in general. We present Truman, a novel universal fuzzing engine that automatically infers dependencies from open-source OS drivers to construct device behavior models (DBMs) for virtual device fuzzing, regardless of whether target virtual devices are open-source or binaries. The DBM includes inter- and intra-message dependencies and fine-grained state dependency of virtual device messages. Based on the DBM, Truman generates and mutates quality seeds that satisfy the dependencies encoded in the DBM. We evaluate the prototype of Truman on the latest version of hypervisors. In terms of coverage, Truman outperformed start-of-the-art fuzzers for 19/29 QEMU devices and obtained a relative coverage boost of 34% compared to Morphuzz for virtio devices. Additionally, Truman discovered 54 new bugs in QEMU, VirtualBox, VMware Workstation Pro, and Parallels, with 6 CVEs assigned.

Fuzzing evolved into the most popular technique to detect bugs in software. Its combination with sanitizers has shown tremendous efficacy in uncovering memory safety errors, such as buffer overflows, that haunt C and C++ programmers. However, an important class of such issues, the so-called use-of-uninitialized-memory (UUM) errors, struggles to gain similar benefits from fuzzing endeavors. The only fuzzer-compatible UUM sanitizer available to date, MSan, requires that all libraries are fully instrumented. Unlike address sanitization, for which partial instrumentation results in false negatives (missed detection of bugs), UUM sanitizers require complete instrumentation to avoid false positives, hampering testing at scale. Yet, full-stack compiler-based instrumentation can be a daunting prospect for compatibility and practicality. As a result, many programs are left untested for UUM bugs.

In this paper, we propose an efficient multi-layer, opportunistic design that does not require (source-based) recompilation of all code without harming accuracy. The multiplicity of executions when fuzzing offers us the opportunity to learn what any encountered false positive looks like, and later ignore them when we meet them again with new test cases. Such an avenue is feasible only if one can resort to fast techniques to effectively discriminate candidate errors, or false negatives will then occur.

We show how to realize this design by using the dynamic binary translation of QEMU for compatibility and lightweight code analysis techniques to achieve scalability and accuracy. As a result, we obtain a fuzzer-friendly, performant sanitizer, QMSan, that effectively tackles current practicality challenges of UUM error detection. On a collection of 10 open-source and 5 proprietary programs, QMSan exposed 44 new UUM bugs. In our tests, QMSan incurs slowdowns of 1.51x over QEMU and 1.55x over the compiler-based instrumentation of MSan, showing no false positives and false negatives. QMSan is open-source.

Web browsers are ubiquitous and execute untrusted JavaScript (JS) code. JS engines optimize frequently executed code through just-in-time (JIT) compilation. Subtly conflicting assumptions between optimizations frequently result in JS engine vulnerabilities. Attackers can take advantage of such diverging assumptions and use the flexibility of JS to craft exploits that produce a miscalculation, remove bounds checks in JIT compiled code, and ultimately gain arbitrary code execution. Classical fuzzing approaches for JS engines only detect bugs if the engine crashes or a runtime assertion fails. Differential fuzzing can compare interpreted code against optimized JIT compiled code to detect differences in execution. Recent approaches probe the execution states of JS programs through ad-hoc JS functions that read the value of variables at runtime. However, these approaches have limited capabilities to detect diverging executions and inhibit optimizations during JIT compilation, thus leaving JS engines under-tested.

We propose DUMPLING, a differential fuzzer that compares the full state of optimized and unoptimized execution for arbitrary JS programs. Instead of instrumenting the JS input, DUMPLING instruments the JS engine itself, enabling deep and precise introspection. These extracted fine-grained execution states, coined as (frame) dumps, are extracted at a high frequency even in the middle of JIT compiled functions. DUMPLING finds eight new bugs in the thoroughly tested V8 engine, where previous differential fuzzing approaches struggled to discover new bugs. We receive $11,000 from Google’s Vulnerability Rewards Program for reporting the vulnerabilities found by DUMPLING.

Type confusion, or bad casting, is a common C++ attack vector. Such vulnerabilities cause a program to interpret an object as belonging to a different type, enabling powerful attacks, like control-flow hijacking. C++ limits runtime checks to polymorphic classes because only those have inline type information. The lack of runtime type information throughout an object’s lifetime makes it challenging to enforce continuous checks and thereby prevent type confusion during downcasting. Current solutions either record type information for all objects disjointly, incurring prohibitive runtime overhead, or restrict protection to a fraction of all objects. Our C++ dialect, type++, enforces the paradigm that each allocated object involved in downcasting carries type information throughout its lifetime, ensuring correctness by enabling type checks wherever and whenever necessary. As not just polymorphic objects but all objects are typed, all down-to casts can now be dynamically verified. Compared to existing solutions, our strategy greatly reduces runtime cost and enables type++ usage both during testing and as mitigation. Targeting SPEC CPU2006 and CPU2017, we compile and run 2,040 kLoC, while changing only 314 LoC. To help developers, our static analysis warns where code changes in target programs may be necessary. Running the compiled benchmarks results in negligible performance overhead (1.19% on SPEC CPU2006 and 0.82% on SPEC CPU2017) verifying a total of 90B casts (compared to 3.8B for the state-of-the-art, a 23× improvement). type++ discovers 122 type confusion issues in the SPEC CPU benchmarks among which 62 are new. Targeting Chromium, we change 229 LoC out of 35 MLoC to protect 94.6% of the classes that could be involved in downcasting vulnerabilities, while incurring only 0.98% runtime overhead compared to the baseline.

CMG2/ANTXR2 functions as a Collagen VI receptor and the primary portal for anthrax toxin entry. We find that CMG2 is regulated by ordered cycles of S-acylation and deacylation throughout its life cycle. Following synthesis in the endoplasmic reticulum, acylation by ZDHHC7 on two juxtamembrane cysteines protects folding intermediates from ER-associated degradation, resulting in a 5-fold increase in CMG2 biogenesis. The cytosolic thioesterase APT2 can remove these acyl chains, thereby controlling CMG2 levels. In the Golgi, CMG2 acylation by ZDHHC3 on a third cysteine to permit Arf6-dependent transport to the plasma membrane. At the cell surface, S-acylated CMG2 recruits APT2 in response to ligand binding, enabling release from the actin cytoskeleton and endocytosis. Accordingly, blocking APT2 suppresses the intracellular delivery of anthrax toxin, and inhibits CMG2-dependent Collagen VI degradation. These results define S-acylation-deacylation cycles as key regulators of CMG2 biogenesis and function, and highlight APT2 inhibition as a strategy to modulate CMG2 levels or prevent anthrax intoxication.

In response to current climate considerations, the geotechnical sector is increasingly exploring multiphysical approaches to develop sustainable engineering projects. This thesis focuses on two promising applications that combine the sector's decarbonisation objectives with geomechanical functionality: energy geostructures and biocementation. Multiphysical modelling has played a key role in their development and remains an essential tool to advance innovation adoption.

For energy geostructures, modelling and design frameworks are well established. The remaining challenge is to extend the domain of application beyond its well-known uses, primarily energy pile foundations in balanced climates, and demonstrate their effectiveness in alternative scenarios. In contrast, modelling frameworks for biocementation are often not validated at the large scale. Upscaling also continues to pose challenges, and recommendations for overcoming these are not readily available. These problems obstruct the step towards design principles.

In this setting, the thesis has a twofold objective: (i) to extend the range of applications for energy geostructures by moving beyond conventional considerations and (ii) to move towards developing design approaches for biocementation treatment. By building on existing knowledge and using multiphysical modelling as a central tool, the research offers new insights into the design, optimisation, and application of these technologies and aims to support future sustainable geotechnics.

Numerical analyses are used to look at three different settings for the application of energy geostructures. The potential of energy piles to provide cooling energy in hot-dominated climates is demonstrated through simulations, which show that, despite unbalanced thermal demands, temperatures stabilise over time and respect heat pump limitations. Simulations reveal that geothermal activation of an underground data centre can reduce internal air temperatures, and this effect can be used to optimise ventilation system performance but requires a case-by-case evaluation. Finally, models are used to understand the internal air dynamics of an energy metro station, and recommendations are provided for how such factors can be accounted for in its design. These works demonstrate that through modelling, the technology can be optimised to maximise its impact in providing renewable thermal energy.

The work then demonstrates how multiphysical modelling can aid in achieving standardisation of biocementation. A modelling framework is benchmarked against upscaling experiments to assess its performance. Recommendations for achieving homogeneity in biocementation soil improvement are provided, highlighting the benefits of using high, consistent injection rates and demonstrating the effect of using novel injection geometries. The framework is then adapted to simulate treatments using an ex situ hydrolysis method, revealing a shift in the governing precipitation mechanism from urea hydrolysis to calcium carbonate precipitation kinetics in relation to mixing patterns. Last, the effects of different treatment configurations for slope stabilisation via biocementation are analysed using both limit equilibrium and finite element methods. While biocementation can improve slope stability, its success largely depends on the manner of application. These findings highlight the contribution of modelling in developing effective approaches for biocementation treatment.

The electrochemical reduction of CO2 for producing carbon-neutral fuels and chemicals is a pivotal technology for the future decarbonization of the chemical and energy sectors. Among the various electrolyzer architectures, forward bias bipolar membrane (BPM) systems have emerged as a promising solution, combining high selectivity and CO2 utilization with a pure-water feed, salt-free design. However, performance limitations arising from water transport imbalance, membrane degradation, and kinetics overpotentials at the membrane interface hinder their industrial deployment.

This thesis addresses these challenges through systematic investigations across four key axes. The first part focuses on water transport, using synchrotron-based, operando X-ray tomography and diffusion measurements to analyze water management and hydration dynamics in a commercial BPM-based cell. It reveals that cathode flooding is not a limiting factor at the investigated current densities (up to about 200 mA cmâ 2), as the GDL saturation stays below 10%. Instead, insufficient water supply to the cathode and membrane over-swelling are identified as the dominant challenges for high current density operation.

The second part focuses on degradation mechanisms within the membrane-electrode assembly. Pre- and post-mortem analysis, again by X-ray tomography, reveals the evolution of structural damage. We show that membrane delamination occurs at current densities above 100 mA cmâ 2. To resolve this, a series of semi-porous BPMs with engineered gas-permeable anion exchange layers is designed and investigated. These structures enable enhanced back-transport of recombined CO2 toward the cathode, thereby relieving interfacial pressure buildup at the membrane junction. Among the tested designs, a microporous ionomer-nanoparticle composite layer proves particularly effective in suppressing membrane delamination under up to 200 mA cmâ 2 and reducing anode catalyst layer damage area by up to 90% compared to commercial BPMs.

The third part addresses the kinetic limitations at the BPM junction by integrating metal-oxide catalyst layers (e.g., TiO2, SiO2, IrO2) directly at the membrane interface. Complete catalyst coverage at 20â 30 ÎŒg/cm2 enables up to 100% higher current density at similar iR-corrected voltages, providing a scalable strategy to overcome interfacial limitations.

Finally, the fourth part investigates the role of the catalyst-layer microenvironment and its influence on reaction kinetics in cation-free systems. By designing a fully gas-fed setup with pretreated membranes, the study isolates and evaluates the impact of trace alkali cations migrating across the membrane. Furthermore, the use of anion exchange ionomers with high ion-exchange capacity (IEC) is found to promote beneficial double-layer capacitance and stabilize the catalytic interface over multi-hour operation. The results support a mechanism where both mobile and fixed cations shape the local environment on the silver catalyst surface, thereby enabling stable CO production under pure-water-fed, salt-free conditions.

Altogether, this work presents a comprehensive framework for enhancing the performance of a pure-water fed CO2 electrolyzer with a BPM in forward bias. By coupling advanced diagnostics with targeted material strategies, the thesis contributes to both fundamental understanding and practical design principles to guide the next generation of scalable and durable CO2 electrolysis systems.

The development of soft electronic and optoelectronic systems is essential for the growth and industrial deployment of research fields such as smart textiles and wearables. However, it remains challenging to identify materials that reconcile the required mechanical attributes (e.g. softness and stretchability) with functional metrics essential to develop high-performance devices (e.g. electrical conductivity or photoconductivity). Moreover, it is equally important to find processing routes that can produce such devices at large scale and low-cost while ensuring an accurate arrangement of materials with disparate electronic properties into complex architectures with small feature sizes.

In this thesis, we investigate the thermal drawing technique as a strategy to produce soft, multifunctional electronic and optoelectronic fibers. Originally developed for optical fibers, this process leverages the viscous flow of materials to transform macroscopic assemblies into long fibers while preserving the initial cross-sectional architecture. In particular, this work focuses on implementing the three fundamental pillars of soft optoelectronic devices within thermally drawn fibers: (i) Stretchable and electrically conductive materials serving as electrodes for charge collection and transport: Two composite systems relying on rigid and liquid fillers dispersed in a soft matrix are proposed, and their performance is demonstrated through various types of mechanical sensors. In particular, highly stretchable and conductance-stable electrodes are established in thermally drawn fibers by relying on liquid metal embedded elastomers. (ii) Transparent conductors that simultaneously enable light transmission and charge transport: The introduction of ionogels as a novel materials system in thermally drawn fibers is demonstrated. The versatility of this material enables a fine tuning of its mechanical, electrical and thermal properties to match the targeted attributes for specific applications. In particular, meters-long stretchable fibers encompassing a transparent and conductive core are produced, marking the first implementation of a transparent electrode in thermally drawn fibers. (iii) Stretchable semiconductors functioning as active layers with light-responsive properties: Blends of organic semiconductors with thermoplastic elastomers are envisioned as a facile route to produce soft semiconductors with low processing temperature that can be easily introduced into thermally drawn fibers. Each material system is first studied individually, then integrated into a single soft fiber demonstrating optoelectronic properties. This work paves the way towards the development of complex multi-functional soft fibers to fabricate smart textiles and wearables with advanced electronic and optoelectronic properties.

This thesis presents a comprehensive study on the application of composite magnets in two industrial scenarios: the development of a linear actuator and the enhancement of ski equipment with magnetic attachment systems. In both cases, non-uniform magnetization is employed, specifically in the form of a continuous adaptation of the Halbach array.

In the first application, the thesis explores the replacement of sintered magnets, typically used in permanent magnet motors, with composite magnets to reduce production costs while maintaining performance. The final prototype is functional but delivers a force reduced by 30% compared to the original. The thesis also presents an optimization of the stator's ferrous body, allowing for in-situ winding and further reducing manufacturing costs.

The second application aims to replace the traditional glue used with climbing skins with a flexible composite magnetic strip to enhance user experience. This technology is applied to two cases: cross-country skiing and mountaineering. For the latter, the results are promising but require further development, particularly in achieving sufficient magnetic force. In contrast, the cross-country skiing case result in a working product and to the development of a production line ready to supply the first batch of magnetic skins.

The thesis covers the design and optimization of magnetic patterns, as well as the development of a magnetizer tailored to this application, using finite element analysis and prototyping. For the cross-country case, a cooling system is developed to achieve a production rate that aligns with industrial needs.

The results of the thesis demonstrate the potential of composite magnets and the possibilities offered by unconventional magnetic patterns, paving the way for future advancements in magnetic material technology.

The Marshall Plan was the cornerstone of U.S. Cold War economic and foreign policy. It was a program of technical assistance to set up free trade through European integration, of mutual security to prevent the spread of communism through the Atlantic Alliance, and of information, education, and cultural exchange to promote the American Way of Life. Devoting specific attention to labor affairs to prevent strikes and communist tendencies among workers, Labor Division of the Marshall Plan advocated a transnational workers' housing program, notably in France, Greece, Italy, and Turkey as well as the Allied-controlled Austria and Allied-occupied Germany, for union-sponsored housebuilding and homeownership.

The program was framed around the discourse of non-communism, free labor, and the free world, disseminated jointly with the International Confederation of Free Trade Unions (ICFTU). Providing financial and technical assistance to trade unions for establishing and managing housing cooperatives, training construction labor in housing development and design, and industrializing building trades were key elements of this program, jointly organized with the ICFTU's European Regional Organisation, trade union federations and governments.

Union-sponsored cooperative housing was promoted for industrial workers as a means to assert authority and autonomy in housing provision, to build a home of one's own and a house with a garden, symbolizing upward mobility. This vision promised labor stability through union affiliation and mortgage loans, while advertising domestic ideals of a preindustrial way of life. Built on U.S. New Deal and wartime legacies on union-sponsored housing and advanced by Scandinavian models of non-profit housing associations, the program functioned as a multilateral U.S. Cold War Project, and grafted real-estate dynamics onto public housing, regardless of local policies and housebuilding traditions of the participating countries.

This thesis offers methodological insights to architectural history and theory. First, I survey non-architectural archives of governmental and multilateral organizations and develop my arguments using correspondence, memoranda, reports and "picture files" as well as photographs, press clippings, and a film script, rather than architectural archives and drawings. Second, I explore trade unions as authors of housing development, design and construction, and integrate social history into architectural historiography for a wider understanding of built environment production and its non-architect agents. The thesis also offers new historical findings and theoretical perspectives. First, I introduce the Marshall Plan's workers' housing program, overlooked except some country-specific studies on its involvement in housing and domesticity. Second, I present the U.S. architect-planner Donald Monson and U.S. labor advisors as key actors, along with previously unexamined U.S. housing consultants. Third, I discuss self-help housing as a tool of postwar imperialism in Western Europe, beyond postcolonial frameworks dividing the "global west/north" from the "global east/south," and I propose multilateral imperialism as an alternative to scholarship that interprets cross-cultural exchange through the lens of global governance. Finally, I demonstrate the agency of union-sponsored cooperatives in shifting workers' housing toward a real-estate market through self-acquisition of land and tenant-ownership model.