240106 research outputs found
Sort by
Innovation as Resistance: Grabbing Back Life on the Coast in the Blue-Green Economy
No full textThis thesis explores how coastal communities in Denmark respond to socio-economic marginalization through locally anchored innovation within the emerging blue-green economy. Drawing on Ethnographic fieldwork in and around Hirtshals, North Jutland, it examines how social innovation initiatives leverage belonging as a spatial, temporal, affective, and political quality to resist depopulation and peripheralization. The thesis focuses on three cases: The North Sea Free Trade School, the Workforce of the Future program, and Hirtshals Winter Festival. These efforts reveal how future imaginaries, rooted in the lived experiences of coastal life, serve as tools for ”grabbing back” agency and value locally. By extending the concept of adjacency beyond geography to include the politics of belonging, this thesis illustrates how innovation becomes a means of resistance, serving both diagnostic and reparative function in coastal spaces. The findings contribute to broader discussions on sustainability, marine governance, and equitable participation in the blue-green economy
The Evolution of the Female Role in Horror Cinema
No full textThis thesis examines the evolving representation of women in horror cinema, tracing the shift from passive victims to complex figures who are given agency. By using feminist film theory as its foundation, this study examines eight key films across decades spanning from Alfred Hitchcock’s Psycho (1960) to Coralie Fargeat’s The Substance (2024). Drawing on theoretical frameworks from Laura Mulvey, Carol J. Clover, Barbara Creed, Julia Kristeva, and Cynthia A. Freeland, the thesis investigates how female characters have been framed, fetishized, and ultimately reimagined within the horror genre. The analysis highlights how early horror films often reinforced patriarchal narratives, positioning women as objects of violence and desire. However, later works, especially those written and directed by women, such as Jennifer’s Body, Promising Young Woman, and The Substance, subvert these conventions by presenting female characters who confront, embody, or weaponize their monstrosity. These films use horror not only as a genre of fear, but as a political and affective space in which the traumas and injustices of gendered experience can be critically exposed. This thesis argues that contemporary feminist horror, what Barbara Creed terms the “Feminist New Wave,” reclaims female monstrosity as a symbol of empowerment. By exaggerating, duplicating, or undermining traditional horror tropes, these films resist the voyeuristic dynamics of the “male gaze” and rewrite the cultural scripts around femininity
Leveraging LLMs to Generate Business Plans
No full textIn today's competitive business environment, well-structured business plans are essential for startups to be successful and secure funding, yet many small businesses lack the resources to create them effectively. To provide a convenient solution for the stated problem, this research investigates how Large Language Models (LLMs) can be utilized to generate business plans and examines the benefits and limitations compared to traditional planning methods. A multi-agent system was developed using the CrewAI framework, employing six specialized agents for different business plan sections, with Google's Gemini 2.5 Pro for content generation and Meta's Llama 3.3 70B for initial evaluation. The system was tested using both expert-provided fictional company data and synthetic data generated from existing business plans using GPT 4.1. Evaluation employed both human expert judgment and LLM-as-a-Judge techniques across five metrics including relevance, completeness, correctness, consistency, and clarity. Results demonstrate that LLMs can generate structurally coherent business plans with relevant content across all sections, offering significant advantages in speed, cost-effectiveness, and accessibility for startups. However, the generated plans can be utilized as drafts only, since limitations include hallucination tendencies, inconsistencies between sections, and lack of deep business insights that human expertise provides. The research concludes that AI-generated business plans serve best as starting points and still requires human review and refinement, representing a complementary tool rather than a complete replacement for traditional business planning methods
From Earth to Education: A Sustainable Approach to Climate Resilient School Design in Rural Zambia
No full textThis thesis proposes the design of a secondary boarding school in rural Zambia, in which different strategies are integrated to make the design climate resilient as well as providing a high-quality learning environment. The project is designed with a community driven approach, ensuring the needs of the users are fulfilled. An additional focus area is the use of simple construction methods from locally sourced materials, enabling the local community to take part in the building process. Furthermore, this thesis examines the relationship between a comfortable indoor climate and the environmental costs associated with achieving it in the context of a developing country. The final design seeks to reach a balance between environmental impact, indoor climate and user comfort, offering a solution that is both contextually relevant while still providing the design with architecturally qualitative spaces. Ultimately, this work advocates for a thoughtful approach to school design. An approach that embraces environmental responsibility, strengthens community engagement and involvement, and creates an educational environment through climate-conscious decisions
Development of a Digital Information Material for Women with Inflammatory Bowel Disease in Relation to Pregnancy: A Case Study
No full text
Truncated Bike Lanes at Signalized Intersections - An Investigation of the Relationship Between Behavior, Design, and Perceived Safety
No full textControl of Satellite Swarms
No full textThis project aims to test the feasbility of controlling two cube satellites with magnetorquers for swarming capabilities. This novel concept for swarm control based on magnetic actuation is investigated. The success criterion is to make the controller generate enough electromagnetic force to have two satellites attract and settle at some desired distance in between.A linear state-space control strategy is developed for a highly non-linear system by a means of translation and scheduled gains. An analysis of this non-linear is conducted and an approximation of the model is done. This is followed by linearization at the desired operating point and also at different points in order to schedule gains for the linear controller. A fictitious force is calculated that acts on the satellites in different orbits around Earth, which acts as the minimum requirement for the feasibility of this project. This is followed by the controller design.The strategy, namely, observer based control with full state-feedback is finally implemented in MATLAB Simulink platform and tested for the derived requirements
Capturing Transformer Overloads through Ensemble-Based Forecasting of EV Charging Station Consumption: An Evaluation of Model Diversity and Its Impact on Detection Performance
No full textGivet den store udvikling af elbiler og de globale FN's verdensmål om Bæredygdige byer og samfund, stiger antallet af ladestandere markant, hvilket belaster elnetværket mere for hvert år som går. Derfor vil studiet imødekomme udfordringerne ved at forudsige ladestandernes elforbrug for at mindske de risici som overbelastningerne kan forårsage på transformere. Ved at bruge tidsserie data aggregeret på time basis, vil dette studie analysere data fra virkelighedens verden fra byen Boulder i Colorado samt for et syntetisk dataset for danske hjemmeladestandere. Rapporten vil gennemgå to forsøg, hvilket er: (1) at forudsige elforbruget vha. MAE og Huber loss samt (2) opfange overbelastninger vha. recall og MAE/Huber loss. Til dette har rapporten udarbejdet en ensemble model, som med forskellige base learners (xPatch, PatchMixer, GRU, LSTM, AdaBoost, Random Forest, Gradient Boosting) trænes med bootstrap sampling. Studiets resultater peger på at deep learning modeller, især PatchMixer og xPatch, udkonkurrerer ensemble modellen og de individuelle modeller, både i forudsigelser og i at opfange overbelastninger, dog med udfordringer i at fange høje værdier. Idet ensemble modellens resultater ikke var fyldestgørende, var problemer såsom ustabile konfigurationer samt dårligt ydende base learners en af hovedårsagerne til dens dårlige præstation.The rapid adoption of electric vehicles (EVs), driven by global sustainability goals, has led to widespread deployment of electric vehicle charging stations (EVCS), raising concerns about the stability of existing power infrastructure. This study addresses the challenge of forecasting EVCS electricity consumption to help mitigate transformer overload risks. Using time series data aggregated at an hourly resolution, we analyze a real-world dataset from Boulder, Colorado and a synthetic Danish residential dataset.Two main tasks are explored: (1) consumption forecasting using MAE and Huber loss, and (2) overload detection using recall and MAE/Huber Loss. An ensemble model is proposed, combining diverse base learners (e.g. xPatch, PatchMixer, GRU, LSTM, AdaBoost, Random Forest, Gradient Boosting) trained on bootstrap samples. Results show that deep learning models, particularly PatchMixer and xPatch, outperform ensemble and standalone models in both forecasting and overload detection, though all models struggled with peak consumption values. While ensemble performance was mixed, issues were largely attributed to unstable configurations and underperforming base learners