Blekinge Institute of Technology
Not a member yet
13576 research outputs found
Sort by
ThreMoLIA : Threat Modeling of Large Language Model-Integrated Applications
No full textLarge Language Models (LLMs) are currently being integrated into industrial software applications to help users perform more complex tasks in less time. However, these LLM-Integrated Applications (LIA) expand the attack surface and introduce new kinds of threats. Threat modeling is commonly used to identify these threats and suggest mitigations. However, it is a time-consuming practice that requires the involvement of a security practitioner. Our goals are to 1) provide a method for performing threat modeling for LIAs early in their lifecycle, (2) develop a threat modeling tool that integrates existing threat models, and (3) ensure high-quality threat modeling. To achieve the goals, we work in collaboration with our industry partner. Our proposed way of performing threat modeling will benefit industry by requiring fewer security experts' participation and reducing the time spent on this activity. Our proposed tool combines LLMs and Retrieval Augmented Generation (RAG) and uses sources such as existing threat models and application architecture repositories to continuously create and update threat models. We propose to evaluate the tool offline - i.e., using benchmarking - and online with practitioners in the field. We conducted an early evaluation using ChatGPT on a simple LIA and obtained results that encouraged us to proceed with our research efforts.
Correcting market failure for no-regret electric road investments under uncertainty
No full textSeveral electric road system technologies that enable in-motion charging of electric vehicles are nearing market readiness. However, substantial contribution to decarbonization requires rapid deployment on an international scale. Investment is discouraged by prior research that has identified that declining battery costs may eventually leave the infrastructure a stranded asset. We explore under what circumstances electric roads offer effective and low-risk decarbonization of European heavy-duty road freight. Transport system dynamics are explored and quantified, through pairwise comparison of scenarios with and without electric road incorporation, using a purpose-built agent-based simulation (MOSTACHI). Prior stranded asset risks are confirmed, but we show that policy that encourages high electric road utilization can correct for market failures and make the infrastructure a no-regret investment in much of Europe – never yielding worse outcomes than not investing. Electric roads are shown to be an effective risk mitigation strategy, achieving market-driven phase-out of fossil fuels before 2050, also in pessimistic scenarios where static charging alone would be insufficient. Electric roads reduce levelized system cost by 0–17%, greenhouse gas emissions by 7–63% (2030 to 2050, cumulative) and battery mineral demand by 20–40%. Benefits are maximized with early and predictable deployment. Genomförbarhetsstudie elvägspilot E2
Bridging Cultural Divides: Tensions and Remedies in a Cross-Cultural Project Setting : Insights from a Chinese-Swedish Collaboration
No full textBackground: Multinational corporations and cross-cultural challenges have been extensively studied from various perspectives. As the number of Chinese multinational enterprises (MNEs) operating in Europe continues to grow, it becomes increasingly important to understand how these Chinese MNEs influence their western subsidiaries and the challenges associated with it. This thesis explores these challenges and the cross-cultural impact they have in a project setting. Purpose: The primary purpose of this thesis is to identify success factors to improve cross-cultural project management collaboration within a Chinese-Swedish setting. To achieve this, the thesis examines the cultural differences, assesses their impact on Chinese-Swedish collaboration, and identifies potential solutions to address these challenges. Method: The thesis employs an explorative qualitative approach based on a single case study of a Chinese original equipment manufacturer (OEM), investigating the dynamics between a Swedish subsidiary and its Chinese owner. Data was collected through four semi-structured interviews, primarily focusing on the Chinese perspective of collaboration. The data was analyzed using a hybrid thematic analysismethod. Results and analysis: The findings illustrate how national cultural differences influence decision making, communication and project management which shapes the perception on authority, collaboration, and project teamwork. These differences lead to inefficiencies in collaboration, project alignment, and timing. Identified solutions to these challenges include early project alignment, consistent communication, clear decision-making, and diligent follow-through to improve collaboration and trust. Conclusions: Enhanced collaboration and communication are key elements for success in a cross-cultural business setting. By understanding the needs of a cross-cultural organization and effectively adapting to differences, companies can not only reduce friction in collaboration but also unlock the full potential of a diverse team working toward a common goal. Recommendations for future research: How can an Eastern perspective contribute to established Western project management paradigms and what implications would this have for international project management? Can cross-cultural collaboration be improved by standardization? How does the cross-cultural organizational setup impact the relative significance of different cultural dimensions?
Augmentation of Friskvård : Potential Integration of AI Therapy with Sweden’s Friskvård Program
No full textBackground: Employee wellbeing plays a critical role in shaping organizational productivity and overall performance. Prior research has consistently demonstrated strong links between employee wellbeing, job satisfaction, and organizational outcomes. Swedish wellness benefit program, Friskvård, has traditionally focused on promoting physical health through activities such as gym memberships and sports. However, given the rising prevalence of mental health challenges in the workplace, this study explores whether AI-based mental health interventions—specifically AI therapy—could serve as a complementary component to the existing Friskvård framework, thereby addressing both physical and psychological aspects of employee wellbeing. Purpose: The purpose of this research is to study the potential integration of AI therapy into Sweden’s existing wellness benefit system, Friskvård. Specifically, the study aims to investigate whether AI therapy can coexist with and complement the current Friskvård offerings, which primarily focus on physical wellbeing. The research seeks to deepen understanding of the organizational impact of providing accessible AI-driven mental health support, while also identifying potential barriers to its adoption within Swedish workplaces and the Friskvård framework. Method: This study is based on a qualitative research design, incorporating semi-structured interviews with 14 participants from a single Swedish organization—comprising 7 managers and 7 individual contributors. In addition to primary data, the research builds on existing literature related to mental health and its organizational impact like overall employee mental health and absenteeism. A thematic analysis was conducted to identify and interpret key patterns within the data, resulting in the identification of seven central themes relevant to the study’s focus. Results and analysis: There is a strong, established emphasis on physical activities within Swedish organizations in the application of Friskvård benefits. A more comprehensive understanding and supportive regulatory framework is required for broader integration of AI therapy as part of employee wellness initiatives. Insights from the interview analysis indicate that improved employee wellbeing—through accessible mental health support—can contribute to reduced absenteeism and turnover rates, ultimately enhancing overall organizational performance. Conclusions: AI therapy can be integrated into Friskvård as a complementary tool, in Swedish organizations. Offering cost-effective mental health support to reduce work-related stress, boost employee engagement and satisfaction, and reduce absenteeism. The organizations could benefit from AI therapy as a strategic asset, aligning with organizational goals, talent retention advantages, increased productivity and better workplace culture although measuring the impact is complex. Recommendations for future research: A cross-organizational study with attention to habits, regulations and trust issues is needed for a further understanding of the implementation of AI therapy in organizations. The long-term effect of AI therapy on employees and cost effectiveness could be further studied to analyze workplace adoption
Evaluating Automation Investments : Integrating Cost-Benefit Analysis into theDynamo++ Decision-Support Model
No full textIn today’s manufacturing landscape, automation is widely recognized as a key driver of productivity, efficiency, and competitiveness. However, despite its potential companies often face challenges justifying such investments due to limited integration of financial analysis in existing decision-support tools. While the Dynamo++ framework provides structured evaluation of Levels of Automation (LoA) across mechanical and informational dimensions, it lacks a financial component essential for strategic investment decisions. This thesis addresses that gap by integrating Cost-Benefit Analysis (CBA) into the Dynamo++ method. A mixed-method case study was conducted at a Swedish truck manufacturer, focusing on two assembly stations. The study combined Dynamo++ with the Analytic Hierarchy Process (AHP), Benefit Change Scoring (BCS), and CBA. Data was collected using a mixed method approach. AHP was used to prioritize operational criteria, while BCS provided scenario-specific benefit scores. These were compared against investment costs using CBA. The results demonstrate that the integrated framework supports more balanced automation decisions by combining operational priorities with financial reasoning. Some lower-cost automation scenarios delivered strong benefit-to-cost ratios, while others with higher costs showed limited added value. Overall, the study demonstrates that combining Dynamo++ with CBA and supporting tools results in a more holistic and ransparent evaluation framework. This integrated approach bridges the gap between technical analysis and financialjustification, offering practical value for companies planning future automation investments in manufacturing environments
Chess Training for Cognitive and Social Enhancement
No full textChess has long been recognized as a powerful cognitive training tool for enhancing working memory and problem-solving. AI-based platforms have provided valuable tools for more structured and interactive chess training experiences. Previous research suggests that chess training can improve individuals' quality of life (QoL). This study investigates the question: How does chess training influence physical development and social engagement? Furthermore, it explores the broader inquiry of how serious games impact human physical and social engagement, serving as the foundation for this research. Over 14 days, participants followed structured digital chess AI-tool exercises, with half additionally engaging in AI-guided daily workouts. The purpose was to examine differences in engagement, skill development, and cognitive improvements between self-paced chess gameplay and AI-generated workout plans.Results indicate improved social engagement, focus, and problem-solving skills across various age groups and backgrounds. Beyond strategic thinking, chess training fosters motivation for daily activities such as exercise and social interactions. However, given the study’s limited scope, further research with a larger sample size and extended duration is required to explore these effects in depth.Schack har länge varit känt som ett kraftfullt verktyg för kognitiv träning, särskilt för att förbättra arbetsminne, problemlösning, mönsterigenkänning samt socialt och fysiskt engagemang. AI-drivna plattformar har ytterligare strukturerat schackträning och skapat interaktiva inlärningsupplevelser som stödjer färdighetsutveckling. Tidigare forskning visar att schackträning kan ha en positiv inverkan på livskvalitet (QoL) genom att främja både kognitiv och social utveckling. Denna studie undersöker den centrala frågan: Hur påverkar schackträning fysisk utveckling och socialt engagemang? Dessutom utforskar den bredare frågeställningen om hur serious games, särskilt AI-drivna schackverktyg, påverkar mänskligt engagemang och färdighetsförvärv. Under en 14-dagarsperiod deltog deltagarna i strukturerade AI-assisterade digitala schackövningar, där hälften även genomförde AI-genererade dagliga träningspass. Studien syftade till att analysera skillnader i engagemang, färdighetsutveckling och kognitiv förbättring mellan självstyrt schackspel och AI-genererade träningsprogram. Resultaten visar på förbättrad koncentration, problemlösningsförmåga och socialt engagemang över olika åldersgrupper och bakgrunder. Utöver de intellektuella fördelarna verkar schackträning även öka motivationen för dagliga aktiviteter, inklusive fysisk träning och sociala interaktioner. Dock krävs ytterligare forskning med större urval och längre tidsperioder för att utforska dessa effekter mer ingående och validera dess långsiktiga påverkan
Samspel inom havsplaneringen : En studie av Sveriges havsplan
No full textBehovet av att nyttja havets resurser växer varje dag, därför behövs det en väl fungerande planering över var och hur havet får nyttjas. Flera användningar och deras behov samexisterar med att nyttja havets resurser inom den svenska havsplaneringen. Eftersom användningarnas vikt varierar mellan de olika områdena och inom svensk lagstiftning kan det innebära att planeraren ställs inför ett utmanande uppdrag att balansera användningarna och deras behov. Därför är det viktigt att planeraren har en förståelse för hur användningar samspelar med varandra och att den har koll på vilka hjälpmedel som finns tillgängliga. Uppsatsen undersöker havsplanering där stort fokus var på användningar och planerings hjälpmedel. Syftet med uppsatsen har varit att undersöka samspelet mellan olika användningar och hur planeringshjälpmedel kan användas vid samordning. Utifrån syftet har två frågeställningar tagits fram: “Hur samspelar olika användningar i havsplanen” och “I vilken utsträckning kan olika planeringshjälpmedel bidra till samordning av olika användningar inom havsplanering”. Frågeställningarna medförde att forskningsstrategin som valdes var fallstudie där havsplanen utgjorde fallet. För att besvara första frågan har fallet avgränsats till havsplaneområdena Södra Östersjön och Sydvästra Östersjön- och Öresund. För att besvara den andra frågan valdes olika former av planeringshjälpmedel ur havsplanen samt från Havs- och vattenmyndighetens hemsida. Analysen visar att användningar samspelar inom olika nivåer och att dessa tas fram genom en tolkning av innehållet i havsplanen och havsplanekartan. De två havsplane områden som har undersökts visar på liknande förutsättningar, men olika prioriteringar och bestämmelser. Havsplaneområdena identifierar inte konflikter som förekommer utan visar endast på slut resultatet utan vidare förklaring till varför ett beslut har tagits. Under havsplaneringsprocessen används olika planeringshjälpmedel som underlättar vid avvägning mellan användningar. Sydvästra Östersjön- och Öresund visar på en större variation av användningar, vilket beror på dess geografiska plats och nationella roll. Däremot visar Södra Östersjön på mindre variation i användningar och speglar istället Sveriges vikt av försvaret. Analysen visar hur samexistens uppkommer och hur planeringshjälpmedel används i syftet för att uppnå en god havsplanering för Sverige.
Satellitdriven miljörekonstruktion i Unreal Engine : En förenklad process för bred användning
No full textPhotorealistic 3D environments are widely used in film, architecture, and AI training, yet recreating real-world locations using satellite data often requires expert knowledge and extensive time consuming manual effort. This limits the accessibility and range of applications for the technology. This bachelor's project aims to simplify the process by developing a user-friendly tool in Unreal Engine that automates environment generation using satellite imagery and elevation data. The tool is designed for users with only basic experience, while still offering advanced configuration options for more experienced users. The study follows an iterative design method, testing and refining technical solutions. The outcome is a functional prototype that generates landscapes, environments and vegetation with minimal manual input. The project demonstrates time savings and outlines the potential for further development of automated pipelines for reality-based 3D environments. Fotorealistiska 3D-miljöer används inom film, arkitektur och AI-träning, men processen för att återskapa verkliga platser baserat på satellitdata kräver ofta expertkunskap och omfattande tidskrävande manuellt arbete. Detta begränsar teknologins tillgänglighet och användningsområden. Detta kandidatarbete syftar till att förenkla denna process genom att utveckla ett användarvänligt verktyg i Unreal Engine som automatiserar miljögenerering med hjälp av satellitbilder och höjddata. Verktyget kräver endast grundläggande användarkunskaper, men erbjuder även avancerade inställningar för erfarna användare. Studien bygger på en iterativ designmetod där tekniska lösningar testas och förfinas. Resultatet är ett fungerande prototypverktyg som genererar landskap, miljöer och vegetation med minimal manuell insats. Projektet visar på tidsbesparingar och möjligheter till vidareutveckling av automatiserade pipelines för verklighetsbaserade 3D-miljöer.
Algorithms used for procedurallygenerated dungeons : A comparison between Binary Space Partitioning,Depth-First Search, 2D Delaunay Triangulation, and3D Delaunay Triangulation.
No full textContinuous SBOM Generation for Development Workflows : An Empirical Comparison with Build-Time Approaches and Runtime Dependency Detection
No full textSoftware Bill of Materials (SBOM) have become essential for software supply chain security, driven by regulatory mandates and increasing supply chain attacks. Traditional SBOM generation occurs at build time, creating static snapshots that quickly become outdated during active development and fail to capture runtime-loaded dependencies. This thesis addresses the critical gap between build-time accuracy and development-time security feedback by investigating continuous SBOM generation approaches integrated into developer workflows. This research employs Design Science Research methodology to develop and evaluate a continuous SBOM generation plugin for the Node.js ecosystem. The artifact integrates with Visual Studio Code to provide real-time dependency monitoring, automatic SBOM regeneration, and immediate vulnerability feedback during development activities. We conducted a systematic three-sprint empirical evaluation comparing continuous generation against established static approaches (Syft, CDXGen, Trivy) across direct dependencies, transitive dependencies, and runtime detection scenarios. Our empirical findings demonstrate that continuous SBOM generation maintains equivalent accuracy for direct dependencies (100% detection rate) while providing superior component discovery (8.8% increase, 1,674 vs 1,539 components) and significantly faster vulnerability feedback (12.8 seconds vs 5-15 minutes for CI pipelines). However, continuous generation exhibits reduced transitive dependency coverage (87.5% vs 95.3% for best static tools) and potential scalability limitations including memory accumulation patterns (23% increase across test runs). Runtime dependency detection achieves excellent discovery rates (100% for dynamic and environment[1]specific dependencies) but suffers from severe metadata completeness degradation (10-42% version coverage) that limits practical deployment for compliance or com[1]prehensive security assessment. The research contributes the first systematic empirical comparison of continuous vs. static SBOM generation methodologies, providing evidence-based guidance for tool selection and deployment strategies. We demonstrate that continuous generation is most suitable for small-to-medium development projects requiring immediate security feedback, while enterprises should maintain hybrid approaches combining continuous generation for development environments with static generation for comprehensive compliance coverage. Technical limitations including scale constraints (evaluation limited to 22 packages), ecosystem specificity (Node.js only), and unknown enterprise performance characteristics constrain generalizability and indicate clear directions for future research. This work provides a foundation for informed decision-making in enterprise SBOM implementation while identifying specific technical and methodological challenges that require continued research investment to achieve comprehensive software supply chain security