Publikationer från Linköpings universitet
Not a member yet
58266 research outputs found
Sort by
Mathematical Modeling of a Hematopoietic Progenitor Cell under Mechanical Load
No full textAseptic loosening, the failure of prosthetic fixation, is a common complication following joint replacement surgery and often requires revision procedures. This condition is linked to cellular stress caused by mechanical loading, which alters cellular behavior. One consequence is a change in the release of extracellular vesicles: under normal conditions, cells predominantly release microvesicles, but under increased mechanical load, they instead secrete exosomes. Additional intracellular responses include elevated ATP release, altered expression of key proteins, and shifts in intracellular signaling. Key components of this cellular response include Ca2+ homeostasis, where SERCA2, an intracellular calcium pump, plays an important role, as well as the ATP release and the expression of ACP5/TRAP, a marker linked to bone remodeling. In this thesis, a mathematical model of a hematopoietic cell was developed using ordinary differential equations (ODEs) to simulate intracellular dynamics under mechanical loading. Due to the complexity of biological systems, the model was constructed part-wise, with iterative hypothesis testing aimed at aligning the model with experimental observations. The available data reflected two mechanical loads, high and low, which guided model development toward capturing how different levels of stress affect cellular output. Under the project, various models were developed, and several were rejected during the process. The hypothesis about how the system works was modified under the process, and additional biological states were added, which led to a more complex model. In the end, a model that could predict some of the key states well was developed. The model provides valuable insight into how the intracellular mechanisms of a hematopoietic progenitor cell react to mechanical load, but there is still a lot that can be improved or expanded upon. Further refinement of the model structure would be required to investigate whether the switch between the loads could be identified. As the model developed in this project is the first of its kind, there are endless possibilities for how it can be extended. Some of the main weaknesses of the final model are the high uncertainty in predicting certain states and its inability to predict the expression of active SERCA2, a particularly important state. Despite its limitations, the model represents a novel and promising foundation for future work, with broad potential for extension and application
The Path to Market : Development of a framework for go-to-market-strategies to guide startups in the manufacturing industry in commercializing their technology solution on a target market.
No full textStartups inom tillverkningsindustrin står inför utmaningar att omsätta sina innovationer till kommersiellt framgångsrika produkter. Ett centralt hinder är det tydliga gapet mellan strategisk planering och faktisk implementering, särskilt i samband med marknadsinträde. Trots att teorier om marknadsinträdesstrategier är väl etablerade, finns det en brist på praktiskt användbara och generaliserbara ramverk anpassade för startups i denna kontext. Med detta som utgångspunkt formulerades studiens syfte till att utveckla ett ramverk för marknadsinträdesstrategier som vägleder startups inom tillverkningsindustrin att kommersialisera sin tekniklösning på en marknad. Med vetenskaplig litteratur som grund togs en analysmodell fram. Analysmodellen innefattar tre huvudsakliga analyssteg som tillsammans syftar till att skapa förståelse för hur startups inom tillverkningsindustrin kan strukturera sitt marknadsinträde. De tre analysstegen består av undersökning av marknadsattribut, faktorer för utveckling av värdeerbjudande och intäktsmodell. Inom respektive steg analyseras faktorer som påverkar strategi för marknadsinträde. Den kvalitativa empiriska insamlingen består av elva semistrukturerade intervjuer, varav en med studiens fallföretag och tio med legotillverkare inom tillverkningsindustrin. Genom studien identifieras flera centrala marknadsattribut och faktorer som startups inom tillverkningsindustrin bör ta i beaktning för att lyckas med sitt marknadsinträde. Studien resulterade i tre slutsatser. Den första slutsatsen visar att det finns fyra centrala marknadsattribut för segmentering av marknaden inom tillverkningsindustrin, vilka varierar mellan olika aktörer och påverkar förutsättningarna för marknadsinträde. Den andra slutsatsen visar 19 centrala faktorer för att utveckla ett kundanpassat värdeerbjudande som svarar mot olika behov och utmaningar. Den tredje slutsatsen visar 17 faktorer att beakta vid utvecklingen av en intäktsmodell, där kundernas värderingar vid investering i maskinlösningar belyses.Startups in the manufacturing industry face challenges in transforming their innovations into commercially successful products. A key obstacle is the evident gap between strategic planning and actual implementation, particularly in the context of entering a market. Although theories on go-to-market-strategies are well-established, there is a lack of practically applicable and generalizable frameworks tailored to startups in this context. Based on this, the aim of the study was to develop a go-to-market-strategy framework that guides startups in the manufacturing industry in commercializing their technology solutions in a target market. Grounded in academic literature, an analytical model was developed. This model comprises three main analytical steps, which together aim to build an understanding of how manufacturing startups can structure their go-to-market-strategy. The three steps include an examination of market attributes, factors for the development of a value proposition and revenue model. Within each step, factors influencing the go-to-market-strategy are analyzed. The qualitative empirical data collection consists of eleven semi-structured interviews, one with the case company of the study and ten with contract manufacturers in the manufacturing industry. Through the study, several key market attributes and factors were identified that startups in the manufacturing industry should consider in order to succeed in their go-to-market-strategy. The study resulted in three main conclusions. The first conclusion identifies four key market attributes for segmenting the manufacturing market, which vary across different actors and influence the conditions for entering a market. The second conclusion highlights 19 critical factors for developing a customer-tailored value proposition that addresses diverse needs and challenges. The third conclusion outlines 17 factors to consider when developing a revenue model, emphasizing how customers assess value when investing in machinery solutions
LiU Home : Designing for Trust
No full textFinding secure and reliable student housing remains a challenge for many students atLinköping University. Existing online platforms such as Facebook groups and Blocket areoften unregulated, lack verification systems, and offer limited support for the specific needsof student renters and landlords. This thesis explores how trust is established in an onlinehousing rental platform, with a focus on three factors: social presence, ease of use, and se-curity. To investigate this, a prototype platform "liuHome" was developed and evaluatedthrough two user tests with a total of 51 student participants. The two versions differedin key features with respect to the login method, payment integration, user profiles, andfiltering functionality. Participants completed predefined tasks and responded to Likert-based surveys and questions about trust. The data were later analyzed using mean scoresand Mann-Whitney U tests to evaluate differences. The results indicate that the inclusionof LiU Single Sign-On (SSO) and Swish payment integration significantly improved users’perception of security. Similarly, the introduction of user profiles and review systems im-proved the social presence and trust of other users on the platform. Filtering tools andimproved designs positively influence ease of use. The findings suggest that targeted de-sign innovation that focuses on the chosen factors in trust can meaningfully enhance userconfidence in online housing marketplaces.Att hitta tryggt och pålitligt studentboende är fortfarande en utmaning för många stu-denter vid Linköpings universitet. Befintliga digitala plattformar, såsom Facebookgrupperoch Blocket, är ofta oreglerade, saknar verifieringssystem och erbjuder begränsat stöd förde särskilda behov som studenter har som hyresgäster och uthyrare. Denna uppsats un-dersöker hur förtroende etableras i en digital bostadsplattform, med fokus på tre centralafaktorer: social närvaro, användarvänlighet och säkerhet. För att undersöka detta utveck-lades en prototypplattform – "liuHome" – som utvärderades genom två användartestermed totalt 51 deltagande studenter. De två versionerna av plattformen skilde sig åt vadgäller inloggningsmetod, betalningsintegration, användarprofiler och filtreringsfunktion-alitet. Deltagarna genomförde fördefinierade uppgifter och besvarade Likert-baseradeenkäter samt frågor om förtroende. Det insamlade materialet analyserades med hjälp avmeddelvärden och Mann-Whitney U-tester för att utvärdera skillnader. Resultaten visaratt införandet av LiU:s Single Sign-On (SSO) och Swish-betalning signifikant förbättradeanvändarnas upplevda säkerhet. Även införandet av användarprofiler och recensionssys-tem ökade den sociala närvaron och tilliten till andra användare på plattformen. Fil-treringsverktyg och förbättrade designval påverkade användarvänligheten positivt. Slut-satsen är att riktade designåtgärder, med fokus på dessa tre tillitsfaktorer, på ett menings-fullt sätt kan stärka användares förtroende för digitala bostadsmarknadsplatser
6DoF Object Pose Estimation using Gaussian Splatting and NeRFs
No full text6DoF object pose estimation is important in many applications, such as robotics, autonomous driving, and virtual reality. Many methods today require expensive equipment, such as depth sensors, or high-quality hand-made 3D models, such as CAD models, to achieve high accuracy. NeRF and Gaussian Splatting are two recent methods for Novel View Synthesis, which solve these problems by creating 3D reconstructions of objects using only RGB images. This thesis explores whether the two state-of-the-art camera pose estimation methods, IFFNeRF and 6DGS, which make use of NeRF and Gaussian Splatting, can be used for 6DoF object pose estimation, and how they are affected by images containing occlusion or background. This report shows that IFFNeRF and 6DGS can be used for object pose estimation, with 6DGS generally performing better than IFFNeRF. The results also indicate that IFFNeRF and 6DGS are resistant to occlusion but highly sensitive to changes in the background
Wait, we can use AI for code reviews? : Investigating the usefulness of AI-generated code review comments
No full textRecent advancements in artificial intelligence (AI), especially large language models (LLMs), have sparked significant interest in their potential applications. Code review is a widely adopted software development practice that helps participants understand the codebase and maintain software quality. However, this process is highly time-consuming, prompting interest in improving efficiency through advances in code review tools. This study investigates the applicability and effectiveness of LLMs as assistive tools in the code review process. We generate and systematically evaluate AI-generated code review comments, which are assessed by both the study authors (using scientifically derived criteria) and professional software engineers (reflecting industrial best practices). A review of scientific literature and a series of interviews with engineers reveal substantial theoretical alignment on what constitutes a good code review comment, particularly regarding the identification of functional errors, edge cases, maintainability concerns, and suggestions for more efficient solutions. However, analysis of inter-rater reliability and thematic categorization indicates a substantial divergence of what constitutes a useful code review comment between study authors and engineers in practice. Our results show that current state-of-the-art LLMs can generate code review comments that are often perceived as useful by both academics and practitioners. Specifically, 69.4% (95% CI [65.5%, 73.1%]) of comments were found useful by the study authors, and 92.8% (95% CI [88.9%, 95.4%]) by professional software engineers. While LLMs are not positioned to replace humans in the code review process, they show promise as effective assistive tools. We highlight challenges related to prompt engineering, subjectivity in usefulness assessments, and the potential of LLMs as a component in future code reviewing tools
Buy now, save later? : When impulse meets the future in Generation Z: Buy now, Pay later from a behavioral economics perspective
No full textKonsumtion på kredit har vuxit snabbt på den svenska e-handelsmarknaden, särskilt genom tjänster som Buy Now, Pay Later (BNPL). Generation Z som präglas av teknologisk utveckling, snabb informationsåtkomst och otålighet, utgör den största användargruppen. Kombinationen av generationens egenskaper och BNPL:s lättillgänglighet kan bidra till förändrade ekonomiska beteenden, särskilt i en tid där konsumtion kan ske utan omedelbara ekonomiska konsekvenser. Mot denna bakgrund syftar uppsatsen till att undersöka hur tillgången samt attityden till BNPL-tjänster påverkar Generation Z:s förutsättningar att hantera konflikten mellan kortsiktiga konsumtionsköp och långsiktiga ekonomiska mål. För att besvara detta genomfördes tolv intervjuer med personer födda mellan år 1995 och år 2005 vilket utgör en grupp som både befinner sig i aktiv konsumtionsålder och har haft möjlighet att använda BNPL-tjänster. Studien utgår från beteendeekonomiska teorier för att förstå hur BNPL:s utformning samspelar med psykologiska mekanismer som kan underminera långsiktig ekonomisk planering. Studien visar att BNPL-tjänster på kort tid har etablerats som ett standardalternativ inom e-handeln, särskilt bland Generation Z. Tjänsternas utformning och marknadsföring sänker tröskeln för konsumtion och skapar en betalningslogik som uppmuntrar snabba beslut. Trots att många respondenter uppger att de konsumerar mer till följd av BNPL, upplevs detta inte som problematiskt och sparandet anses opåverkat. Detta tyder på att BNPL-köp ofta hålls mentalt åtskilda från övrig ekonomi, vilket skapar en kognitiv distans mellan impulsiva beslut och långsiktiga mål. Tjänsten framställs som rationell och flexibel, vilket förstärker känslan av ekonomisk kontroll, en känsla som inte alltid utgår från faktisk reflektion. Att respondenter lättare identifierar riskbeteenden hos andra än hos sig själva tyder även på en övertro på den egna förmågan, och pekar på hur BNPL inte bara påverkar konsumtion, utan även hur unga strukturerar och förstår sin ekonomiska verklighet. Slutsatsen är att BNPL skapar nya villkor för ungas ekonomiska disciplinering. Även om tjänsterna inte direkt förhindrar sparande, kräver de ett ökat mått av självdisciplin och ansvarstagande för att motverka att omedelbar konsumtion prioriteras framför långsiktig ekonomisk nytta
GNSS-Free Navigation Using Magnetic Anomaly and Topographic Maps
No full textMost modern navigation systems rely heavily on global navigation satellite systems (GNSS), which are vulnerable to jamming, spoofing and interference. To overcome these issues, alternative methods for navigation are needed. This thesis presents a GNSS-free framework for fusing inertial measurements with geophysical map data for airborne navigation. Specifically, measurements from two map modalities, terrain elevation and magnetic anomaly maps, are fused with inertial sensor measurements within a marginalized particle filter (MPF). Real-world terrain and magnetic anomaly data from surveys over Sweden are used, while sensor measurements are generated synthetically. Four geographic scenarios, ranging from mountainous to lake environments, are used to analyze how the performance varies with different map characteristics. The results show that the use of both topographic and magnetic anomaly data significantly improves estimation accuracy and reduces inertial sensor requirements compared to using the maps individually. The results also highlight the complementary nature of the two map modalities. In a simulated flight over lake Vättern in Sweden, using only the terrain map data gave poor results, as expected. But by also leveraging the magnetic anomaly map in the filter, performance was drastically improved. By fusing the two map modalities, the root mean square error (RMSE) of the estimates is reduced, but at the cost of decreased filter robustness in some scenarios. Moreover, robustness is further reduced in some scenarios with large map gradients. To address this, two robust sensor fusion techniques are evaluated: Huber’s M-estimation and P-norm opinion pooling. The results show that robustness can be greatly improved using these methods, though at the expense of estimation accuracy. Overall, the thesis concludes that fusion of inertial sensor measurements with two map modalities using an MPF framework is a promising alternative to GNSS-based navigation
Outdoor Education as a 'stimulating challenge' for pre-school teachers' motivation : the influence of Outdoor Education on pre-school teachers' well-being and motivation, through the lens of Self-Determination Theory.
No full textThis qualitative study inquires about how Outdoor Education can be used as a tool to support motivation and well-being of pre-school teachers. Eleven pre-school teachers, five from Sweden and six from Italy, with experience in Outdoor Education, were interviewed about their perspectives and experiences of teaching in the outdoors. Data was analysed using thematic analysis, through the lens of Self-Determination Theory and its three core needs: autonomy, competence and relatedness. The results of this study show an overall positive contribution of Outdoor Education on teachers’ perceived autonomy, competence and relatedness between teachers and students, teachers and teachers, students and students, and a closer relationship with nature. An interconnection between the well-being of teachers and the well-being of students also emerged, suggesting the need for future investigations about this topic. However, further research on Outdoor Education and its importance for teachers should be conducted in order to fill the gap that is currently present in the field of Outdoor Education research, especially when it comes to investigate its impact on teachers
Från plast till papper och tyg? – En studie av plastpåseskattens långsiktiga konsumtionseffekter : En kvantitativ analys skatteinförandets påverkan på det svenska konsumtionsmönstret för bärkassar
No full textPrediktion av ankomsttid med användning av djupamaskininlärningsmodeller tränade på realtidsdata från bussar
No full textEstimating bus arrival times is critical both for passengers’ trip planning and foroperators’ real-time monitoring of bus network performance. Recent advances in real-timesensor technology and low cost data storage has enabled teh storage of vast amounts ofdata. Deep machine learning models offer new opportunities for leveraging this data.Whereas prior work has focused on static schedule information or route-level time series,this report combines publicly available schedule data with high-frequency sensor signalsfrom buses in northern Stockholm to train models to predict the delay to future stops.The thesis evaluates two architectures — Long Short-Term Memory (LSTM) networks andtransformers — under two training regimes: route-specific models and a single generalmodel trained on all routes. LSTMs were optimized per route, while transformers weretrained both per route and jointly across the full dataset. The results show that the LSTMstrained with and without sensor data consistently outperform the baseline predictions butthe performance varies by which route is used during training. Averaged across multipleroutes, the LSTM achieves an RMSE that is 24.5% lower than the best-performing baselinemodel when using sensor data, and 24.0% lower without sensor data. Transformer modelsstruggle to fit to the data but can still compete with the baseline predictions. When trainedon a single route, the transformer beats the baseline by 8.9%, but when generalizingfor all routes, excluding the evaluation route, the performance of the transformer dropsand is 40.2% worse compared to the baseline. The most important sensor data from thebus when predicting future delay are accelerator time proportion, speed median and brakestandard deviation. The thesis finds that incorporating sensor data when training LSTMor transformer-based models positively affects multi-horizon predictions when applied toa specific route. These methods provide a proof of concept and lay a valuable foundationfor future research aimed at improving prediction accuracy and reliabilit