Changes take place every day in business and society. Some are relatively slow like for example changes related to globalization, demographics, and the environment. Other changes related to technology can be faster and more disruptive. Some events, like the Covid-19 pandemic, are especially disruptive and may trigger rapid changes with extensive impact. Regardless of the change it is valuable for organizations to be able to adapt. The public sector is now challenged by increasing cost pressures, an ageing population, new demands from various groups as well as new workforce requirements and this creates a need for transformation. Previous research with regards to transformation capability focused on concepts like dynamic capabilities, organizational ambidexterity, organizational change capacity, organizational change competence, or organizational agility but there seem to be a lack of studies of the concept specifically related to the public sector. In addition, a recent study by the Swedish Agency for Work Environment Expertise concluded that the research area is fragmented and that it lacks studies on how employees can contribute to organizational change competence. The purpose of this thesis was thus to investigate how transformation capability of public sector organizations can be improved from an employee- or actor perspective. To explore the research question a qualitative and explorative method was used. Theoretical literature related to transformation as well as transformation capability, was used to build knowledge around the study topic. Two different surveys with open-ended questions were used to collect data from the Swedish public sector organization selected for the study. The first survey was used as a probe to build some initial understanding of transformation practice in the organization. The second one focused specifically on how different actors could contribute to improve transformation capability. The results gave a good general overview of how the transformation capability can be improved from different actor perspectives related to the organization. By using the results together with the existing theoretical frameworks of transformation capability a new more people-centric model of transformation capability was proposed and used to explore ways of improvement. The conclusion was that transformation capability of public sector organizations can be improved by leveraging the performance measures, the actors, the process, the culture/mindset, and the technology that are involved in transformation work. A proactive, collaborative and cross-organizational approach to improvement of transformation capability over time should decrease the risk for negative effects from future disruptions and create a more viable and sustainable organization.

Die Zunahme technologischer Innovationen in unserem Alltag bietet uns neue Chancen und Herausforderungen - auch als Autofahrer. Wenngleich die breitere Verfügbarkeit und Funktionsvielfalt von sowohl Fahrerinformations- (Human Machine Interface, HMI) und -assistenzsystemen als auch mobilen Endgeräten der Erfüllung von Nutzerbedürfnissen dienen und bestenfalls der Fahrerablenkung entgegenwirken sollen, wird der Fahrer durch diese nicht notwendigerweise bei der Fahraufgabe unterstützt. Die Anforderungen an die Informationsverarbeitung steigen hingegen, gleichzeitig besteht die Versuchung die Aufmerksamkeit nicht-fahrrelevanten Tätigkeiten zuzuwenden. Fahrerablenkung ist seit jeher Thema der Verkehrsforschung, denn die Ablenkung von der Fahraufgabe kann schwerwiegende Konsequenzen haben. Innerhalb der Interaktion von Fahrer, Fahrzeug und Umwelt spielt das HMI zur sicheren Erfüllung der primären Fahraufgabe hinsichtlich des Stabilisierens, Manövrierens und Navigierens eine essenzielle Rolle. Eine fahrsituationsabhängige Anpassung der dargestellten Informationen kann während dieser Interaktion variierender Anforderungen eine Unterstützung für den Fahrer darstellen. Die vorliegende Dissertation hatte die Identifikation einer Gestaltungslösung für zukünftige Fahrerinformationssysteme, welche eine sichere und komfortable Nutzung während des Fahrens erlauben, zum Ziel. Hierfür wurde der nutzerzentrierte Gestaltprozess (User-Centred Design, UCD, DIN EN ISO 9241-210, 2010) verfolgt. Im ersten Schritt, den Vor-Studien, wurden der Kontext und die Nutzeranforderungen in der Interaktion mit fahrfremden Tätigkeiten in einem explorativen Ansatz untersucht. Am häufigsten gaben die Fahrer an, ihr Smartphone während der Fahrt zur Navigation und Kommunikation zu nutzen. Die Bereitschaft sich einer fahrfremden Tätigkeit zuzuwenden, war maßgeblich von der Modalität dieser Tätigkeit in Interaktion mit dem aktuellen Fahrszenario abhängig. Insbesondere der Straßentyp beeinflusste die Entscheidung. Die berichtete Bereitschaft zu fahrfremden Tätigkeiten wurde im zweiten Schritt des UCDs in einer Fahrsimulatorstudie, Studie I, untersucht. Sowohl für visuell-manuelle als auch kognitiv- auditive Nebentätigkeiten war die zuvor berichtete Nutzungsbereitschaft Prädiktor für das Fahrererleben und -verhalten. In Fahrszenarien, in denen die Bereitschaft für Nebentätigkeiten gering war, wurde eine höhere Beanspruchung wahrgenommen. Die Reaktionszeiten für die visuell-manuelle Aufgabe stiegen in Fahrszenarien geringer Bereitschaft. Für kognitiv-auditive Aufgaben wurden geringere Spurabweichungen und Geschwindigkeitsvarianzen in Fahrszenarien hoher Bereitschaft gefunden. Diese Befunde flossen in die Gestaltung erster adaptiver HMIs ein, Schritt drei des UCDs, und wurden im vierten Schritt in einem Fahrsimulator gegen ein statisches sowie ein konfigurierbares HMI getestet, Studie II. Abhängig von der Komplexität des Fahrszenarios wurde die Informationsmenge der adaptiven HMIs erhöht bzw. reduziert. Keines der getesteten Konzepte beeinträchtigte die Bedienbarkeit des Systems oder das Fahrverhalten und Blickverhalten. Die adaptiven und das statische HMI unterschieden sich nicht hinsichtlich der subjektiven Beurteilungen und der Verhaltensmetriken. Das konfigurierbare HMI war den adaptiven und dem statischem hinsichtlich der User Experience (UX), Beanspruchung, dem Bedienkomfort und der Akzeptanz überlegen und zog weniger Aufmerksamkeit auf das Mittelkonsolendisplay. Im Rahmen des UCDs wurden die adaptiven HMIs in einem iterativen dritten und vierten Schritt auf Basis der Ergebnisse der Studie II modifiziert und in einer Realfahrtstudie evaluiert, Studie III. Die adaptiven HMIs wurden gegen ein reduziertes, ein informationsreiches, und ein kontrollierbares HMI getestet. Einhergehend mit den Ergebnissen aus Studie II beeinträchtigte keines der Konzepte die Systembedienbarkeit oder das Fahr- und Blickverhalten. Alle Konzepte wurden gleichermaßen hoch in der UX und dem Bedienkomfort bewertet. Die Akzeptanz für das kontrollierbare Konzept war höher als für die adaptiven Konzepte. Das informationsreiche HMI erzeugte eine höhere subjektive Beanspruchung, insbesondere durch die wahrgenommene visuelle Belastung. Ebenso wurden längere Blickabwendungen von der Straße gefunden. Das Adaptieren der Nutzeroberfläche ohne den direkten Einfluss des Nutzers birgt das Risiko ein Gefühl der Bevormundung hervorzurufen. Im Rahmen dieser Dissertation wurde ein Verfahren zur Erfassung der Bevormundung durch Technologie mittels eines kombinierten Experten- und Novizenansatzes durch Interviews, Onlineumfragen und eines psycho- lexikalischen Ansatzes entwickelt. Die Ergebnisse des finalen Fragebogens (Technology Paternalism Questionnaire, TPaQ) zeigten eine höhere wahrgenommene Bevormundung der adaptiven Systeme im Vergleich zu den konfigurierbaren und kontrollierbaren. Kein Unterschied bestand hingegen zu den statischen Systemen, was für ein allgemein erhöhtes Bevormundungsgefühl durch vordefinierte Systeme spricht. Die Befunde dieser Dissertation liefern eine Gestaltungslösung zur Unterstützung des Fahrers in der Interaktion mit dem Fahrzeug und der Umwelt. Informationen im HMI konnten reduziert und situationsabhängig adaptiv dargestellt werden, ohne das Fahrererleben und -verhalten zu beeinträchtigen. Um dem Gefühl der Bevormundung entgegenzuwirken und die UX zu erhöhen, sollte der Inhalt des HMIs für den Nutzer konfigurierbar, bzw. kontrollierbar sein.:1 INTRODUCTION 2 THEORETICAL BACKGROUND 2.1 Driving Models and Driving Tasks 2.1.1 Information Processing 2.1.2 Driver Distraction 2.2 System Attitude Model 2.2.1 User Experience 2.2.2 Behavioural Adaptations 2.2.3 Driver Capabilities 2.2.4 Driving Task Demands 2.3 HMI Design Solutions 2.3.1 Indication of Adaptation 2.3.2 Adaptation Contexts 2.3.3 Adaptive User Interfaces 2.3.4 Challenges for Adaptive User Interfaces 3 AIM OF THE THESIS AND RESEARCH QUESTIONS 4 PRE-STUDIES: USER REQUIREMENTS AND NEEDS 4.1 Creativity Workshops 4.1.1 Background and Research Questions 4.1.2 Method 4.1.3 Results 4.1.4 Summary 4.2 Focus Group 4.2.1 Background and Research Questions 4.2.2 Method 4.2.3 Results 4.2.4 Summary 4.3 Online Survey 4.3.1 Background and Research Questions 4.3.2 Method 4.3.3 Results 4.3.4 Summary 4.4 Discussion 5 STUDY I: INTERACTION OF DRIVING SCENARIOS AND NON-DRIVING-RELATED TASKS 5.1 Background and Aim of the Study 5.1.1 Driver Distraction Effects under different NDRT Modalities 5.1.2 Driving Scenarios 5.2 Method 5.2.1 Independent Variables 5.2.2 Dependent Variables 5.2.3 Study Procedure 5.3 Results 5.3.1 Manipulation Check 5.3.2 Subjective Perceptions 5.3.3 Behavioural Adaptations 5.4 Discussion 6 TECHNOLOGY PATERNALISM 6.1 Theoretical Background 6.1.1 Paternalism 6.1.2 Previous Research 6.1.3 Research Questions 6.2 Methodological Approach 6.3 Expert Approach 6.3.1 Method 6.3.2 Results 6.3.3 Summary 6.4 Novice Approach 6.4.1 Interviews: Perception of (Technology) Paternalism 6.4.2 First Online Survey: Psycholexical Approach 6.4.3 Second Online Survey: Experiences with Patronisation by Technology 6.5 Technology Paternalism Model 6.5.1 Identified Definition 6.5.2 Model Assumptions 6.6 First Validation 6.7 Test Statistics 6.7.1 Psychometric Criteria 6.7.2 Factor Structure 6.7.3 Moderation and Mediation Analyses 6.8 Conclusion and Implications 6.9 Final Questionnaire 7 STUDY II: ADAPTIVE HUMAN MACHINE INTERFACES 7.1 Background and Aim of the Study 7.1.1 Needed Information in an Automotive HMI 7.1.2 HMI Designs for the Simulator Study 7.1.3 Adaptive, static and configurable HMI Concepts in the Simulator Study 7.2 Method 7.2.1 Independent Variables 7.2.2 Dependent Variables 7.2.3 Study Procedure 7.3 Results 7.3.1 Manipulation Check 7.3.2 Subjective Perceptions 7.3.3 Behavioural Adaptations 7.3.4 Configuration of the HMI 7.4 Discussion 8 STUDY III: ADAPTIVE HUMAN MACHINE INTERFACES IN THE FIELD 8.1 Background and Aim of the Study 8.1.1 HMI Designs for the Field Study 8.1.2 Adaptive, non-adaptive and controllable HMI Concepts in the Field Study 8.2 Method 8.2.1 Independent Variables 8.2.2 Dependent Variables 8.2.3 Study Procedure 8.3 Results 8.3.1 Manipulation Check 8.3.2 Subjective Perceptions 8.3.3 Behavioural Adaptations 8.3.4 Controllable HMI 8.4 Discussion 9 GENERAL DISCUSSION 9.1 Effects of the Driving Scenario on NDRT Engagement Willingness and Driver Behaviour 9.2 Impacts of Adaptive HMI Concepts 9.3 Evaluation of Configurable and Controllable HMI Concepts 9.4 Patronisation through Technology 9.5 Limitations 9.6 Future Research and Theoretical Implications 9.7 Practical Implications 10 CONCLUSION BIBLIOGRAPHY APPENDIX CURRICULUM VITAE PUBLICATIONS LIST / As technology increases throughout our daily lives we are faced with new chances and challenges, also as car drivers. Extended availability and functionalities of in-vehicle information systems (IVIS) and driver assistance systems (ADAS) in addition to mobile device features tempt the driver into distraction from the driving task. Although the intent of the integration of new technologies is the fulfilment of drivers’ needs, it does not necessarily truly support the driver or counteract driver distraction. Driver distraction is an everlasting topic in driving research as failures in managing the driving task can have severe effects. Within the interaction of the driver, the vehicle and the environment, in-vehicle information systems are crucial in guaranteeing a fulfilment of the primary driving tasks of navigating, stabilizing and manoeuvring. To support the driver in this interaction of varying demands a contextual, driving scenario-based adaptation of the content in the information system poses an opportunity. The present doctoral thesis aimed at identifying a potential design solution for an IVIS to provide a safe and comfortable usage while maintaining safe driving behaviour. Therefore, an iterative user-centred design approach (UCD, DIN EN ISO 9241-210, 2010) was pursued. As a first step, the context and user requirements in interacting with non-driving-related tasks (NDRTs) were identified in the Pre-Studies following an explorative approach. Most prevalently, drivers engaged in using their smartphones for navigation and communication. The willingness to engage in an NDRT was reported to be dependent on the modality of the task in interaction with the driving scenario, where the road type had the highest impact. As the second step, the reported engagement willingness was evaluated in a driving simulator, Study I. For both visual-manual and cognitive-auditory NDRTs the reported engagement willingness predicted the perceived workload. A higher workload was perceived in driving scenarios of a lower reported engagement willingness for NDRTs. This finding was supported by the driving behaviour. For the visual-manual task, behavioural data showed an increase in reaction times in the scenarios where engagement willingness was reported to be lower. For cognitive-auditory tasks lower lane departures and speed variances were observed in the driving scenarios with a higher reported engagement willingness. These results fed into the third and fourth steps of the user-centred design process, Study II, in which first designs of an adaptive Human Machine Interface (HMI) were realised and tested in a driving simulator against a static and a configurable HMI. Based on the complexity of the driving scenario the information content in the HMI was reduced or increased. None of the HMI concepts adversely affected system usability, driving performance, or eye glance behaviour. The adaptive and the static concepts did not differ in subjective perceptions and behavioural adaptations. The configurable HMI was superior to the adaptive and static HMI concepts in user experience (UX), perceived usability, subjective workload, as well as acceptance, and attracted less glances to the centre stack display (CSD). Iteratively, the results of Study II were integrated in the adaptive HMI concepts and then tested in an on-road study, Study III. Two adaptive HMI concepts of higher adaptation degrees, a reduced, a loaded and a controllable HMI concept were evaluated regarding subjective perceptions and behavioural adaptations. Supporting the findings of Study II, the HMI concepts did not adversely affect system usability and driver behaviour. All concepts were equally high in UX and perceived usability. The controllable HMI showed some advances, as acceptance was higher than for the adaptive concepts. The loaded HMI provoked a higher workload, especially due to the visual load, and longer glances away from the road scene. Changing the user interface (UI) without the users’ direct control poses responsibility on the system designer: Users can feel patronised by the technical system. A method to assess Technology Paternalism was developed applying a combined expert and novice approach, including interviews, online surveys and a psycholexical approach. The final questionnaire, the Technology Paternalism Questionnaire (TPaQ), showed good reliability and validity. The results of the TPaQ indicated that the adaptive HMI concepts were perceived more patronising than the configurable and controllable HMI concepts. Though, the adaptive concepts did not differ to the static concepts, speaking for a general feeling of patronisation through pre-defined concepts. Solely the control over the visualisation reduced the perceived patronisation by a technical system. The findings within this doctoral thesis provide a design solution to face the issue of driver distraction and support the interaction of the driver, vehicle and environment. Information content in the HMI could be reduced and contextually adapted, as no adverse effects on drivers’ subjective perceptions and behavioural adaptations were found. In order to counteract a feeling of patronisation and increase the UX, the information content in the HMI should be configurable or controllable.:1 INTRODUCTION 2 THEORETICAL BACKGROUND 2.1 Driving Models and Driving Tasks 2.1.1 Information Processing 2.1.2 Driver Distraction 2.2 System Attitude Model 2.2.1 User Experience 2.2.2 Behavioural Adaptations 2.2.3 Driver Capabilities 2.2.4 Driving Task Demands 2.3 HMI Design Solutions 2.3.1 Indication of Adaptation 2.3.2 Adaptation Contexts 2.3.3 Adaptive User Interfaces 2.3.4 Challenges for Adaptive User Interfaces 3 AIM OF THE THESIS AND RESEARCH QUESTIONS 4 PRE-STUDIES: USER REQUIREMENTS AND NEEDS 4.1 Creativity Workshops 4.1.1 Background and Research Questions 4.1.2 Method 4.1.3 Results 4.1.4 Summary 4.2 Focus Group 4.2.1 Background and Research Questions 4.2.2 Method 4.2.3 Results 4.2.4 Summary 4.3 Online Survey 4.3.1 Background and Research Questions 4.3.2 Method 4.3.3 Results 4.3.4 Summary 4.4 Discussion 5 STUDY I: INTERACTION OF DRIVING SCENARIOS AND NON-DRIVING-RELATED TASKS 5.1 Background and Aim of the Study 5.1.1 Driver Distraction Effects under different NDRT Modalities 5.1.2 Driving Scenarios 5.2 Method 5.2.1 Independent Variables 5.2.2 Dependent Variables 5.2.3 Study Procedure 5.3 Results 5.3.1 Manipulation Check 5.3.2 Subjective Perceptions 5.3.3 Behavioural Adaptations 5.4 Discussion 6 TECHNOLOGY PATERNALISM 6.1 Theoretical Background 6.1.1 Paternalism 6.1.2 Previous Research 6.1.3 Research Questions 6.2 Methodological Approach 6.3 Expert Approach 6.3.1 Method 6.3.2 Results 6.3.3 Summary 6.4 Novice Approach 6.4.1 Interviews: Perception of (Technology) Paternalism 6.4.2 First Online Survey: Psycholexical Approach 6.4.3 Second Online Survey: Experiences with Patronisation by Technology 6.5 Technology Paternalism Model 6.5.1 Identified Definition 6.5.2 Model Assumptions 6.6 First Validation 6.7 Test Statistics 6.7.1 Psychometric Criteria 6.7.2 Factor Structure 6.7.3 Moderation and Mediation Analyses 6.8 Conclusion and Implications 6.9 Final Questionnaire 7 STUDY II: ADAPTIVE HUMAN MACHINE INTERFACES 7.1 Background and Aim of the Study 7.1.1 Needed Information in an Automotive HMI 7.1.2 HMI Designs for the Simulator Study 7.1.3 Adaptive, static and configurable HMI Concepts in the Simulator Study 7.2 Method 7.2.1 Independent Variables 7.2.2 Dependent Variables 7.2.3 Study Procedure 7.3 Results 7.3.1 Manipulation Check 7.3.2 Subjective Perceptions 7.3.3 Behavioural Adaptations 7.3.4 Configuration of the HMI 7.4 Discussion 8 STUDY III: ADAPTIVE HUMAN MACHINE INTERFACES IN THE FIELD 8.1 Background and Aim of the Study 8.1.1 HMI Designs for the Field Study 8.1.2 Adaptive, non-adaptive and controllable HMI Concepts in the Field Study 8.2 Method 8.2.1 Independent Variables 8.2.2 Dependent Variables 8.2.3 Study Procedure 8.3 Results 8.3.1 Manipulation Check 8.3.2 Subjective Perceptions 8.3.3 Behavioural Adaptations 8.3.4 Controllable HMI 8.4 Discussion 9 GENERAL DISCUSSION 9.1 Effects of the Driving Scenario on NDRT Engagement Willingness and Driver Behaviour 9.2 Impacts of Adaptive HMI Concepts 9.3 Evaluation of Configurable and Controllable HMI Concepts 9.4 Patronisation through Technology 9.5 Limitations 9.6 Future Research and Theoretical Implications 9.7 Practical Implications 10 CONCLUSION BIBLIOGRAPHY APPENDIX CURRICULUM VITAE PUBLICATIONS LIST

Die Zugriffslokalität referenzierter Daten ist eine wichtige Eigenschaft verteilter Anwendungen. Lokales Zwischenspeichern abgefragter entfernter Daten (Caching) wird vielfach bei der Entwicklung solcher Anwendungen eingesetzt, um diese Eigenschaft auszunutzen. Anschliessende Zugriffe auf diese Daten können so beschleunigt werden, indem sie aus dem lokalen Zwischenspeicher bedient werden. Gegenwärtige Middleware-Architekturen bieten dem Anwendungsprogrammierer jedoch kaum Unterstützung für diesen nicht-funktionalen Aspekt. Die vorliegende Arbeit versucht deshalb, Caching als separaten, konfigurierbaren Middleware-Dienst auszulagern. Durch die Einbindung in den Softwareentwicklungsprozess wird die frühzeitige Modellierung und spätere Wiederverwendung caching-spezifischer Metadaten gewährleistet. Zur Laufzeit kann sich das entwickelte System außerdem bezüglich der Cachebarkeit von Daten adaptiv an geändertes Nutzungsverhalten anpassen. / Locality of reference is an important property of distributed applications. Caching is typically employed during the development of such applications to exploit this property by locally storing queried data: Subsequent accesses can be accelerated by serving their results immediately form the local store. Current middleware architectures however hardly support this non-functional aspect. The thesis at hand thus tries outsource caching as a separate, configurable middleware service. Integration into the software development lifecycle provides for early capturing, modeling, and later reuse of cachingrelated metadata. At runtime, the implemented system can adapt to caching access characteristics with respect to data cacheability properties, thus healing misconfigurations and optimizing itself to an appropriate configuration. Speculative prefetching of data probably queried in the immediate future complements the presented approach.

Die komponentenbasierten Softwareentwicklung verspricht die vereinfachte Entwicklung von komplexen Anwendungen. Um die Wiederverwendbarkeit zu verbessern und die Flexibilität zu erhöhen, müssen Komponenten dazu möglichst an verschiedene Umgebungsbedingungen angepasst werden können, sowohl innerhalb einer als auch in unterschiedlichen Anwendungen. Diese Prozesse werden als Komponentenadaption bezeichnet. In dieser Arbeit wird ein neues Adaptionskonzept für Komponenten entwickelt. Die sogenannten Adaptierbare Komponenten verwenden ein hierarchisches Komponentenmodell und werden aus einer Menge von Subkomponenten zusammengesetzt. Die Kernidee zur Umsetzung der Adaptivität besteht darin, bestimmte Parameterwerte einer Adaptierbaren Komponente auf unterschiedliche interne Konfigurationen der Subkomponenten abzubilden. Zur Beschreibung der möglichen internen Konfigurationen von Adaptierbaren Komponenten werden vier verschiedene graphische Modellierungstechniken entwickelt, die alle auf der graphischen Notation von UML-Komponentendiagrammen basieren und diese erweitern. Eine sogenannte Parameterabbildung definiert die Zuordnung von Parameterwerten auf bestimmte Konfigurationen. Die Konzepte Adaptierbarer Komponenten setzen keine neue Komponentenplattform voraus, sondern werden durch eine Kombination von Modelltransformation und spezieller Laufzeitunterstützung auf existierende Komponentenplattformen abgebildet. Ein dazu entwickeltes generisches Verfahren definiert die Schritte zur Unterstützung einer neuen Komponentenplattform. Mit Hilfe von zwei Fallstudien wird gezeigt, dass sich die Modellierungskonzepte von Adaptierbaren Komponenten für komplexe Beispiele anwenden lassen. Mit EJB, JavaBeans und Microsoft COM wird die Modelltransformation und Laufzeitunterstützung anhand des generischen Verfahrens exemplarisch für populäre Komponentenplattformen demonstriert.

No description available.

The thesis has aimed to study the emergence of a new weapons technology, also known as ‘killer robots’ or lethal autonomous weapon system. It seeks to answer what factors drive the development and deployment of this weapon system without ‘meaningful human control’, a component that allows the decision to kill to be delegated to machines. The research question focuses on seeking the motivations to develop and deploy LAWS, as well as the consequences this would have on military conduct and conflict characteristics.The incentives they bring up and the way of adopting them has been studied by synthesizing antinomic democratic peace theory and adoption capacity theory respectively. The findings of this qualitative content analysis lead to two major conclusions. (1) That LAWS present severe risk avoidance and costs reduction potential for the user. These factors have a more prevalent pull on democracies than autocracies, since they stand to benefit from LAWS’ specific capabilities more in comparison. (2) That their adoption is aided by low financial intensity needed to adopt it, due to the high commercial profitability and applicability of AI technology, and the ease of a spillover to military sphere. Their adoption is hindered by high organizational capital needed to implement the drastic changes LAWS bring. All of this leads to the prediction that LAWS are likely to proliferate further, at a medium speed, and potentially upset the balance of power.

Edge cloud is a distributed computing architecture that is growing in popularity. It aims to bring the cloud closer to the edge of a network, reducing latency and improving performance through the use of distributed servers (edge nodes) spread out geographically. However, in the case of sudden increases in user requests, a node may run short of resources and need to implement a strategy that allows the node's service to degrade its service quality to a level that requires fewer resources so that the service can still be delivered. One such strategy is brownout, a control theory-based algorithm that dynamically controls the node's service quality in order to meet e.g., a latency goal. This thesis explores the use of brownout, previously used in combination with load balancing in the cloud, in conjunction with load balancing in an edge-cloud environment. In this thesis, four load-balancing strategies are evaluated in a Kubernetes-based edge-cloud environment, along with an application that implements the brownout feature. Two of the strategies are originally designed to be used with brownout but made for the regular cloud, one is a recently introduced strategy that performs well in the edge cloud but is brownout unaware, and the last is a random load balancer used as a baseline (also brownout unaware). The goal of the evaluation is to determine the efficiency of these strategies in different edge-cloud scenarios, with regard to service quality-weighted throughput, average latency, adherence to a set latency goal, and outsourcing (requests load balanced to another edge node). The results show that the first two strategies perform worse than the random load balancer in many regards. Their performance is also less predictable and tends to get worse with increasing network delays. The edge cloud strategy, however, shows an improvement in performance when the brownout is introduced in the majority of the test scenarios.  Furthermore, the thesis introduced three possible modifications to make one of the cloud-based strategies perform better in the edge cloud. These modifications were tested in the same environment as the other load-balancing strategies and compared against each other. The first modification consisted of making the load-balancing logic treat its own node differently from other edge nodes.  The second version was devised to only outsource when a certain resource threshold is exceeded and the last version was designed to prioritize its own node when below a certain resource threshold. The last version improved on the others and performed better than the base version in all measured metrics. Compared to the edge cloud strategy with brownout, it performed better with regard to service quality-weighted throughput but was outperformed in all other metrics.

<p>  </p> <p>Evolution is the foundation for all of biology. However, our approaches and understanding of evolution—simply, the change of allele frequencies from one generation to the next—have themselves evolved over time. In this dissertation I explore multiple approaches to understand evolution and the consequences of evolution across variable scales and study organisms. First, I use meta-analytic techniques and Bayesian hierarchical models to investigate the phenotypic consequences of two forms of cryptic local adaptation, co- and countergradient variation, by leveraging a decades-old quantitative genetics approach (Chapter 1). I find large effects for both co- and countergradient variation, however they are obscured in natural settings by concurrent large environmental effects. I also show that these large effects are ubiquitous across phenotypic traits, organisms, and environmental gradients, suggesting that while similar phenotypes may be the evolutionary end point, the mechanisms to achieve those phenotypes likely vary. In the following chapter I explore the rapid evolution of a unique and understudied species introduction, pink salmon (<em>Oncorhynchus gorbuscha</em>) in the Great Lakes. Pink salmon were introduced into Lake Superior in a single introduction event and have broken two obligate life histories, anadromy (though they treat the Great Lakes like surrogate oceans) and their fixed two-year life cycle, making them ripe subjects for contemporary evolution. Using whole-genome sequence data, I first investigate the effects of a genetic drift in the form of a bottleneck at introduction and characterize the subsequent loss of genetic diversity (Chapter 2). I show that despite a large loss of genetic diversity, pink salmon also rapidly adapted to their novel environment based on signals of putative selection across numerous regions of the genome, particularly in a period gene associated with their daily circadian clock (<em>per2</em>). Next, I explore how genome structure likely aided adaptation by pink salmon to the Great Lakes, providing evidence that a supergene (~29 Mbp) containing an inversion on chromosome 10 swept to near fixation in the Great Lakes (Chapter 3) and likely aided in osmoregulatory adaptation to this novel environment. Finally, I end with a short perspective chapter (Chapter 4) where I highlight potential future research directions for each of the previous chapters. Together, this research investigates the drivers and consequences of evolution across multiple scales and shows the powerful effect of genetic drift and genetic adaptation in shaping the genomic and phenotypic attributes of populations.</p>

This study analyzes the semiotic resources in two versions of Vårdguiden's web text about tick bites: one longer version and one in easy Swedish. The aim is to identify and compare the prior knowledge and information need of each versions' model reader. This study is theoretically grounded in social semiotics and health literacy. The method consists of three multimodal text analyzes with ideational, interpersonal and textual sub-analyses.  Both versions presupposing several actual, practical and mental prior knowledge of the model reader. The information need of the versions' model reader is both common and separate. Both versions also presupposing that the model reader needs informing, explanatory, instructive and reassuring messages with a multimodal and multi-sequential presentation. The longer version presupposes more comprehensive prior knowledge and larger information needs. The version in easy Swedish though presupposes some prior knowledge which is not assumed in the longer version. For example, it is prior knowledge of how a tick bite looks like and the prior knowledges which is required when photographic close-ups should be interpreted. The version in easy Swedish also and presupposes larger needs for explanations and instructions. The results hinting what the versions presuppose of the empirical reader and prompting a discussion about how the versions high demands for health literacy can impede the reader friendliness. However, the results show that the versions' design creates possibilities for broad groups to assimilate the content. The study highlights some important challenges which are relevant when texts should be adapted to broad groups. / I studien analyseras två versioner av 1177 Vårdguidens text om fästingbett: en längre version och en version på lätt svenska. Syftet är att identifiera och jämföra de förkunskaper och behov av information som förutsätts hos versionernas modelläsare. Studien är teoretiskt grundad i sociosemiotik och hälsolitteracitet. Metoden består av tre multimodala textanalyser med ideationella, interpersonella och textuella delanalyser. Båda versionerna förutsätter flera faktiska, praktiska och mentala förkunskaper hos modelläsaren. Behoven av information hos versionernas modelläsare är dels gemensamma, dels skilda. Båda versionerna förutsätter att modelläsaren behöver informerande, förklarande, instruerande och lugnande budskap som presenteras multimodalt och multisekventiellt. Den längre versionen förutsätter mer omfattande förkunskaper och större behov av information. Versionen på lätt svenska förutsätter dock några förkunskaper som inte förutsätts i den längre versionen. Det är till exempel förkunskaper om hur ett fästingbett ser ut och de förkunskaper som krävs för att kunna tolka fotografiska närbilder. Versionen på lätt svenska förutsätter även att modelläsaren behöver fler förklaringar och instruerande budskap. Resultaten antyder vad versionerna förutsätter hos de empiriska mottagarna och föranleder en diskussion om hur versionernas höga krav på hälsolitteracitet kan hindra brukarvänligheten. Resultaten visar dock att versionernas utformning ger möjligheter för breda grupper att tillgodogöra sig innehållet. Studien belyser några viktiga utmaningar som är relevanta när texter ska mottagaranpassas till breda grupper. / <p>Uppsatsen ingår i kursen språkrådgivning och textvård magisterkurs, 30 hp.</p>

I dag anses molntjänster vara ett omtalat ämne som har ändrat hur IT-tjänster levereras och som skapat nya affärsmodeller. Några av molntjänsternas mest frekvent nämnda fördelar är flexibilitet och skalbarhet. Molntjänster är i dagsläget extensivt använda av privatpersoner genom tjänster så som Google Drive och Dropbox. Å andra sidan kan en viss försiktighet gentemot molntjänster uppmärksammas hos de organisationer som innehar känslig data. Denna försiktighet kan anses leda till en långsammare tillämpningshastighet för dessa organisationer. Detta examensarbete har som syfte att undersöka sambandet mellan molntjänster och känslig data för att kunna erbjuda stöd och kunskapsbas för organisationer som överväger en övergång till molntjänster. Känslig data är definierat som information som omfattas av den svenska Personuppgiftslagen. Tidigare studier visar att organisationer värdesätter en hög säkerhetsgrad vid en övergång till molntjänster och ofta föredrar att leverantören kan erbjuda ett antal säkerhetsmekanismer. En molntjänsts lagliga överensstämmelse är en annan faktor som uppmärksammas. Datainsamlingen skedde genom en enkät, som var riktad till 101 av de svenska organisationerna i syfte att kartlägga användningen av molntjänster samt att identifiera möjliga bromsande faktorer. Dessutom genomfördes tre (3) intervjuer med experter och forskare inom IT-lag och/eller molnlösningar. En analys och diskussion, baserad på resultaten, har genomförts, vilket ledde till slutsatserna att en molnlösning av hybrid karaktär är bäst lämpad för den försiktiga organisationen, de olika villkoren i serviceavtalet bör grundligt diskuteras innan en överenskommelse mellan parter uppnås samt att i syfte att undvika att lösningen blir oförenlig med lagen bör främst en leverantör som är väl etablerad i Sverige väljas. Slutligen, bör varje organisation utvärdera om molntjänster kan tillgodose organisationens säkerhetsbehov, då det i stor mån berör ett risktagande. / Cloud computing is today a hot topic, which has changed how IT is delivered and created new business models to pursue. The main listed benefits of Cloud computing are, among others, flexibility and scalability. It is widely adopted by individuals in services, such as Google Drive and Dropbox. However, there exist a certain degree of precaution towards Cloud computing at organizations, which possess sensitive data, which may decelerate the adoption. Hence, this master thesis aims to investigate the topic of Cloud computing in a combination with sensitive data in order to support organizations in their decision making with a base of knowledge when a transition into the Cloud is considered. Sensitive data is defined as information protected by the Swedish Personal Data Act. Previous studies show that organizations value high degree of security when making a transition into Cloud computing, and request several measures to be implemented by the Cloud computing service provider. Legislative conformation of a Cloud computing service is another important aspect. The data gathering activities consisted of a survey, directed towards 101 Swedish organizations in order to map their usage of Cloud computing services and to identify aspects, which may decelerate the adoption. Moreover, interviews with three (3) experts within the fields of law and Cloud computing were conducted. The results were analyzed and discussed, which led to conclusions that hybrid Cloud is a well chosen alternative for a precautious organization, the SLA between the organizations should be thoroughly negotiated and that primarily providers well established on the Swedish market should be chosen in order to minimize the risk of legally non-consisting solution. Finally, each organization should decide whether the security provided by the Cloud computing provider is sufficient for organization’s purposes.