Högskolebiblioteket i Halmstad Publikationer
Not a member yet
17134 research outputs found
Sort by
Samverkan mellan hem och förskola - en nyckel till barnens utveckling? : En kvantitativ studie om hur samverkan skattas av vårdnadshavare och förskolepersonal.
No full textSamverkan mellan vårdnadshavarna och förskolepersonalen lyfts oftast fram som en positiv kraft i barns utveckling och lärande. Den här studiens syfte innefattar vårdnadshavares och förskolepersonalens skattning av samverkan, samt om det finns några variationer. Detta gjordes med hjälp av en webbenkät som 22 förskolepersonal och 37 vårdnadshavare i två kommuner i sydvästra Sverige svarade på. Resultatet i studien visar på att samverkan skattas högt men att det finns vissa variationer i deras skattning. Den dagliga kommunikationen, vad som anses vara relevant i innehållet men även vad som ansågs vara relevant information varierade. En slutsats i studien var att det krävs ett kontinuerligt arbete för att skapa en gemensam förståelse mellan vårdnadshavare och förskolepersonal och den dagliga kommunikationen. Genom den kan en god samverkan byggas. Collaboration between guardians and preschool staff is often highlighted as a positive force in children`s development and learning. The purpose of this study includes the assessment of collaboration by both guardians and preschool staff, as well as whether any variations exist. This was done through a web survey answered by 22 preschool staff and 37 guardians in two municipalities in southwestern Sweden. The results of the study show that collaboration is rated high, but that there are certain variations in their assessment. Daily communication, what is considered relevant in the content and what was regarded as relevant information varied somewhat. One conclusion of the study was that continuous effort and mutual respect are required to create a shared understanding between guardians and preschool staff, especially through daily communication. Through this, good collaboration can be built.
Dance for people with Parkinson's disease brings happiness, well-being, and enhanced mobility
No full textThis project has its starting point in a dance class with 14 people with Parkinson’s Disease who participated in Dance for Parkinson’s Disease® (Dance for PD). PD is a progressive neurodegenerative disorder associated with symptoms such as tremors, freezing, slowness of motion, and non-motor symptoms such as fatigue, sleep disturbances, and neuropsychiatric disorders. The purpose of this study was to explore the dancers’ experiences of participating in Dance for PD. The study was designed with a qualitative structure (framework). Data collection included diary entries after every dance session and, finally, two focus group interviews. All data were analyzed with thematic analysis. Three themes from all the data emerged: Maintaining everyday life, Working with health conditions, and Transformation to a happier life. The findings suggest that Dance for PD can be a complement to treatment. Attending the dance classes was an important activity in the lives of the participants. The participants in Dance for PD experienced improved physical mobility and well-being, and through the dance they gained an increased confidence in their own abilities and dared to challenge other aspects of everyday life. The dance teacher’s pedagogy with varying music and movements adapted to the participants became important for the actual experience of the dance. The dance promoted social interactions, and the participants experienced togetherness. In the study, Community of Practice (CoP) serves as a lens for discussing the findings. © 2025 Elenita Forsberg, Kristina Ziegert, Lars Kristén
Time-Series Anomaly Detection in Heavy-Duty Electric Vehicles : In collaboration with Volvo Trucks
No full textThe increasing electrification of heavy-duty vehicles creates new de- mands for reliable monitoring and diagnostics of fleet operations. De- tecting anomalies in energy consumption and operational behaviour is particularly challenging due to the scarcity of labeled fault data and the variability introduced by routes, loads, and environmental conditions. This thesis investigates how representation learning, and specifically Echo State Networks (ESNs), can support unsupervised anomaly detection in this context. The study combines controlled synthetic benchmarks with real- world telemetry from Volvo Trucks. On the synthetic side, faults of varying types and difficulties were injected into canonical signals. Several representation-detector pipelines were compared, including rolling statistical moments, principal component analysis, and ESN- derived features, coupled with detectors such as Isolation Forest, CO- POD, and one-class SVM. Results showed that ESN-based representa- tions consistently improved anomaly detection, particularly for com- plex fault dynamics, while rolling statistics provided a competitive baseline for slope-type deviations. On the real dataset, two applications were pursued. First, route- level screening demonstrated that ESN embeddings distinguished atypical trajectories within start-end groups, aligning with indepen- dent dynamic time warping clusters. Second, energy-based anomaly detection integrated ESNs with the Consensus Self-Organized Mod- els (COSMO) framework. This approach identified trips with atypi- cal energy-demand patterns that correlated with interpretable oper- ational factors such as load, road gradient, and temperature. These findings highlight the value of ESNs as compact and effective tempo- ral encoders, and of COSMO as a consensus-based scoring method that stabilizes anomaly decisions across heterogeneous fleets. Beyond anomaly detection, the thesis introduces a complexity ar- chitecture mapping that relates reservoir topology to signal complex- ity. A synthetic ladder of signals was used to calibrate reconstruction fidelity across reservoir families, showing that compact reservoirs suf- fice for simple dynamics while structured multi-layer topologies are advantageous for nonlinear mixtures. Applying the same analysis to real fleet energy data confirmed that architectural organization, rather than connection density, is the key driver of performance. Overall, the contributions of this thesis are threefold: (i) systematic benchmarking of representation-detector pipelines for anomaly de- tection in synthetic signals; (ii) demonstration of ESN- and COSMO- based methods for interpretable fleet-level diagnostics on real electric truck telemetry; and (iii) an empirical mapping between reservoir structure and signal complexity, providing guidance for architectural selection. Together, these results advance the development of scalable and interpretable monitoring solutions for heavy-duty electric vehi- cle
The Impact of AI on Employment Trends in Sweden : A Quantitative Secondary Data Analysis within the Nordic Welfare Framework
No full textThis analysis considers the effect of Artificial Intelligence (AI) technology on Sweden’s work opportunities between 2015 and 2024, viewed through the lens of its Nordic welfare model. AI adoption during the projected timeframe will put 46% of Swedish jobs at high or medium risk of losing automation by 2040, mainly impacting retail, manufacturing, healthcare, and finance. The analysis takes a quantitative approach by assessing sectoral employment changes, skill-level alterations, and the impact of labour market policies. Guided by the Routine-Biased Technological Change (RBTC) and Skill-Biased Technological Change (SBTC) frameworks, the AI impacts mid-skilled versus high-skilled employment negatively and positively, respectively. Sweden’s institutional weaknesses high union density and strong collective bargaining combined with Active Labour Market Policies (ALMPs) help lessen the damage in mid- to high-skilled employment through effective retraining and transition support. Comparative analysis against other Nordic countries shows Sweden’s subdued but efficient policy response as moderately unique. The adoption of AI alongside proposed tailored policies and interventions focused on equitable employment shifts highlights the importance of not only equitable demographic targeting but also adaptive welfare frameworks. This study documents the steps that proactive welfare states can take to reduce the socio-economic risks of automation for developing a proactive policy approach to AI
Usefulness of Synthetic Data in Biometric Recognition
No full textOcular recognition plays a crucial role in scenarios where masks or partial faces limit the use of face recognition. However, privacy regulations and restrictions have hindered access to large-scale public databases necessary for training and evaluating recognition models. This thesis investigates whether synthetic face data can substitute real images while retaining the detail required for ocular recognition. Ocular crops were extracted from two real datasets, VGGFace2 and AgeDB, and from two synthetic datasets, GanDiffFace and DCFace. We evaluated five ResNet50 models trained under different conditions: VGGFace2 (face), VGGFace2 (ocular), GanDiffFace (ocular), Glint360K (face baseline), and ImageNet (baseline). Identification performance was measured using Rank-1 Accuracy, and verification was measured using Equal-Error Rate (EER). The model trained only on synthetic ocular data achieved 99% Rank-1 Accuracy and 4.9% EER on synthetic test data, but performance dropped to below 15% Rank-1 Accuracy and approximately 32% EER on real ocular images. In contrast, the model trained on real ocular images reached 68% Rank-1 Accuracy and 11.6% EER on real data, and 84% Rank-1 Accuracy and 9.6% EER on synthetic data. These results suggest that while the ocular region of current synthetic faces is suitable for benchmarking, they are not yet a suitable replacement for real data in training and evaluating models intended for real-world deployment
QuickTrust : Produktförnyelse av vakuumassisterad biopsinålhållare med tillhörande adapter
No full textThis bachelor ́s thesis was conducted in collaboration with Turon Medtech AB and aimed to innovate and improve a vacuum-assisted biopsy (VAB) needle holder and its corresponding adapter. These components are used to mount a vacuum-assisted biopsy needle during breast biopsy procedures. The goal was to improve usability for healthcare professionals and enhance patient comfort by further developing the company ́s existing needle holder and adapter. The development process was carried out through an iterative design process, beginning with a needs identification phase based on work analysis in a clinical setting. This was followed by the creating of a requirement specification and conceptual design phase, where analogy modeling was used as a supporting tool. The generated concepts were systematically evaluated using the Pugh matrix and expert opinions from clinical professionals and mechanical engineers. Several plastic prototypes were manufactured in plastic to enable practical testing and further development. In the final design stage, aluminum 7075-T6 along with a surface treatment of 20 natural anodizing was selected as the material for the final product, and the solution was verified through a Finite Element Method (FEM) analysis. The solution developed in this bachelor ́s thesis was named QuickTrust, a mechanical locking device consisting of a VAB-needle holder, an adapter, and two removable locking arms. The design enables quick and secure one-handed assembly and disassembly, along with locking indicators that ensure correct position during biopsy. The modular design in aluminum 7075-T6 along with a surface treatment is optimized for efficient cleaning and withstands the expected mechanical loads. The bachelor ́s thesis thus presents a technically verified design solution with the potential to improve workflow, increase patient safety, and enhance both user and patient experience during breast biopsy procedures. Kandidatuppsatsen gjordes i samarbete med företaget Turon Medtech AB och syftade till att produktförnya en vakuumassisterad biopsi (VAB) nålhållare och en tillhörande adapter. Dessa används för att montera en vakuumassisterad biopsinål vid bröstbiopsier. Målet var att förbättra användarvänligheten för vårdpersonal samt öka patientkomforten genom att vidareutveckla företagets befintliga VAB- nålhållare och tillhörande adapter. Utvecklingsarbetet bedrevs genom en iterativ designprocess som inleddes med behovsidentifiering baserad på arbetsanalys inom klinisk miljö. Därefter togs en kravspecifikation fram, följt av konceptuell utformning där analogimodellering användes som stöd. De framtagna koncepten utvärderades systematiskt med hjälp av Pugh-matris och expertutlåtanden från både klinisk expertis och konstruktörer. Flera prototyper tillverkades i plast för att möjliggöra praktisk testning och vidareutveckling. I det avslutade konstruktionsarbetet valdes aluminium 7075-T6 ytbehandlat med 20 naturlig anodisering som material för den färdiga produkten, lösningen verifierades genom en Finita Elementmetod (FEM) analys. Lösningen som utvecklades i kandidatuppsatsen namngavs QuickTrust, en mekanisk låsanordning bestående av en VAB-nålhållare, en adapter och två avtagbara låsarmar. Konstruktionen möjliggör snabb och säker montering och demontering med enhandsfattning samt tydliga låsindikatorer som säkerställer korrekt positionering vid biopsi. Den modulära designen i aluminium 7075-T6 med ytbehandlig är anpassad för effektiv rengöring och de förväntade mekaniska påfrestningarna. Kandidatuppsatsen presenterar därmed en tekniskt verifierad designlösning med potential att förbättra arbetsflöden, öka patientsäkerheten och höja både användar- och patientupplevelsen vid bröstbiopsier.
Design and welding optimization for steel reinforcements of forklifts
No full textFatigue failure is a significant challenge in welded structures, particularly in forkliftreinforcements subjected to cyclic loading. This study aims to optimize the design andwelding configurations of steel reinforcements to enhance fatigue life and reduce stressconcentrations at critical weld locations. The research involves numerical analysis usingFinite Element Method (FEM) simulations, hand calculations, and topology optimization toassess and improve the structural performance of welded structures.A comparative analysis is conducted between a baseline model without reinforcement and animproved design incorporating welded stiffeners. Initial results indicate that reinforcementsignificantly reduces stress concentrations, thereby increasing fatigue resistance. FEMsimulations and analytical calculations are used iteratively to validate the structural integrityand determine the optimal reinforcement configuration.Furthermore, the study investigates whether weld sizes can be reduced in laser-cut steelreinforcements without compromising fatigue strength. Design recommendations aredeveloped based on guidelines from SSAB, Eurocode, and IIW standards. The outcome isexpected to provide a refined reinforcement design that enhances durability and performancewhile optimizing material usage and manufacturing costs.This research employs an Agile methodology, ensuring continuous improvement throughiterative cycles of design, analysis, and validation. The results contribute to the developmentof optimized forklift reinforcements, offering a practical solution for fatigue-related failures inheavy-duty industrial applications
Long-Term Effect Of Ombrotrophic Peatland Rewetting On Soil Organic Carbon, Bulk Density, Peat Thickness And Carbon Stock
No full textEn jämförande kvalitativ studie av cellulosa- och stenullsisolering
No full textThe construction industry is currently facing increasing demands to contribute to a more sustainable future, as global challenges such as climate change and the increase in carbon emissions are becoming increasingly critical. Awareness of the environmental impact of the building sector has grown significantly, leading to a heightened interest in circular solutions, resource efficiency, and environmentally friendly material choices. Traditional construction methods are increasingly being questioned, and the industry is shifting towards new approaches where sustainability is a central focus. This study aims to investigate the potential of replacing conventional stone wool insulation in a wall element with recycled cellulose insulation. The objective is to reduce environmental impact without compromising quality or performance. By comparing the materials in terms of moisture performance, fire classification, energy efficiency, and cost, the study explores whether cellulose can serve as a viable alternative in contemporary building constructions. The research is based on a practically applicable case and relates its findings to current regulations and industry standards, highlighting both the possibilities and challenges involved in the transition toward more sustainable insulation solutions
Design och prototyputveckling av en väder- och vattentålig elektronisk inkapsling för utomhusbruk
No full textThis bachelor thesis, carried out in collaboration with Jelmtech Produktutveckling AB, aimed to develop a wall-mounted electronic enclosure that fulfills IPX7 classification requirements, user-friendliness, sustainability, and suitability for large-scale production. The development process included literature review, material selection using the Ashby method, CAD modeling in 3DEXPERIENCE, and prototype manufacturing through 3D printing. The final design is developed to be manufactured through injection molding of the material ASA with Molded-In-Place Gasket (MIPG) sealing application of the material TPE. A simplified water test in 20 cm deep water demonstrated successful resistance to ingress, even without the use of snap-fit features, indicating promising design performance. However, the formal IPX7 immersion test failed due to deviations in material and manual sealing application, and the 3D printer’ s ability to print out the modular design fully, emphasizing the importance of correct manufacturing methods. Despite these limitations, the project is considered successful, and the developed solution shows strong potential for industrial implementation.