This thesis explores the use of tensor network methods—specifically the imaginary Time-evolving Block Decimation (iTeBD) algorithm—to evolve one-dimensional finite transverse field Ising model (TFIM) spin chains and compute their ground state energies. Using the Matrix Product State (MPS) formalism, a compact representation of the system state is implemented, so that operations on the state can be performed efficiently. After simulating the evolution for different chain lengths, the computed ground state energies are compared with the known analytical solutions. The method demonstrates high accuracy and linear error scaling with system size, outperforming brute-force diagonalization.

Hydropower is a renewable, low-carbon and dispatchable energy source, which makes it a crucial component in combating climate change. This motivates and necessitates research on operating hydropower plants optimally. In this thesis, several different models of the energy production of the hydropower plants in the Skellefte River over a seven-day period are developed and compared. This study is partly based on a previous report which compares a linear programming model (LP) of connected power plants with a more accurate mixed-integer linear programming model (MILP) that, by utilizing integer variables, incorporates forbidden discharge intervals. What distinguishes this study is that restrictions on how frequently the power plants are started are incorporated into the model. This is motivated by the fact that excessive starting and stopping is damaging to a hydropower plant. The models are formulated and solved in GAMS.

Thus, the aim of this study is to compare the less detailed and computationally demanding LP-model with the MILP-model that incorporates restrictions on the number of start-ups and evaluate the agreement of the optimal profits and the operational schedules. This is done, partly, in an effort to gauge the LP-model’s suitability as an approximation of a system of connected hydropower plants. 

The results suggest that applying a restriction on the number of start-ups had a marginal effect on the profit and, although the impact on the operational schedule was significant, the similarities in the overall pattern of the discharge plans support the conclusion that the linear model (LP) is an adequate approximation of the model with restrictions on the number of start-ups.

This thesis primarily discusses the linear quantum kicked rotor and its energy growth over long periods of time. The linear kicked rotor is a useful model due to being exactly solvable. It has the Hamiltonian $H=2\pi \alpha p + V(\theta)\sum^\infty_{-\infty} \delta(t-n),\space (0\leq\theta\leq2\pi)$ as written by Berry [Physica D 10 (1984) 369–378]. The quasi-energy eigenstates of the system are localized in $p$ if $\alpha$ is irrational and extended in $p$ if $\alpha$ is rational. The re-derivations of the quasi-energy eigenstates, wave-function and kinetic energy expectation aim at reproducing previous results. The result in Berry's paper is for the pure momentum initial state, where the energy grows quadratically in time when $\alpha$ is rational and remains bounded when $\alpha$ is irrational. A new derivation generalizes the kinetic energy expectation and momentum expectation for arbitrary initial states. The generalized result reduces to Berry's in the case of a pure momentum initial state. Without a pure momentum initial state, the energy growth behaviour over long times still remains the same as Berry's.

This thesis focuses on the derivation and numerical solution of the shallow water equations (SWE), which is a system of partial differential equations used to model fluid flow in situations where the horizontal length scales are much larger than the vertical depth. Since the SWE generally cannot be solved analytically, a high-order numerical method is introduced. Specifically, the Summation-By-Parts (SBP) framework combined with the Simultaneous Approximation Term (SAT) technique is presented and implemented to solve the equations with accuracy and stability. The scheme is implemented in Python. Different test cases were used to analyze the accuracy and stability of the implemented scheme. It was concluded that the scheme worked well for basic wave behaviors, while it struggled with more challenging scenarios involving complex water flow and shocks.

This report presents a master’s thesis in Integrated Product Design at the Royal Institute of Technology (KTH), conducted during the spring of 2025. The project was commissioned by the Swedish National Resource Centre for Deafblindness (Nkcdb). The purpose of the project was to develop a deeper understanding of the everyday challenges faced by individuals with congenital deafblindness and to use these insights to contribute to improvements in their daily lives. The project began with a background study in which relevant literature was reviewed. This was followed by a qualitative investigation, consisting of interviews with professionals working in the field of deafblindness, as well as with parents of children with congenital deafblindness. In addition, two contextual observations were conducted to further deepen the understanding of the target group. The results from the interviews were analysed and categorised into key themes. Insights from the background research and interviews formed the foundation for idea generation and subsequent concept development. The concept development process was iterative and included continuous dialogue with experts in the field of deafblindness to ensure that the proposed solution was well suited to the needs of the target group. The project resulted in Time Buddy, a conceptual time aid designed to support individuals with congenital deafblindness by improving their sense of time. The device aims to reduce feelings of stress and anxiety, especially during unstructured periods or sudden activity transitions, by fostering a greater sense of predictability in daily life.

Artificial Intelligence (AI) is no longer a vision of the future, it is a revolutionary technological force with the potential to reshape how humans live. Driven by exponential data growth, advanced algorithms, and powerful hardware, we are in the middle of an AI boom that is fundamentally transforming how companies organize their processes, make decisions, and create value.This study examines how AI is being integrated into CAD (Computer-Aided Design) tools to streamline design processes in technology-intensive industries, with a particular focus on the defense sector. As product development faces increasing demands for innovation, shorter lead times, and greater sustainability, AI emerges as a key component in meeting these challenges. Through interviews, observations, and literature reviews, insights have been gathered on how AI is practically impacting workflows, decision-making, and overall efficiency.The aim of the study is to map relevant AI tools, analyze their current and future applications, and identify how they can contribute to a more modern and competitive development process.The findings show that AI holds significant potential to redefine the role of the designer by analyzing large datasets, generating innovative design proposals, and optimizing solutions that previously consumed a disproportionate amount of engineering time. Features such as generative design, predictive commands, and AI assistants are highlighted as key enablers in future CAD environments.At the same time, the study makes clear that the full impact of AI cannot be achieved through technical investments alone. A strategic, holistic approach is required, one that combines technological development with organizational maturity. Success depends on broadening skill sets, adapting digital platforms, and ensuring the security and transparency of AI systems.The analysis presents three possible development pathways, Minimal, Intermediate, and Maximum, illustrating how different levels of AI integration influence a company’s innovation capacity and long-term competitiveness. The study shows that even limited implementation can lead to some efficiency gains, but it is the more ambitious strategies that pave the way for real success. AI in CAD, therefore, emerges not merely as a technical enhancement, but as a transformative force reshaping workflows, skill requirements, and business strategies.The recommendations presented in the study are intended to support companies in implementing AI in a strategic and goal-oriented manner, with a focus on strengthening innovation and streamlining product development. Successful integration requires not only the right technology, but also the right conditions and a strong willingness to change.

This thesis explores the interplay between combinatorics, geometry, and algebraic systems through the lens of integer lattice structures and polynomial root counting. Beginning with the classical Frobenius problem, we investigate the space of linear Diophantine equations and demonstrate how solutions can be characterized geometrically using lattice polytopes. In the two-variable case, we derive the closed-form expression for the Frobenius number and build a foundation in the geometry of lattice polygons, later introducing Pick's formula to calculate the Euclidean area of any lattice polygon given its lattice points.

Extending to higher dimensions, we introduce Ehrhart theory as a framework for understanding the enumeration of lattice points in dilations of rational polytopes. We develop the Ehrhart series and its connections to generating functions, setting the stage for a deeper exploration into algebraic geometry. The focus then shifts to systems of polynomial equations and the Newton polytopes associated with them. We present Bernstein’s Theorem and its geometric proof via mixed volumes, revealing a profound connection between the combinatorics of Newton polytopes and the algebraic structure of polynomial systems in the algebraic torus.

Finally, we derive Bézout’s classical Theorem as a special case of the BKK bound and explore the emerging field of mixed Ehrhart theory, highlighting how it naturally bridges discrete geometry and algebraic complexity. Throughout, emphasis is placed on intuition, geometric interpretation, and the deep structural unification between counting, volume, and algebraic solutions.

This report focuses on one of the most important tests for the validity of the theory ofgeneral relativity, more precisely the problem of Mercury’s anomalous perihelion preces-sion. This problem, as well as a few possible solutions for it, are introduced and describedin further detail in the introduction. Four of these solutions, which in this report arecalled the Newtonian solution, the relativistic method, Brown’s method and a numericalsolution in form of a simulation, are then described further in the investigation. TheNewtonian and relativistic methods are derived after a short introduction to general rel-ativity. Subsequently, each of the four methods is used to determine the perihelion shiftof a few celestial bodies, after which the results from each method are compared to eachother and to observed values for the shift.

In this thesis, traffic is modeled as a minimum cost network flow problem, which routes traffic in an optimal way. Different vehicle types with unique characteristics are modeled by expanding the idea of multiple commodities. By transforming the network flow problem into a linear programming problem, the model is solved using the simplex method. The model is applied to a traffic scenario in central Stockholm, which show that the model can replicate interactions between different vehicle types. Limitations and strengths of the model are discussed by looking at two potential applications.

We investigate a novel thermal light-based LiDAR system that offers high accuracy and robust noise resistance while relying on inexpensive light sources. Furthermore, this approach offers a theoretically unlimited ambiguity range due to the inherent chaotic nature of thermal light sources. In this work, we demonstrate that temporal photon correlations in sunlight, as a thermal light source, enable distance measurements in a thermal LiDAR system. Bunched photons are spectrally filtered around 1550 nm with a notch filter in combination with a Fabry-Pérot interferometer to yield a narrow bandwidth of 0.03 nm FWHM. Superconducting nanowire single-photon detectors (SNSPDs) are employed, enabling high-precision range sensing. Peak second-order correlation values of approximately 1.8 were achieved in free-space propagation and across various single mode optical fibers, up to 13.5 km. Measurement precision varied from centimeter to sub-millimeter scales. Additionally, we introduce a signal quality metric to increase reliability in assessing second-order time-correlation measurements compared to the peak second-order correlation value alone.

Byggbranschen befinner sig i ett förändringstillstånd där ökade krav på effektivitet, hållbarhet och korta projekteringstider driver behovet av nya digitala arbetssätt. Detta är inget undantag för projektering inom industribyggande där den beställande kunden ofta är under tidspress för att etablera eller expandera sin produktionsverksamhet. En vanlig konstruktion inom industribyggande är stålhallar då de tillhandahåller stora invändiga ytor och flexibla lösningar som industriverksamheter ofta efterfrågar. Trots att kunskapen om stålhallar är väl etablerad, bygger dimensioneringen fortfarande ofta på manuella och erfarenhetsbaserade metoder, vilket gör det svårt att säkerställa optimala lösningar sett till ekonomi och miljö. Till följd av att stålhallen normalt har en repetitiv utformning lämpar den sig väl för digital modellering. I detta sammanhang framträder parametrisk design och optimeringsalgoritmer som kraftfulla verktyg.

Genom att skapa en parametrisk beräkningsmodell för en stålhall, med en integrerad optimeringsprocess, har detta arbete utforskat hur generativ design kan användas för att skapa förutsättningar för snabbare, mer informerade beslut i tidiga projekteringsfaser. Med hjälp av den generativa designmiljön som utvecklats har även optimala utformningar av olika stålhallar utvärderats, med målet att minimera stomstålets massa. I samband med detta utfördes även en känslighetsanalys för att utvärdera hur stålmassan beror av de geometriska parametrarna pelaravstånd, fackverkshöjd samt fackindelning. Resultaten från arbetet påvisar att den utvecklade modellen är väl lämpad till att skapa och utvärdera olika designalternativ i tidiga projekteringsskeden. Känslighetsanalysen visade att pelaravståndet hade störst påverkan på stålmassan, jämfört med övriga utvärderade parametrar.

This report gives an account of the relativistic corrections necessary for a satellite navigation system. In particular, it concentrates on the corrections arising due to general relativity. The main effect is the gravitational blueshift of the clocks’ frequency on board the satellite, as observed by Earth. Furthermore, the time dilation is derived from two metrics, the weak- field approximation and the Schwarzschild solution. The resulting expressions are solved and compared numerically in order to assess the robustness of the weak-field approximation. A theoretical model is proposed to evaluate the time dilation for different celestial bodies. Using Earth-parameters it amounted to approximately +39 μs/day. This result is in agreement with previous studies. For other rocky planets the limit of the weak-field approximation is found at masses of Jupiter’s magnitude – no such rocky planets are currently known. Hence, there is no evidence that the time dilation derived from the Schwarzschild solution is necessary for satellite navigation systems.

Particle dynamics in turbulent fluid flows, particularly under gravity-capillary wave regimes, plays a pivotal role in environmental and engineering challenges such as marine pollutant dispersion and plastic aggregation. However, the interplay between turbulent wave forcing, particle interactions, and clustering remains underexplored in controlled experimental settings. This study combines laboratory experiments and stochastic modeling to investigate the dispersion and clustering of floating particles in gravity-capillary wave turbulence. A transparent wave tank equipped with programmable paddles and cylindrical obstacles generated controlled turbulence, while a water and titanium-dioxide mixture enhanced optical contrast for surface measurements. Two particle sizes (8 mm and 15 mm) were tracked using Fourier Transform Profilometry (FTP) for wave-height reconstruction and Particle Tracking Velocimetry (PTV) for trajectory analysis. A stochastic model incorporating turbulence-driven velocity fluctuations, wave-induced energy transfer, and particle trapping was developed to simulate dynamics.

Caveolin is an integral monotopic membrane protein associated with curved regions of cell membranes. In mammalian cells caveolins oligomerize into large complexeses which together with another group of proteins called cavins form small invaginations of the membrane, so called caveolae. However, experimental studies using model membranes have demonstrated the ability of the human Caveolin-1 complex (7sc0) to induce curved regions of the membrane without cavins. This membrane shaping ability is lacking in caveolin from the species S. purpuratus (9dn0) and S. rosetta (9dn1). Motivated by these experimental results, we used computational methods to examine the membrane shaping ability of the human Caveolin-1 complex by comparing lipid-protein interactions between these three species. The dynamics of the caveolin complexes embedded in simple bilayers were studied using coarse-grained molecular dynamics simulations. Similar computational studies have indicated a correlation between curved regions and the splaying of the lipid chains in the distal leaflet right across from the complex. The results from our simulations agree with the experimental observations in that 7sc0 does associate with regions of high curvature while 9dn0 and 9dn1 do not, but we did not see the expected dependence on lipid splay in the distal leaflet. Varying the lipid number imbalance between the two leaflets did not provide further mechanistic insights. However, analysis of the lipid-protein interactions in the proximal protein-containing leaflet indicated differences between the three caveolin variants. Detailed hydrophobicity analysis of the structure of each complex showed that both 9dn0 and 9dn1 have a clear boundary between hydrophobic and hydrophilic regions preventing lipids from sliding down the rim of the disk-shaped complex. This boundary is not present for 7sc0, allowing the proximal lipids to instead change their relative positioning and more easily accommodate bilayer curvature. These findings help further our understanding of the membrane shaping ability of the human Caveolin-1 complex.

This thesis presents the design and construction of a light-sheet fluorescencemicroscope intended for educational use. Inspired by the open-sourcedescSPIM project, the microscope was built to allow students to gain practicalexperience with optical systems and microscopy principles. The systemenables three types of imaging experiments: visualization of the light-sheetitself for resolution assessment, imaging of small samples in cuvettes, andimaging of larger cleared samples in an index-matched chamber. Keychallenges included developing functional control software and compensatingfor refractive index mismatches during imaging. The microscope wassuccessfully used to produce 3D reconstructions of biological samples,demonstrating both the strengths and limitations of the system. The setupemphasizes practical learning, offering students valuable insights into samplepreparation, light-sheet alignment, and volumetric image acquisition.

This thesis work investigates the possibility of 3-D printing an ionization chamber for use in the field of microdosimetry. The aim is to achieve a proof of concept with a detector printed with a combination of conductive and non-conductive materials. The design can then be coupled with existing electronics to provide radiation measurements. In addition to printing a working detector, the work aims to suggest potential design changes in geometry and dimensioning. A spherical detector was designed and printed using PLA and Conductive PLA filaments. The diameter of the active region is approximately 5 cm, while the entire detector with stem is 11 cm tall, with a wall thickness is 3 mm. The printed detector was equipped with a BNC cable and banana connectors to be coupled with electronics. Measurements were performed with the detector at the Swedish Radiation Safety Authority, with the goal of using the variance method for determining dose mean lineal energy at the micrometer scale. Under irradiation from a cesium-137 source, the variance method gave a dose mean lineal energy of 0.685 ± 0.055 keV/μm for cesium-137 in a simulated site diameter of 53 μm. This is in line with literature values obtained with commercial detectors. The dose rate was also calculated to validate the performance of the detector. This study will give a basis for the use of such 3-D printed models in the field of micro- and nanodosimetry.

Recent efforts have been made to extend the well-understood classification of symmetry protected topological (SPT) phases in pure states to that of mixed states. This raises the need for new tools that can be used to define SPT in mixed states. In this thesis we study the open boundary ground states of the Affleck–Kennedy–Lieb–Tasaki (AKLT) model, belonging to the Haldane phase, using the information lattice. We characterize the states on the information lattice and link the distribution of information in the information lattice to the double degeneracy in the entanglement spectrum, a previously known indicator of the Haldane phase. The ground states are then subjected to noise with different symmetries and we study the evolution of the information lattice. We find that strongly symmetric noise preserves 2 bits of information at the top of the information lattice whilst weakly symmetric noise does not. These results indicate that strong symmetry can be used to define a mixed state SPT equivalence relation.

I urbaniseringen har stadskärnorna blivit alltmer förtätade vilket ökar invånarnas exponering av luftföroreningar. Denna studie undersöker därav hur bilfria områden i städer kan bidra till en mer hållbar stadsplanering utifrån alla tre aspekterna.

Under 2025 införs en ny stadsmiljözon på Södermalm i Stockholm med syftet att skifta området mellan husen från trafiken till ökad mänsklig aktivitet. Zonen regleras enligt bestämmelser för gågator med de undantag som uttrycks i trafikförordningen. Platsinventeringarna visade att Nytorget redan är en livlig mötesplats medan områden i närheten som Greta Garbos torg och Axel Landquist park hade mycket lägre aktivitet. Därav är en viktig utmaning med zonen att se till att dessa områden får en ökadrörelse så de inte upplevs som otrygga i jämförelse med närliggande platser. 

Slutsatsen är att bilfria zoner har potential att bidra till en hållbar stadsutveckling ur många aspekter, både i och med minskade luftföroreningar vilket kan bidra till en ökad hälsa bland invånarna och ett mindre klimatavtryck för städer i stort. Mindre trafik kan även ge ökade mänskliga interaktioner och aktiviteter på platser i staden vilket har ett socialt värde.

In modern natural language processing tasks, the transformer architecture has set significant benchmarks across common language modeling metrics. Transformer-based language models typically process linguistic information implicitly through self-attention mechanisms, offering little insight into linguistic priors such as syntactic structure. This thesis investigates the enhancement of transformer performance via explicit incorporation of Part of-Speech (POS), a form of syntactic knowledge. Specifically, we examine different methods of fusing tokens with POS in the embedding layer to provide a computationally inexpensive injection method. We amend a transformer-decoder with dense POS embeddings and compare it against a baseline model on the WikiText-2 dataset in terms of perplexity and next-token prediction. Our experiments indicate that integrating POS information through concatenation and projection leads to statistically significant improvement, decreasing perplexity by approximately 15.1% (from 42.70 to 36.26) and increasing next-token prediction accuracy by 0.69 percentage points (from 55.18% to 55.86%) on the WikiText-2 test set. We find the concatenation and projection method particularly effective, outperforming simple element-wise addition. We discuss the findings, linking the performance gains to the role of POS tags as an inductive bias within an information-theoretic framework. Furthermore, we provide methodological details for aligning tags with subword tokens and acknowledge limitations concerning model scale and tagger accuracy. This research provides evidence and implementation insights for incorporating explicit syntactic knowledge through lightweight embedding fusion into decoder-only transformer architectures.

Denna studie har undersökt hur snabbare pendelbåtar kan påverka antalet passagerare, med fokus på linje 89 som går mellan Tappström på Ekerö och Klara Mälarstrand iStockholm. Studien utgår från båtmodellen Candela P12 som kan och får köra snabbare än dagens båtar och som ska testas i trafik på sträckan. Med hjälp avtransportplaneringsprogrammet PTV Visum simulerades tre scenarier med olika restider: 55 minuter (vilket är nuvarande restid), 35 minuter och 25 minuter. Resultaten visar på en ökning av antalet passagerare med ungefär 80-140% jämfört med grundscenariot. Den största ökningen kom från hållplats Tappström, vilket tyder på att restiden är en avgörande faktor för resenärers val av färdmedel. Studien visar att snabbare pendelbåtar har potential att avsevärt öka antalet passagerare och stärka vattenvägarnas roll i Stockholms kollektivtrafik.

Studien tar inte hänsyn till nygenerering av resor eller passagerare som byter från bil till pendelbåt. Fordonens passagerarkapacitet betraktades inte heller. Trots dessabegränsningar ger studien insikter om potentialen för snabbare pendelbåtar att öka kollektivtrafikens attraktivitet och effektivitet.

As the development of Generation IV lead-cooled fast reactors (LFRs) progresses,ensuring the mechanical integrity of structural materials in harsh core environmentsis essential, especially under accident scenarios involving fission product release.One such concern is Liquid Metal Embrittlement (LME), a phenomenon wherea liquid metal causes a transition from ductile to brittle behavior in structuralalloys. This thesis examines the susceptibility of the ferritic alumina-forming alloyFe10Cr4Al (10-4) to LME when exposed to molten lead contaminated with tellurium(Te), a fission product of particular relevance. Slow Strain Rate Testing (SSRT) isconducted on 10-4 specimens across a range of temperatures, strain rates, and Teconcentrations, using both unnotched and notched samples. For reference, 316Lstainless steel is tested under similar conditions. Topographic and chemical analysis(SEM/EDS) on 10-4 specimens show that LME is most evident in notched samples,where stress concentration damages the oxide layer that forms on the steel surface,promoting liquid-solid contact. The strongest embrittlement occurs at intermediateTe concentrations ( 0.3 wt%), where brittle fracture features dominate. Ductility fullyrecovers at the highest Te content (1.2 wt%), due to the supersaturated solution thatleads to the formation of Te-rich precipitates that limit wetting or reduce effectivephase interaction. Moreover, LME effects are not observed in unnotched specimens atlower strain rates, suggesting that slower deformation helps preserve oxide integrityand enables the formation of protective precipitates that hinder penetration.

The Bidirectional Scattering Distribution Function (BSDF) provides a comprehensive description of how light interacts with a material surface, expressing the fraction of optical power redistributed as a function of incident and scattered angles, detection area, and detection distance. Accurate BSDF data are critical in applications for material characterization, quality control, optical simulations, and photorealistic rendering.

This thesis presents the design and construction of a prototype scatterometer tailored for precise BSDF measurements. The work concentrates on the optical subsystems, light source and detector and was conducted in close collaboration with mechanical and software teams, as well as with Andréa Nordmark (Uppsala University) on detector and light-source development. The finalized light source design delivers a collimated beam with a divergence of $\approx 0.14^\circ$ and a diameter of 7.36 mm, while limiting angular deviation to $0.5^\circ$. It accommodates incident angles up to $70^\circ$ and detection angles up to $82^\circ$, powered by a minimum incident flux of 21.6 mW to ensure robust signal-to-noise performance.

Prototype validation included linearity tests of detector response versus exposure time and incident power, along with angular scattering scans on representative greenish and reddish samples. The results confirm the system’s ability to capture BRDF signatures with high fidelity, while revealing practical challenges—such as LED spectral aberrations and detector saturation—that will inform future refinements.

The study demonstrates that detailed BSDF measurements can be achieved with cost-effective components when core optical requirements—collimation quality, angular precision, and detector linearity—are rigorously met. The platform establishes a solid foundation for extending capabilities to Bidirectional Transmittance Distribution Function (BTDF) measurements and for improving the way measurement data is processed and analyzed in future efforts.

Cardiovascular disease is the leading cause of death worldwide, hence, computer-aided research has been increasingly more popular in the field. Medical AI could be used to find early signs of disease, diagnose, and simulate disease effects in scans. With a medical image generation model, one could potentially use it for educational purposes during medical training, but also to understand how diseases affect the body and even potential causes. Image generation models are generally trained on natural images, but this project aimed to fine-tune a stable diffuser on medical chest x-rays. Counterfactuals is an image-to-image generation aiming to transition the image from one label to another. In this case, counterfactuals were generated with cardiomegaly present or not, depending on the original label. Datasets were generated from a chest X-ray dataset to represent cardiomegaly and non-cardiomegaly patients in equal numbers. Many diffusion models have the option of segmentation masks, which could minimize the change in areas irrelevant to the disease. Segmentation methods were tested, and the approach led to HybridGNet, which can generate both lung and heart segmentation masks and is therefore multipurpose depending on the application. Multiple pretrained classification models were tested and fine-tuned for a binary cardiomegaly classification. The search for the best approach for the type of task did not show a clear winner since the different ResNet, DenseNet, and EfficientNet models performed approximately equally, around 83-85\% test accuracy. The best test accuracy was for EfficientNet V2, and it was therefore used to evaluate the generated images. The computational demands and training times did, however, differ, and for practical applications, ResNet18 and EfficientNetB0 will suffice because of their low computational demands, low training time, and comparable test accuracy to the other models. For the generative models, the Stable Diffusion v1-4 was used to fine-tune. The small dataset and lower computational power did not lead to great results with an SSIM of 0.12 and a relative SSIM of around 0.29. The counterfactual images were generated with varying strengths and tested to see the class switching abilities using the classification model. The results never achieved the class switching abilities while still maintaining similarities between real and synthetic images. The testing was, however, done on relatively small datasets, and the batch size had to be kept lower to not computationally exhausting the model. Although it did not fulfill the goal of being able to perform class switching, it shows promise for further testing. More data is needed than just the dataset, as it does not contain many patients with cardiomegaly. Combined with better computing power, this could potentially lead to improved results. This could in the future be used for teaching purposes. With further development, it could include features such as patient gender, age, or any other underlying diseases.

The digitalization of healthcare has led to an increased use of connected medicaldevices, which in turn has introduced new cybersecurity risks. This study examinesthe security risks and vulnerabilities of pacemakers and ventilators, with a particularfocus on their technical functionality, communication systems, and potential threatsfrom cyberattacks. Through a combination of literature review and semi-structuredexpert interviews, several weaknesses have been identified, including insufficientauthentication, vulnerable communication protocols, and the use of outdatedsoftware. To reduce these risks, the importance of security in the design phase ishighlighted, along with regular updates, network segmentation, and enhancedcollaboration between cybersecurity experts and medical device professionals. Thestudy demonstrates that a systematic approach to cybersecurity is essential toprotect both device functionality and patient safety. 

Since sustainability has long been a challenge in the property sector, and energy efficiency isbecoming increasingly important, the potential of energy efficiency in two commercialbuildings is explored. Additionally, it will examine the implementation of energy efficiencyfrom a tenant's perspective and their role in achieving sustainable outcomes. The thesis alsoconsiders the influence of the EU’s Energy Performance of Buildings Directive (EPBD) andrelated legislation, which place growing demands on property owners and tenants alike tomeet higher energy performance standards.

Therefore, the purpose of this thesis is to evaluate whether an older commercial building canachieve energy performance levels comparable to a newer constructed building, consideringcurrent and evolving regulations. It further aims to investigate tenants' engagement in energyrelated efforts, with a focus on how their behavior, motivation, and interaction with propertyowners impact the practical implementation of energy efficiency measures. To fulfill thispurpose, a mixed-method approach is adopted, which combines qualitative semi-structuredinterviews with tenants and a quantitative comparative analysis of key energy performanceindicators of two properties.

The results show that with targeted retrofitting measures, the older building can achieveenergy performance levels comparable to the new construction, achieving up to a 26%reduction in total energy use. Moreover, tenants' willingness to pay extra for energy efficientbuildings is often financially limited and is therefore more driven by operational savings orcomfort improvements rather than environmental considerations. Another key challengeidentified is the lack of accessible energy information and the misalignment of incentivesbetween property owners and tenants. The thesis concludes that greater communication,transparency, access to KPIs, and collaborative investment models are essential to bridge thegap between regulatory ambitions, property owners, and tenants. 

This study compares the management and distribution of medical devices in the home between the Stockholm Region and the Västmanland Region, focusing on equipment for sleep apnea, life-support ventilation, nutrition and oxygen. The aim is to identify regional differences in the division of responsibilities, preparedness and efficiency to improve patient safety and resource use. Through interviews and document analysis, the results show that Stockholm's centralized model provides predictability but high costs, while Västmanland’s flexible model is cost-effective but vulnerable to supply disruptions. Both regions achieve high patient safety, but disposal of materials and lack of remote monitoring are challenges. The study recommends digital solutions, standardized protocols and sustainable materials to optimize management. The results are relevant for regions that want to develop effective and sustainable systems for medical devices. 

Industrial real estate has become an increasingly important part of the property market,particularly as e-commerce and global trade drive demand for warehouse and productionfacilities. Despite this, research has primarily focused on institutional investors and largescale actors, while mid-sized private property owners remain relatively underexplored. Theseowners represent a significant share of the market, and their decisions influence the sector’sdynamics and future development. The purpose of this study is to examine the investmentbehavior of mid-sized private industrial property owners in Stockholm and to analyze theirperceived investment performance and outlook. This has been done through a quantitativeanalysis of the owner distribution in Länna industrial areas in Stockholm to map out theownership dynamics in the area. A qualitative analysis has also been done through interviewswith property owners and companies managing industrial real estate with the aim to identifythe factors influencing their investment decisions, including personal goals, market dynamics,and future outlook.

This thesis explores how blockchain-enabled tokenization can facilitate sustainable real estate development by addressing financing barriers in green building retrofits. Integrating extended utility theory with quasi-empirical calibration, the study evaluates how key blockchain features—liquidity, transparency, and environmental return—impact investor decision-making and capital allocation.

A theoretical model is developed that extends the classical mean-variance utility framework to include liquidity premiums, smart contract transparency, and monetized environmental benefits (e.g., carbon credits). Sensitivity analyses reveal that both liquidity and transparency have independent and synergistic effects on investor utility, suggesting tokenization reduces perceived risk while improving market accessibility.

To enhance external validity, the model is calibrated using real-world data from the Grev55 tokenized residential project in Denmark. Results show that tokenized green assets can deliver competitive returns and substantial utility gains, particularly when transparent reporting and secondary trading are enabled.

This study contributes to the emerging literature on green digital finance by demonstrating how blockchain can improve the economic viability of sustainable investments. The findings offer practical implications for policymakers, developers, and ESG-focused investors aiming to scale up climate-aligned real estate finance through digital innovation.

This project aims to provide performance data for the motor + propeller used in the C-3 (Combat Climate Change) collaboration between KTH and EAFIT using wind tunnel testing. The project involves computer assisted design of components, electrical integration, experiment planning, and data analysis. A testing rig consisting of a motor-balance adapter mounted to a sting was set up in the KTH L2000 wind tunnel, where the motor + propeller were tested at 0° angle of attack with 0, 15, 20, and 25 m/s freestream velocity at different motor power levels. The 6-axis balance telemetry data was converted into measurements thrust and torque. These were further used to generate coefficients of power and thrust. The results show that propeller efficiency increases with higher power settings, peaking around 15 to 20 m/s freestream velocity. Additionally, thrust significantly exceeds drag up to 25 m/s, but the UAV may face difficulty accelerating far beyond this speed. Further testing using propellers mounted downwind of a fuselage, at speeds exceeding 25 m/s, and at different angles of attack, is needed

Digitalization has rapidly increased the availability of property data, yet many property owners struggle to convert this data into actionable insights. This study explores how structured property data can create value and support a shift toward proactive and data-driven property management. Based on two case studies and interviews with industry professionals, a six-step framework was developed to support the transition from unstructured property data to data-driven decision-making. The framework was tested and refined with three large property owners, demonstrating the need to combine technical solutions with organizational change.

The thesis concludes that structured property data not only enhances operational efficiency but also enables better procurement, long-term planning and more sustainable and well-informed decision-making. The findings emphasize that leadership engagement, user-friendly systems and a clear responsibility structure are essential for achieving long-term value. Overall, the study shows that a clear framework can turn data structuring into a practical tool for supporting informed and long-term decision-making in property management.

Denna rapport undersöker stabilitetskriterierna för Dead Ship Condition i Second Generation Intact Stability Criteria (SGISC) som är framtagna av International Maritime Organisation (IMO). Dessa kriterier består bland andra av nivå 1 och nivå 2 som utvärderar fartygs stabilitet utifrån fartygs geometri, lastfall och påverkan av externa krafter som modelleras på olika sätt för de två nivåerna. De olika nivåerna modellerades i Matlab för en redan given skrovform och modelleringen utgick ifrån ett kodpaket med flertalet färdiga funktioner. Konstanter och funktioner angavs i enlighet med IMO:s kriterier. Nivå 1, som i tidigare regelverk kallas Vindkriteriet, bygger på en konstant vindkraft som simulerar en kraftig byvind och nivå 2 modelleras med både en oscillerande vindkraft och en vågkraft som påverkar fartyget. Utifrån denna rapport visas att skrovet som användes är stabilt för nivå 1. Fartygsskrovet stabilitet enligt nivå 2 kunde inte på korrekt sätt tas fram på grund av felaktigheter i koden för nivå 2. Rapporten visar att det är svårt att avgöra stringensen mellan nivåerna då nivå 2 är betydligt mer komplext och därför svårare att implementera.

This thesis develops a machine learning-based forecasting model to predict monthly housing price growth in Shanghai, with the aim of capturing short-term market dynamics through a combination of macroeconomic, financial, policy, and sentiment-based indicators. The model is built using the XGBoost algorithm and is interpreted using SHAP values to ensure transparency in feature attribution.

By reformulating the prediction target as monthly price growth and aligning it with percentage-based input features, the study improves model coherence and performance. The results show that recent price momentum plays a key role in shaping expectations, while liquidity conditions—particularly money supply (M2)—emerge as strong macro-financial predictors. In contrast, other variables such as CPI and policy scores show more limited and variable impact, while consumer sentiment proxies demonstrate moderate influence, with their contribution varying depending on broader market conditions. The inclusion of financial market indicators further confirms the influence of investor sentiment and capital reallocation in short-term housing price movements.

Beyond its technical contribution, the study offers a structured and explainable framework for urban housing market forecasting. It demonstrates the value of integrating data-driven models with economic reasoning and provides insights relevant to policymakers, analysts, and stakeholders seeking to understand and anticipate price changes in complex real estate environments.

Color transfer is an image processing method in which the colors of an input image are mapped to the colors of a target image. The map can be obtained by describing the colors of the images as distributions and formulate the prob- lem into an optimal transport problem using Kantorovich’s formulation. The methods which this report will utilize to solve this transportation problem are the simplex method and Sinkhorn’s algorithm. The performance of the two methods can be improved by reducing the number of colors in the images using a k-means algorithm. The different methods will be tested on several pairs of images, demonstrating that Sinkhorn’s algorithm combined with k-means is the most effective method in terms of computational and qualitative results.

This thesis investigates the stability of polarization states in a 20 km deployed underground fiber network between KTH Royal Institute of Technology and Ericsson in Stockholm, intended for the study of quantum communication. Polarization-encoded quantum states are sensitive to environmental disturbances, making active stabilization essential for maintaining fidelity. The study evaluates the performance of Qunnect’s Automated Polarization Controller (Qu-APC) through three experiments: measuring passive long-term polarization drift, analyzing wavelength-dependent polarization effects, and implementing triggered compensation using an Arduino-based control scheme. Results show that polarization drift occurs consistently over time, with a baseline drift of approximately 1.7% in fidelity per hour, emphasizing the need for active compensation. Triggered compensation every 30 minutes maintained fidelity above 95% and achieved over 99% network up-time. Although the wavelength sweep experiment produced inconclusive results, the findings confirm the feasibility of maintaining stable polarization states in real-world fiber networks, providing a foundation for future quantum communication applications.

Detta kandidatexamensarbete syftar till att ta fram en förstudie till konstruktionen av ett styrsystem för en motorcykelsimulator. Bakgrunden är behovet av en simulator som kan användas både för att träna förare i olycksdrabbade trafiksituationer utan att utsätta dem för faktisk fara och för att utföra experiment i forskningssyfte. Simulatorer har tidigare främst utvecklats för bil- och flygtrafik, men en realistisk motorcykelsimulator kräver särskilt fokus på balans, lutning och framför allt korrekt återgivning av styrkänsla. En viktig del i arbetet är att kunna återge de krafter som uppstår i verklig körning, och att differentiera mellan låg- och högfartskörning, där olika parametrar dominerar motorcykelns beteende.

Arbetet omfattar en litteraturstudie kring befintliga lösningar, olycksstatistik och tekniska krav. Moment- och vinkelmätningar har analyserats, både från tidigare studier och genom egna empiriska mätningar på motorcyklar som Yamaha MT-07 och BMW R 1300 GS. Utifrån detta har krav specificerats för systemets kapacitet, såsom att klara ett vridmoment på minst 50 Nm, styrvinkel om minst 30° åt vardera håll, samt en uppdateringsfrekvens på minst 500 Hz. 

Tre huvudprinciper för motstånd i styret har jämförts: elektriskt, gyroskopiskt och hydrauliskt. Elektriskt motstånd, särskilt med servomotor och utväxling, bedöms ge bäst prestanda i relation till kravspecifikationen, med fördelar i styrbarhet, modularitet och respons. Ett konstruktionsförslag presenteras som grund för framtida implementering i simulatorn. Projektet har även en tydlig koppling till hållbarhetsmål, då ökad säkerhet och minskade olyckor bidrar till både minskade personskador och resursslöseri.

Several gene-editing tools, such as CRISPR/Cas9, offer promising potential for cancer treatment, yet efficently delivering them specifically to tumor cells while sparing healthy tissues remains a significant challenge. This thesis introduces a novel targeted delivery platform that combines lipid nanoparticles (LNPs), nanoscale lipid nanobubbles (NBs), focused ultrasound (FUS), and bioorthogonal trans-cyclooctene (TCO)/methyl-tetrazine (mTz) "click" chemistry to overcome current delivery barriers. In initially experiments, studies with naked LNPs loaded with either a green fluorescent protein (GFP) reporter plasmid or eGFP encoding mRNA showed that the uptake of the nucleic acids in different recipient cells were dependent on both the amount of nucleic acid used during production of the LNPs, the number of LNPs incubated with the cells, and the cell line. Next, different recipient cells were treated with DOPE-PEG2k-TCO, to study the interaction with a fluorescently labelled (Cy5) version of the mTz click chemistry counterpart. Here, a DOPE- PEG2k-TCO concentration of 10 μM was found to be optimal for HEK293T cells and Yummi1.7 cells, whereas a higher concentration of 20 μM was required for CT-2A cells. In a final set of experiments, the uptake of flurorescently labelled (Cy5) and mTz-decorated LNPs into TCO-decorated Yummi1.7 cells was studied. This study showed an increase (approximately 1.78-fold) in targeted cellular uptake compared to non-targeted methods. These results highlight the effectiveness of integrating FUS, NBs, and click chemistry for precise and localized gene-editing therapy delivery. Although time constraints prevented the evaluation of CRISPR/Cas9 and guide RNA targeting miR-21, this method is broadly applicable, and enables future in vivo studies in ultrasound-guided, click-chemistry-assisted gene and drug delivery. 

This thesis investigates the relationship between Environmental, Social, and Governance (ESG) performance and stock market outcomes in the Pan-Scandinavian real estate sector during the period 2019 to 2023. Using panel data regression analysis on a sample of 23 listed real estate firms across Sweden, Norway, and Germany, the study evaluates the impact of both aggregated ESG scores and individual ESG pillars on stock returns and return volatility.

The empirical results reveal a statistically significant negative relationship between overall ESG scores and stock returns, primarily driven by the Social pillar. No significant effects are observed for the Environmental or Governance dimensions. In terms of stock return volatility, higher ESG scores, particularly in the Environmental and Social categories, are associated with increased volatility, suggesting that ESG engagement may introduce additional uncertainty rather than reduce financial risk. The Governance pillar is the only dimension where improved scores are associated with marginally lower volatility, aligning with certain strands of previous literature.

These findings challenge the view that ESG performance enhances shareholder value and reduces investment risk. From a portfolio theory perspective, ESG investments within this sector and timeframe appear to be positioned below the efficient frontier, as they are associated with both lower returns and higher volatility. The study offers relevant insights for investors and policymakers, and highlights the importance of sector-specific analysis in the evolving discourse on ESG and financial performance.

By measuring the polarization of X-ray emission from distant stellar objects such as neutron stars and stellar-mass black holes, information about their geometry and properties can be discerned. XL-Calibur is a balloon borne polarimetric mission that utilizes Compton scattering to determine the polarization of the emitted X-rays. The incoming photons are focused into the polarimeter by an X-ray mirror which is all held in alignment by an optical bench (truss). Factors during flight such as uneven heat exposure or gravitational sag can cause a slight offset in the alignment of incoming photon beam introducing a systematic error in the reconstructed polarization. The aim of this thesis is to analyze if an artificial neural network can use the data collected of photon hits in the polarimeter to correctly predict this offset. The neural network was trained and tested using data simulated to have the same point spread function (PSF) as the X-ray mirror. The model was then applied to the data collected from the XL-Calibur flight. The model performed exceedingly well on simulated data but further studies are required to determine the validity of the network’s predictions on flight data.

The circadian rhythm is a complex biological system driven by molecular feedback loops at the cellular level. This inherent rhythmicity makes the circadian system particularly amenable to mathematical modeling using oscillators. In this study, we investigate a stochastic model of the circadian rhythm in the Suprachiasmatic Nucleus (SCN) using coupled oscillators. In particular, synchronization of these oscillators are induced by implementing a small-world network of oscillators. It can be concluded that the model reliably exhibits known properties of circadian clocks, such as individual oscillators with periods ranging from 24-28 hours, and the synchronized system showing a precise and robust 24 hour period despite stochastic noise in the model. Depending on parameters, both global synchronization and cluster synchronization was observed.

This thesis investigates the securitization process of Swedish housing cooperative loans,which is an underlying asset class that remains underutilized in the Swedish capital marketsdespite its potential. Using a qualitative method, including nine semi-structured interviewswith industry professionals such as institutional investors, originators, and legal experts, thestudy explores the opportunities, risks, and regulatory frameworks associated with theimplementation of securitization.

The empirical findings suggest that securitization could improve financial institutions’ capitalefficiency and liquidity, while also enabling a more diversified investor base. Furthermore,securitization could indirectly benefit housing cooperatives by offering more stable financingoptions and lower margins. However, challenges still exist, particularly regarding regulatoryrequirements such as Basel IV and Solvency II, as well as issues related to transparency andinternational investors’ unfamiliarity with the Swedish housing cooperative structure.

The thesis contributes to a deeper understanding of how housing cooperatives can beintegrated into the broader securitization framework and what measures are needed tofacilitate a more efficient and transparent development of this market segment in Sweden.

This thesis examines whether movements in the Swedish stock market, represented by the OMXS30 index, can predict changes in prime yields for logistics and office real estate in Stockholm, Gothenburg, and Malmö. Using quarterly data from 2009 to 2024, the study applies time-series OLS regression models with lagged OMXS30 values, along with macroeconomic variables such as CPI, 5Y SEK swap rate, and GDP. The theoretical framework includes CAPM, Commercial Real Estate Market Cycles, Capitalization Rate and Real Estate Valuation – Income Approach. Previous research reveals several drivers of real estate returns and cap rate, as well as dynamics between stock returns and real estate returns and values. However, a gap in research in a Swedish context has been identified, which this thesis aims to fill. The results show no consistent and statistically significant relationship between stock market fluctuations and prime real estate yields. Most OMXS30 coefficients were insignificant and varied in direction, indicating weak or limited predictive power. Differences across sectors and cities were observed, where the logistics sectors were revealed to have a closer link to the stock market and the Malmö results stood out. Ultimately, this thesis concludes that fluctuations in the OMXS30 index should not serve as the main predictors of changes in real estate prime yields, emphasising the importance of considering the broader economic and financial landscape, as well as local and property-specific factors.

The Swedish real estate sector, while efficient compared to global standards, remains characterized by manual, paper-heavy processes that introduce administrative delays and transaction costs. This thesis explores the potential of smart contracts, self-executing agreements based on blockchain technology, to reduce time and cost inefficiencies in residential property transactions on the Swedish real estate market. Using a mixed-methods approach that includes semi-structured interviews with industry experts, AI-generated interview results, and Monte Carlo simulations, the study quantifies differences between traditional and smart contract-based transaction systems. Results indicate that while cost savings are modest (approximately 3.8%), smart contracts offer substantial time savings (up to 80%) and significantly improve process consistency in the case of using smart contracts. Key barriers to adoption include regulatory uncertainty, institutional conservatism, and limited public understanding of blockchain technology. Nevertheless, findings suggest that with appropriate regulatory frameworks and cross-sector collaboration, smart contracts could play a complementary role in the ongoing digitalization of the Swedish real estate market.

Stockholms befolkning ökar och med en växande region ökar behovet av en attraktivkollektivtrafik. För att få fler att välja bort bilen är det viktigt att erbjuda en punktlig ochpålitlig kollektivtrafik. Problem med förseningar och inställda tåg har lett till minskatförtroende bland pendeltågsresenärerna vilket även har påverkat synen på SL som helhet.

I Sverige planeras och upphandlas kollektivtrafiken vanligtvis av de regionalakollektivtrafikmyndigheterna. Där finansieringen sker via en kombination av skatter ochbiljettintäkter. I Stockholm ansvarar Trafikförvaltningen för att den trafik som omfattas avallmän trafikplikt, detta sker genom offentlig upphandling. Genom en sådan upphandling fickMTR uppdraget att driva pendeltågen.

Efter problem med en nationell lokförarbrist och eftersatt underhåll misslyckades MTR medatt leverera pendeltågstrafik av god kvalitet. Avtalet avslutades därmed i förtid, och ett nyttoch dyrare nödavtal tecknades med SJ.

Syftet med denna studie är att undersöka om operatören påverkar punktligheten ochtillförlitligheten på Stockholms pendeltåg. Studien bygger på en statistisk analys av data frånTrafikverkets uppföljningssystem för järnväg Lupp.

Resultatet visar att punktligheten förbättrades under vissa månader och försämrades underandra. Operatörsbytet ledde till färre förseningar i september, men fler i januari. Det fannsockså skillnader i punktlighet mellan olika stationer, där vissa hade en positiv effekt efterbytet.

En tydlig förbättring kunde dock ses i antalet inställda avgångar. Dessutom minskade andelenförseningar och inställda tåg orsakade av operatören efter bytet från MTR till SJ. Därmedvisar resultatet att operatören har en påverkan på regulariteten men inte nödvändigtvis påpunktligheten.

This master’s thesis investigates how Swedish Municipalities implement proactive land acquisition strategies and their impact on the long-term urban development. Urbanization and sustainability are central issues in modern society and municipal land ownership provides opportunities to steer physical planning. The research aims to examine how active land acquisition strategies in Swedish municipalities influence long-term urban development, while analyzing practical working methods, success factors and challenges.  

The study consists partly of a document analysis of 18 municipalities and partly of an interview study involving 13 municipalities. The interviews show that 8 of the 13 municipalities have formal policy documents governing land acquisition, while the others rely on internal guidelines. Most of the interviewed municipalities allocate dedicated annual budgets for land acquisition, ranging from 10 to 350 million SEK. Municipalities primarily acquire undeveloped land within development areas designated in their comprehensive plans. The time horizon varies in the municipalities, ranging from a few years to more than 50 years. Results indicate that development processes accelerate significantly when municipalities own the land, and that land ownership enhances control over urban development. The Swedish Land Acquisition Act restricts the ability to acquire agricultural land outside comprehensive plan areas, complicating truly long-term acquisitions. Economically, strategic land acquisitions are viewed as profitable investments, enabling value capture when undeveloped land becomes zoned for development.  

The study concludes that municipal proactive land acquisition strategies strongly positively influence long-term urban development, particularly regarding development pace, design, location, and holistic planning. Municipalities with active land strategies demonstrate greater capacity to achieve long-term objectives in housing, public services and infrastructure. 

This master thesis is exploring the strategic dynamics and challenges in rent negotiations between retail tenants and real estate owners – with game theoretical concepts and models as an analytical tool. The starting point of the study is the challenging status quo that is facing both retail tenants and real estate owners – where lower revenues, digitalisation and changed customer behaviors could be seen as the new normal – forcing both parties to rethink. By applying central game theoretic concepts, such as the prisoners' dilemma, information asymmetry and Rubinstein-bargaining, the study analyses how unequal bargaining positions and inefficient sequential negotiation processes contribute to pareto-inefficient outcomes.

Methodologically, the study builds upon a qualitative case-study in a local center in a suburb of Stockholm – where two semi-structured interviews were conducted – with a retail tenant, RT, and the real estate owner, REO – jointly compiling the empirical foundation of the study. The results show how the REO, through strategic information acquisition and exit-threats, succeeded in obtaining a stronger bargaining position, while the RT, who had limited experience and was emotionally invested, had a weaker bargaining position. The sequentialstructuring of the negotiation agenda further restricted the possibility of accommodating the joint interests – ultimately leading to non-optimal outcomes for both the REO and the RT.

The study also discusses potential measures for improvement – such as increasedtransparency through legislations, a reconstruction of the rent negotiation disposition, but also the utilisation of tenant representative services – with the aim of fostering more sustainable, and symbiotic results. By combining established theoretical concepts with practicalobservations, the study contributes with a deeper understanding of retail rent negotiations and its challenges, as well as its dynamics. In summary, the study highlights the importance of rethinking the contemporary, non-collaborative approach to negotiations – in order to obtain astable and long-term development of the brick-and-mortar retail and commercial real estate.

The European real estate industry is addressing its role in sustainable development by approaching new construction from the perspective of meeting social, financial, and environmental goals, however, the ageing building stock has been left behind and is due for renewal. Against the background of the European Union’s efforts to achieve ambitious climate goals, the existing building stock constitutes a significant latent capacity to contribute to the sustainability goals, with an initial target of 35 million buildings to be renovated by 2030. The aspirational union-wide goals rest on the 2024 Energy Performance of Buildings Directive (EPBD) which mandates the Member States to achieve a net zero emission building sector by 2050, where the improvement of the existing building stock is of special focus. This thesis explores the concept of sustainable real estate development from the perspective of Green Renovations. The main objective of this research was to undertake an exploratory study concerning the national implementation in Swedish context as well as the Swedish real estate industry’s awareness and preparedness for the new directive, including the potential response to its implementation. The underlying problem under investigation concerns the financing of the renovation on a national level. This thesis approached the study from the perspective of four research questions, concerning legislation, the economy in renovations, renovictions and implementation feasibility. This research utilized thematic semi-structured interviews as the basis for a mono method qualitative study. The main findings illustrate that the industry is not specifically prepared for the Directive, legislators will attempt to provide flexibility, profitability guides the actions of property owners, energy efficiency measures are incorporated together with standard improvements and rents are raised, and thus energy targets can predominantly be achieved. This thesis concluded that the Swedish implementation ofthe Directive will result in targets similar to today’s energy class C, allowing for flexibility in accordance to cost-optimal levels. While technically feasible, real estateowners only perform profitable renovations and are concerned about the impact of the Directive in relation to life cycles, and possible forced inefficient renovations.Regardless, tenants are expected to indirectly carry a cost burden associated with the mass renovation of the building stock. 

This report has been conducted to analyze how vibrations in timber floor structures, caused bywalking frequencies, are perceived by the users of the building. A question of interest has beento quantify how many users of a building are likely to experience discomfort due to vibrationsin a structure made of timber versus concrete. The report includes an analysis of how to designfloor systems with respect to vibrations by reviewing different standards that address vibrationperformance. At the time of writing, questions regarding vibrations in timber floors with longerspans frequently arise from developers, particularly concerning which reference values shouldbe used.The report has resulted in a thorough investigation into the limit values provided by, amongothers, Eurocode 5, and the origins of those limits. Through literature studies, it has becomeclear that the key factors influencing vibration performance are the natural frequencies of thefloor system. The parameters influencing natural frequency are mass and stiffness, whichbecome the critical aspects when addressing vibration issues in floors.Analyses conducted in FEM-Design have shown that floor structures with natural frequenciesaround or below 8 Hz have higher acceleration levels when subjected to a walking frequency of1,8 Hz. This supports the Eurocode 5 recommendation that a special investigation should becarried out for floors with a natural frequency below 8 Hz.It has proven difficult to estimate how many more users would perceive disturbances fromvibrations in a cross laminated timber (CLT) floor compared to an hollow-core concrete slab(HD/F) floor. The conclusion of this report is that, from a technical standpoint, it is possible todetermine whether a floor will exhibit dynamic effects that pose a risk of disturbing vibrations.However, it remains challenging to quantify what proportion of users will find such vibrationsuncomfortable or disturbing.

Transitions of transport systems are necessary to meet climate targets and reach sustainable development goals. Sustainable transport needs to be equitable, efficient, safe, and green. Transport systems need to become less resource-intensive, whereby sharing and utilising resources more efficiently can be a route forward. However, for well-functioning sharing, the involved stakeholders need to collaborate. Thus, sharing resources can mean sharing material and physical resources, such as vehicles and places, but it can also mean sharing immaterial and intangible resources, such as time and responsibilities. Through an interview study exploring shared resources in goods deliveries at off-peak hours in Stockholm-based trials, opportunities for organising transport system innovations have been investigated. Interviews were carried out with thirteen respondents involved in, or related to, off-peak logistics, and trigger materials were used to deepen the conversation when discussing the logistics service systems. Off-peak trials can be seen as innovation processes affected by several conditions. Innovation processes are collaborative, and collaborative innovation involves decision-making coordination across organisational boundaries. The concepts of sensemaking and sensegiving are used to analyse the studied collaborative innovation processes. However, stakeholders engaged in innovating sustainable transport systems must find new ways to collaborate, as ambiguity and uncertainty are more extensive than in their day-to-day business development or the incremental development of new services. This study shows that conditions for such collaborative innovation processes vary between different types of collaborative settings, as innovation depends on the transformative capacity to alter types of stakeholder agreements and roles and agree on how to share resources and design such sharing. Designing for resource sharing requires understanding how to design tangible and intangible buffers, and for the involved stakeholders to make sense of such new designs.

The function of proteins is enabled by their conformational landscape, with many proteins transitioning between multiple states that are crucial for their biological activity. While AlphaFold2 (AF2) has transformed protein structure prediction, it often fails to sample alternative conformational states critical for understanding protein function. Here, we evaluate AlphaFold3's (AF3) capability to sample diverse conformational states across 20 proteins with experimentally determined open and closed structures, and investigate whether MSA manipulation techniques developed for AF2, AFCluster and AFSample2, enhance this capability. Our analysis demonstrates that AF3 inherently samples conformational space more effectively than AF2, successfully predicting both conformational states with high accuracy for most targets. The MSA manipulation methods that significantly improved AF2's sampling of alternative states showed less effectiveness when applied to AF3, as opposed to AF2. We speculate that this is due to fundamental architectural differences, particularly AF3's pairformer and diffusion modules which differ from the evoformer in AF2. For proteins where native AF3 failed to capture alternative states, AFSample2 and AFCluster occasionally recovered missing conformations, with AFSample2 performing better on most proteins. AF3's architecture thus provides better conformational sampling than AF2 but shows decreased responsiveness to MSA perturbations.

Att minska bilberoendet och övergå till mer hållbara transportalternativ är en av vår tids stora utmaningar och hållbar mobilitet har blivit en alltmer central fråga i omställningen mot ett fossilfritt samhälle. Det krävs innovativa angreppssätt för att förstå hur förändringar som att minska användningen av privata bilar kan åstadkommas. Det handlar inte bara om att minska koldioxidutsläppen utan också om att skapa mer hållbara samhällen där människor kan leva utan att vara så beroende av bilen. Men hur gör kan det göras i praktiken? Inom Mistra SAMS har vi arbetat med just denna fråga genom att experimentera med framtida lösningar. Vi har utvecklat och testat framtidsprototyper för hållbar mobilitet i verkliga miljöer genom så kallade living labs. Ett av dessa testområden är Riksten, ett bostadsområde i Botkyrka kommun, söder om Stockholm, där vi under nästan ett år har samarbetat med medborgare för att förstå hur man skulle kunna minska antalet bilresor.