Chalmers Open Digital Repository

Welcome to Chalmers Open Digital Repository!

Here you can find:

Student theses and papers
Digital special collections, such as Chalmers modellkammare
Selected project reports

Research publications, reports and dissertations can be found in research.chalmers.se

Communities in Chalmers ODR

Select a community to browse its collections.

Now showing 1 - 2 of 2

Recent Submissions

Edge–Cloud Inference Deployment for Perception-Driven Automation - A systems evaluation of stereo-vision assisted zone-occupancy for building automation control under practical accuracy, latency, and energy constraints
(2026) Palm, Eric; Däldborg, Alexander
This thesis evaluates how edge and cloud inference affect a stereo-depth-assisted, zone-based lighting controller for building automation. The system combines person detection with depth-based spatial assignment to estimate whether predefined lighting zones are occupied. Several model configurations are compared under edge and cloud deployment using controlled scenarios and full-day analytical lighting reconstructions. The results show that deployment placement affects both perception and system behaviour. Cloud inference improves recall in more difficult conditions, such as low light and obstruction, but introduces transport delay, queue behaviour, higher measured processing energy, and greater privacy exposure. Edge inference provides lower processing-energy use, avoids remote communication overhead, and keeps more data local, but is less robust in difficult visual conditions. Across the evaluated configurations, the projected zone-aware controller reduces lighting energy compared with less spatially specific baseline strategies.
Den artificielle juristen- en utvecklad teknisk lösning för framtidens jurister
Sandén, Loke
The rapid development of generative artificial intelligence (AI) is creating new opportunities and challenges within the legal profession. While AI is increasingly used to support legal research, document analysis and information retrieval, there remains a need for educational environments that enable legal professionals to develop practical skills in using AI responsibly and effectively. This study investigates how an AI-based simulation environment can be designed and implemented to support legal education and professional competence development. The project was conducted in collaboration with Semantik AB and followed the Design Science Research (DSR) methodology. A literature review based on the PRISMA framework was carried out to identify current research concerning generative AI in legal education and legal practice. The findings from the literature review informed the design and development of a prototype consisting of a web-based simulation environment where users interact with AI-generated characters representing different actors in a legal case. The system was designed using a flexible architecture that integrates large language models through an adapter-based backend solution. The developed artifact was evaluated through user testing involving legal professionals. Although technical constraints limited the scope of the evaluation, the results indicate that the system was perceived as intuitive, engaging and pedagogically valuable. The evaluation also revealed that users adopted different strategies for information gathering and problem solving, highlighting the importance of supporting effective interaction with AI systems. The findings suggest that generative AI should be viewed primarily as a complement to legal expertise rather than a replacement for human judgment. The study further demonstrates that AI-based simulation environments have the potential to support legal education by providing realistic, interactive and practice-oriented learning experiences. Finally, the study highlights the importance of addressing ethical considerations such as transparency, bias, accountability and data protection when integrating AI into legal education and professional practice.
Beräkningskemisk modellering för fotoemissionsspektroskopi samt nedbrytningsmekanismer för batterielektrolyter
(2026) Grut, Erica; Linder, Cornelia; Janemyr, Noel; Richardsson, Axel; Olofsson, Tor
Computational chemical modeling for photoemission spectroscopy and decomposition products of battery electrolytes. The project examines how quantum chemical methods can simulate x-ray photoelectron spectroscopy (XPS) peaks and theoretically analyze decomposition mechanisms for the electrolyte molecules etyhlene carbonate (EC), propylene carbonate (PC) and fluoroet hylene carbonate (FEC) in lithium-ion batteries. By benchmarking quantum chemical methods against experimental data, suitable methods for simulating XPS-spectra and the reaction energetics of the electrolyte decomposition products that balance accuracy and computational cost could be identified. XPS-simulations of ethyltrifluoroacetate (ESCA) show good agreement with experi mental reference data. Simulations of EC and PC were also made, which exhibit good conformity with values within 0.5 eV of the experimental values. These simulations support computational methods as a reliable way of predicting binding-energies. Further on, electrochemical reaction pathways were analyzed with transition state the ory and Marcus theory. The results shows that EC, PC and FEC has similar decompo sition mechanisms in lithium-ion batteries during operation. Additionally, results show that different calculation methods predict different favorable reactions, where Marcus theory generally predicts lower activation barriers and therefore more favorable electron transfer reactions. Explicit transition state calculations tend to predict high activation barriers, especially in ring-opening reaction steps where the reorganization energies are the driving force. It also indicates that FEC has a higher reduction potential and a faster reduction compared to EC and PC, which supports its function as an additive to enhance the solid electrolyte interface (SEI)-formation. In conclusion, this project shows how quantum chemical methods can be versatile in understanding both XPS-spectra and decomposition mechanisms in lithium-ion battery electrolytes with specific usage in analysis of SEI-formation.
Lokal struktur hos metallhalidperovskiter undersökt med vibrationsspektroskopi
(2026) Curkic, Elma; Eidstedt, Lina; Loman, Janna; Swartz, Emilia
Local Structure in Metal Halide Perovskites studied with Vibrational Spectroscopy The rapid development of metal halide perovskites has during the last decade positio ned the material as one of the most promising candidates for future energy technology such as solar cell technology, light emitting diodes, and photodetectors. Because metal halide perovskites are very flexible and in combination with the ease of synt hesizing the material, they are particularly attractive to use for technical applications. This bachelor’s thesis aims to analyze the local structure of the metal halide perov skite EAxMA1−xPbI3 and in particular to see how the local structure is affected by the cation composition x. By analyzing how the chemical composition affects the structure of the material at room temperature, new knowledge is obtained, which is of purely scientific interest, but can also be used as a reference for the purpose of optimizing stability and efficiency of the material. The scientific method for these studies involves infrared (IR) spectroscopy and Raman spectroscopy, together with studies of the materials photoluminescence. The reason behind this thesis is the stabilization challenges that currently limit the practical application of perovskites. The well-researched material MAPbI3 exhibits good qualities, but is sensitive to chemical degradation. Introducing EA+ is relevant since it shows potential to stabilize the structure, though it has not yet been studied how the substitution affects the properties of the material. This thesis shows how the cation composition has a direct and measurable impact on the local structure of the metal halide perovskites. The results present useful information that corresponds to relevant literature. The thesis might be useful in the next generation’s technical applications with customized features.
Mekanokemisk syntes av metall organiska ramverk för katalytiska väteutlåningsreaktioner
(2026) Andersson, Oskar; Ivarsson, Lydia; Lam, Elina; Svensson, Emil; Wedding, Elsa
Metal-organic frameworks are porous coordination polymers with high specific surface area and flexibility upon functionalisation, which makes them interes ting for heterogeneous catalysis. In this study, metal-organic frameworks were synthesised mechanochemically and subsequently evaluated for their catalytic ac tivity in borrowing hydrogen reaction for N-alkylation of aniline with benzyl alcohol. Mechanochemical synthesis was performed on five metal-organic frameworks, NH2-MIL-88(Fe), NH2-MIL-125(Ti), MIL-100(Fe), Zr-MOF-818, and Cu-MOF-74. They were then characterised using powder diffraction, thermogravimetric analysis, Fourier transform infrared spectroscopy, and gas adsorption to investigate structure, crystallinity, and thermal stability. The results showed that only NH2-MIL-88(Fe) and NH2-MIL-125(Ti) were successfully synthesized. They were then tested for their activity in borrowing hydrogen reactions, also called hydrogen autotransfer reactions, and the reaction mixtures were subsequently analysed with nuclear magnetic resonance spectroscopy. Both frameworks presented catalytic activity, however NH2-MIL-88(Fe) showed the best result under the investigated reaction conditions. Furthermore, the reaction outcome improved with extended reaction time and the addition of a base. The project concludes that mechanochemical synthesis is one possible met hod for producing metal-organic frameworks, but the method entails restrictions with crystallinity and reproducibility. The results show that these materials possess catalytic activity and therefore be suited for borrowing hydrogen reactions.