The rapid growth of centralised social media platforms such as Twitter and Facebook has raised concerns about privacy and the concentration of power held by the companies behind these services. In response, decentralised social media platforms have emerged as alternatives, forming federations of various instances without a single owner. However, these decentralised platforms have faced challenges with centralisation, as users tend to gather on larger instances. This paper examines Mastodon, one of the most popular decentralised platforms, to explore whether its onboarding flow and website design can be modified to encourage less centralisation. The study maps Mastodon's onboarding flow and identifies potential modifications to mitigate centralisation. Additionally, a new design for the official website is proposed. User tests involving five participants were conducted to evaluate the effectiveness of these changes. The results indicate that modifying the design can influence users to choose smaller servers. One effective modification is randomising the order in which servers are presented, eliminating the advantage of larger servers. Another proposed change is to improve user awareness about how feeds and interactions on Mastodon spans servers, enabling users to interact beyond their own server and removing a motivation for choosing a larger one. But the impact of this information on user behaviour remains inconclusive in the user tests. Overall, this research suggests that through design modifications, Mastodon can encourage users to select smaller servers and foster a more decentralised social media environment. / Den snabba tillväxten av centraliserade sociala medieplattformar som Twitter och Facebook har väckt oro över personlig integritet och koncentrationen av makt hos företagen bakom dessa tjänster. Som svar har decentraliserade sociala medieplattformar dykt upp, där plattformen består av en federation av olika instanser utan en enda ägare. Dessa decentraliserade plattformar har dock stött på utmaningar med centralisering eftersom användare tenderar att samlas på större instanser. Den här rapporten undersöker Mastodon, en av de mest populära decentraliserade plattformarna, för att undersöka om dess onboardingflöde och designen av deras webbplats kan modifieras för att uppmuntra mindre centralisering. Studien kartlägger Mastodons nuvarande onboardingflöde och identifierar potentiella ändringar som kan mildra centraliseringen. Dessutom föreslås en ny design för den officiella webbplatsen. Användartester med fem deltagare genomfördes för att utvärdera effektiviteten av dessa förändringar. Resultaten indikerar att den modifierade designen kan påverka användare och få dem att välja mindre servrar. En effektiv förändring är att randomisera ordningen i vilken servrar presenteras, vilket eliminerar fördelen större servrar har i hur synliga de är. En annan föreslagen förändring är att förbättra användarnas medvetenhet om att flöden och interaktioner på Mastodon sträcker sig över servrar, vilket gör det möjligt för användare att interagera utanför sin egen server och tar bort ett av motiven användare har för att välja en större server. Vilken effekt det här har på användarnas beteende när det gäller deras val av server var dock oklart i användartesterna som utfördes. Sammantaget tyder denna undersökning på att Mastodon genom designändringar potentiellt kan uppmuntra användare att välja mindre servrar och främja en mer decentraliserad miljö för sociala media.

Water monitoring is an essential process that managesthe well-being of freshwater ecosystems. However, it isgenerally an inefficient process as most data collection is donemanually. By combining wireless sensor technology and machinelearning techniques, projects such as iWater aim to modernizecurrent methods. The purpose of the iWater project is to developa network of smart sensors capable of collecting and analyzingwater quality-related data in real time.To contribute to this goal, a comparative study between theperformance of a centralized machine learning algorithm thatis currently used, and a distributed model based on a federatedlearning algorithm was done. The data used for training andtesting both models was collected by a wireless sensor developedby the iWater project. The centralized algorithm was used asthe basis for the developed distributed model. Due to lack ofsensors, the distributed model was simulated by down-samplingand dividing the sensor data into six data sets representing anindividual sensor. The results are similar for both models andthe developed algorithm reaches an accuracy of 98.41 %. / Vattenövervakning är en nödvändig processför att få inblick i sötvattensekosystems välmående. Dessvärreär det en kostsam och tidskrävande process då insamling avdata vanligen görs manuellt. Genom att kombinera trådlössensorteknologi och maskininlärnings algoritmer strävar projektsom iWater mot att modernisera befintliga metoder.Syftet med iWater är att skapa ett nätverk av smarta sensorersom kan samla in och analysera vattenkvalitetsrelaterade datai realtid. För att bidra till projektmålet görs en jämförandestudie mellan den prediktiva noggrannheten hos en centraliseradmaskininlärningsalgoritm, som i nuläget används, och endistribuerad modell baserad på federerat lärande. Data somanvänds för träning och testning av båda modellerna samladesin genom en trådlös sensor utvecklad inom iWater-projektet.Den centraliserade algoritmen användes som grund för denutvecklade distribuerade modellen. På grund av brist på sensorersimulerades den distribuerade modellen genom nedprovtagningoch uppdelning av data i sex datamängder som representerarenskilda sensorer. Resultaten för båda modellerna var liknandeoch den utvecklade algoritmen har en noggrannhet på 98.41 % / Kandidatexjobb i elektroteknik 2021, KTH, Stockholm

Algorithmically designed reward functions can influence groups of learning agents toward measurable desired sequential joint behaviours. Influencing learning agents toward desirable behaviours is non-trivial due to the difficulties of assigning credit for global success to the deserving agents and of inducing coordination. Quantifying joint behaviours lets us identify global success by ranking some behaviours as more desirable than others. We propose a real-valued metric for turn-taking, demonstrating how to measure one sequential joint behaviour. We describe how to identify the presence of turn-taking in simulation results and we calculate the quantity of turn-taking that could be observed between independent random agents. We demonstrate our turn-taking metric by reinterpreting previous work on turn-taking in emergent communication and by analysing a recorded human conversation. Given a metric, we can explore the space of reward functions and identify those reward functions that result in global success in groups of learning agents. We describe 'medium access games' as a model for human and machine communication and we present simulation results for an extensive range of reward functions for pairs of Q-learning agents. We use the Nash equilibria of medium access games to develop predictors for determining which reward functions result in turn-taking. Having demonstrated the predictive power of Nash equilibria for turn-taking in medium access games, we focus on synthesis of reward functions for stochastic games that result in arbitrary desirable Nash equilibria. Our method constructs a reward function such that a particular joint behaviour is the unique Nash equilibrium of a stochastic game, provided that such a reward function exists. This method builds on techniques for designing rewards for Markov decision processes and for normal form games. We explain our reward design methods in detail and formally prove that they are correct.

The goal of this dissertation is to evaluate the effectiveness of model predictive control (MPC) for a magnetically suspended flywheel energy storage uninterruptible power supply (FlyUPS). The reason this research topic was selected was to determine if an advanced control technique such as MPC could perform better than a classical control approach such as decentralised Proportional-plus-Differential (PD) control. Based on a literature study of the FlyUPS system and the MPC strategies available, two MPC strategies were used to design two possible MPC controllers were designed for the FlyUPS, namely a classical MPC algorithm that incorporates optimisation techniques and the MPC algorithm used in the MATLAB® MPC toolbox™. In order to take the restrictions of the system into consideration, the model used to derive the controllers was reduced to an order of ten according to the Hankel singular value decomposition of the model. Simulation results indicated that the first controller based on a classical MPC algorithm and optimisation techniques was not verified as a viable control strategy to be implemented on the physical FlyUPS system due to difficulties obtaining the desired response. The second controller derived using the MATLAB® MPC toolbox™ was verified to be a viable control strategy for the FlyUPS by delivering good performance in simulation. The verified MPC controller was then implemented on the FlyUPS. This implementation was then analysed in order to validate that the controller operates as expected through a comparison of the simulation and implementation results. Further analysis was then done by comparing the performance of MPC with decentralised PD control in order to determine the advantages and limitations of using MPC on the FlyUPS. The advantages indicated by the evaluation include the simplicity of the design of the controller that follows directly from the specifications of the system and the dynamics of the system, and the good performance of the controller within the parameters of the controller design. The limitations identified during this evaluation include the high computational load that requires a relatively long execution time, and the inability of the MPC controller to adapt to unmodelled system dynamics. Based on this evaluation MPC can be seen as a viable control strategy for the FlyUPS, however more research is needed to optimise the MPC approach to yield significant advantages over other control techniques such as decentralised PD control. / Thesis (MIng (Computer and Electronic Engineering))--North-West University, Potchefstroom Campus, 2013

La coopération décentralisée, brique de l’Aide publique au développement et action extérieure des collectivités territoriales, existe légalement depuis 1992. L’intérêt pour la thématique "eau et assainissement" ainsi que pour les pays de la Méditerranée, nous a amené à mener notre recherche dans les Territoires palestiniens, territoires souffrant du manque d’eau. Ces actions de coopération nécessitent une coordination des parties prenantes sur le territoire partenaire. C’est cette coordination que questionne cette recherche. Pour cela les enjeux de la coopération décentralisée ont été comparés entre la littérature et le terrain, ce qui nous a permis de nous intéresser plus particulièrement à l’enjeu du transfert de connaissances. En effet, les actions de formation ont une part de plus en plus importante dans les projets de coopération décentralisée, du fait notamment des difficultés financières des collectivités territoriales. La littérature du transfert inter-organisationnel a aussi été mobilisée lors de cette recherche. Nous sommes ensuite revenus sur les différentes pratiques managériales liées à la gestion de projet, au suivi de projet, à la mise en place de consultations publiques, etc. Nous avons pu voir que la communication publique était peu existante et que d’autres difficultés existaient comme la gestion du turn-over. Nous avons souhaité comprendre quels pouvaient être les impacts liés à cela, ainsi que les solutions envisageables. Lorsque nous nous sommes intéressés aux difficultés, la coordination des parties prenantes est apparue comme primordiale à améliorer / Decentralized cooperation projects, an aspect of the development aid and external action of local authorities, has legally existed since 1992. Simultaneous interest in water and sanitation issues and in Mediterranean countries motivated us to focus our research in the Palestinian Territories, many of which suffer of a lack of water. These cooperative actions require a coordination of the stakeholders on the partner territory. This research questions the efficacy of these coordinated efforts applied to the water and sanitation field. Herein, issues of decentralized cooperation have been compared between the literature and a field study, which allowed us to focus more particularly on the issue of knowledge transfer. Indeed, training accounted for an increasingly important share of decentralized cooperation projects, due in particular to financial difficulties of local authorities. This is why we focused on knowledge transfer’s literature, inter-organizational transfer actually as it does not meet the same difficulties as private organizations (with holding information, power games, an unstable relationship, etc.) but conversely it is rarely assessed and thus warrants investigation. We also followed up on the different management practices related to project management : project monitoring, implementation of public consultations, etc. We found that public communication is scarce and found others difficulties such as frequent management turnovers. We wished to understand the impact of frequent turnover and solutions to the problems they may create

Physically decentralised water management systems may contribute to improving the sustainability of urban water management. Any shift toward decentralised systems needs to consider not just physical system design but also social values, knowledge frames, and organisations, and their interconnections to the physical technology. Four cases of recent Australian urban water management improvement projects were researched using qualitative methods. Three cases were of decentralised water management innovation. The other was of a centralised system, although decentralised options had been considered. These cases were studied to identify institutional barriers and enablers for the uptake of decentralised systems, and to better understand how emerging environmental engineering knowledge might be applied to overcome an implementation gap for decentralised urban water technologies. Analysis of each case focused on the institutional elements of urban water management, namely: the values, knowledge frames and organisational structures. These elements were identified through in-depth interviews, document review, and an on-line survey. The alignment of these elements was identified as being a significant contributor to the stability of centralised systems, or to change toward decentralised systems. A new organisational home for innovative knowledge was found to be common to each case where decentralised innovation occurred. ??Institutional entrepreneurs??, strong stakeholder engagement, and inter-organisational networks were all found to be linked to the creation of shared meaning and legitimacy for organisational and technological change. Existing planning frameworks focus on expert justification for change rather than institutional support for change. Institutional factors include shared understandings, values and organisational frameworks, and the alignment of each factor. Principles for, and examples of, appropriate organisational design for enabling and managing decentralised technological innovation for urban water management are proposed. This research contributes to the understanding of the institutional basis and dynamics of urban water management, particularly in relation to physical centralisation and decentralisation of urban water management technologies and, to a lesser extent, in relation to user involvement in urban water management. Understanding of factors that contribute to enabling and constraining decentralised technologies is extended to include institutional and organisational factors. New and practical pathways for change for the implementation of decentralised urban water systems are provided.

Infectious disease can be transmitted via various environmental pathways, many of which are incorporated into our water and wastewater systems. Quantitative microbial risk assessment (QMRA) can be a valuable tool in identifying hazard exposure pathways and estimating their associated health impacts. QMRA can be applied to establish standards and guidelines and has been adopted by the World Health Organisation for the management of risks from water-related infectious diseases. This thesis aims at presenting a holistic approach for the assessment of microbial health risks in urban water and wastewater systems. The procedure of QMRA is presented, together with the data collected for the case studies, and the results are discussed in a risk management framework. Decentralised drinking water treatment with membranes was shown to be competitive with centralised conventional treatment regarding environmental impacts and health. To attain sufficient die-off of pathogens in order to reduce risks to acceptable levels, facilities that permit the long-term storage of locally collected faeces are required. Issues of operation and mangement are likely to determine the health risks in decentralised systems. While failures in distribution are more likely to result in detectable waterborne disease outbreaks, the number of people at risk of becoming infected with pathogens passing normal treatment, calculated on a yearly basis, can be larger. Site-specific pathogen monitoring of source waters was identified as an important factor for the accurate estimation of risk. Noroviruses, an emerging waterborne pathogen, were shown to have fluctuating concentrations in surface water, with significant peaks during the wintertime. Time series analysis has potential as an early warning system if complemented by regular monitoring to discriminate peaks from random fluctuations. Groups already sensitive to infection, i.e. the elderly, the sick and children, were shown to consume higher volumes of cold tap water than the rest of the population, which may call for special atention in the risk management of drinking water systems. Microbial health risks associated with the handling and reuse of wastewater and sludge were shown to be successfully addressed within the management system Hazard Analysis and Critical Control Points (HACCP). Most exposure points identified could be controlled through easy measures. / Copyright Agreement: Figure 6-1, page 49 and figure 6-2, page 50 in the summary/introduction are reprinted from Water Science and Technology: Water Supply 2(2) 11-18, with permission from the copyright holders, IWA. Note: the median values are missing in the article but the figures have been corrected in the summary/introduction.

The goal of this dissertation is to evaluate the effectiveness of model predictive control (MPC) for a magnetically suspended flywheel energy storage uninterruptible power supply (FlyUPS). The reason this research topic was selected was to determine if an advanced control technique such as MPC could perform better than a classical control approach such as decentralised Proportional-plus-Differential (PD) control. Based on a literature study of the FlyUPS system and the MPC strategies available, two MPC strategies were used to design two possible MPC controllers were designed for the FlyUPS, namely a classical MPC algorithm that incorporates optimisation techniques and the MPC algorithm used in the MATLAB® MPC toolbox™. In order to take the restrictions of the system into consideration, the model used to derive the controllers was reduced to an order of ten according to the Hankel singular value decomposition of the model. Simulation results indicated that the first controller based on a classical MPC algorithm and optimisation techniques was not verified as a viable control strategy to be implemented on the physical FlyUPS system due to difficulties obtaining the desired response. The second controller derived using the MATLAB® MPC toolbox™ was verified to be a viable control strategy for the FlyUPS by delivering good performance in simulation. The verified MPC controller was then implemented on the FlyUPS. This implementation was then analysed in order to validate that the controller operates as expected through a comparison of the simulation and implementation results. Further analysis was then done by comparing the performance of MPC with decentralised PD control in order to determine the advantages and limitations of using MPC on the FlyUPS. The advantages indicated by the evaluation include the simplicity of the design of the controller that follows directly from the specifications of the system and the dynamics of the system, and the good performance of the controller within the parameters of the controller design. The limitations identified during this evaluation include the high computational load that requires a relatively long execution time, and the inability of the MPC controller to adapt to unmodelled system dynamics. Based on this evaluation MPC can be seen as a viable control strategy for the FlyUPS, however more research is needed to optimise the MPC approach to yield significant advantages over other control techniques such as decentralised PD control. / Thesis (MIng (Computer and Electronic Engineering))--North-West University, Potchefstroom Campus, 2013

Physically decentralised water management systems may contribute to improving the sustainability of urban water management. Any shift toward decentralised systems needs to consider not just physical system design but also social values, knowledge frames, and organisations, and their interconnections to the physical technology. Four cases of recent Australian urban water management improvement projects were researched using qualitative methods. Three cases were of decentralised water management innovation. The other was of a centralised system, although decentralised options had been considered. These cases were studied to identify institutional barriers and enablers for the uptake of decentralised systems, and to better understand how emerging environmental engineering knowledge might be applied to overcome an implementation gap for decentralised urban water technologies. Analysis of each case focused on the institutional elements of urban water management, namely: the values, knowledge frames and organisational structures. These elements were identified through in-depth interviews, document review, and an on-line survey. The alignment of these elements was identified as being a significant contributor to the stability of centralised systems, or to change toward decentralised systems. A new organisational home for innovative knowledge was found to be common to each case where decentralised innovation occurred. ??Institutional entrepreneurs??, strong stakeholder engagement, and inter-organisational networks were all found to be linked to the creation of shared meaning and legitimacy for organisational and technological change. Existing planning frameworks focus on expert justification for change rather than institutional support for change. Institutional factors include shared understandings, values and organisational frameworks, and the alignment of each factor. Principles for, and examples of, appropriate organisational design for enabling and managing decentralised technological innovation for urban water management are proposed. This research contributes to the understanding of the institutional basis and dynamics of urban water management, particularly in relation to physical centralisation and decentralisation of urban water management technologies and, to a lesser extent, in relation to user involvement in urban water management. Understanding of factors that contribute to enabling and constraining decentralised technologies is extended to include institutional and organisational factors. New and practical pathways for change for the implementation of decentralised urban water systems are provided.

Master of Public Health - MPH / Introduction According (UNAIDS/WHO, 2011a) globally, 34.0 million [31.4 million–35.9 million] people were living with HIV at the end of 2011 out of which 80% live in Sub Saharan Africa. In absolute terms, the burden of HIV increased by 354% between 1990 and 2006 (Ortblad et al, 2010). The prevalence of HIV in Zambia stands at 14.3% in the age group 15 to 49 years. Having realized that HIV/AIDS is no longer a health issue alone Zambia embarked on the multisectoral response initiative which allows many stakeholders to play their role. In order to coordinate this multiplicity of stakeholders the government of Zambia created the National HIV/AIDS/STI/TB Council (NAC) with the core mandate to coordinate the national HIV/AIDS response agenda. In turn, the NAC created the Provincial (PATF) and District AIDS Task Forces (DATF) as decentralized structures through which to coordinate the response at provincial and district levels respectively. However, DAFT is not coordinating the response to HIV/AIDS epidemic effectively but the reason for the ineffective coordination has not yet been described. Aim The aim of the study is to explore factors which impact on the capacity of Lusaka DATF to coordinate an effective and sustainable local HIV/AIDS response. Methods A descriptive qualitative research was undertaken using purposive sampling methods. Data collection methods included a Document Review, Key Informant Interviews using semi structured interview guides, Focus Group Discussions and Field Notes. Results The study has found that the following factors promote effective coordination of DATF stakeholders and their activities at local level: recognition of the DATF by stakeholders as a formal structure which is enshrined within the institutional framework of the government; joint planning and resource mobilisation for activities by stakeholders particularly around national commemoration days such as World AIDS Day and the National VCT day and holding of stakeholder forum meetings to share experiences and lessons learnt. The study has also established that the following factors are responsible for preventing the DATF from coordinating effectively include: DATF does not exist as a legal entity enshrined in any Act of Parliament No. 10 of 2002 which created the NAC; there are multiple reporting channels for local stakeholders instead of having only one reporting framework through the DFATF; inadequate stakeholder forums; lack of memoranda of understanding between the DATF and stakeholders; weak technical capacity of the DACA due to factors such as lack of a training programme; inadequate joint planning and resource mobilisation plans and efforts; poor understanding and different interpretations of the of the concept of coordination by different stakeholders; weak feedback mechanisms from the DATF secretariat to stakeholders; and multiplicity of DATF guidelines by different partners which have at times contradicted each other. Poor political, civic and technical leadership engagement has been identified as key inhibiting factors.