This research paper aims to study the lack of local grassroots initiatives in surfing communities by comparing Ukrainian and German contexts. Hence, a qualitative and inductive approach is used. The scope of the research question is explored, by analysing the effects of individual motivators and societal factors needed for GRI formation in active and connected surfing communities. It is further analysed how these motivators are developed and influenced in Germany and Ukraine taking the theories of ecopsychology, social learning and social capital as the framework and analytical lens. Key findings highlight a certain degree of potential for Ukrainian GRIs in surfing communities, while an intense amount of limiting factors is found. Moreover, the chosen comparative context of Ukraine and Germany reveals relevant findings from the collected data, that reveals a low level of trust towards the governments and a low self-efficacy level in Ukraine, while a high level of trust and high self-efficacy was observed in Germany. The research provides relevant insights about the increasing popularity of surfing as a sport, niche, and potential social movement towards environmental activism and sustainable development. Key words: grassroots initiatives, GRIs, environmental activism, environmentalism, surfing movements, surfing communities, ecopsychology, social learning, social capital, Ukraine, Germany, EU.

Deep Neural Networks (DNNs) have proven to be an alternative for object identification formultiple application areas. They are treated as a critical component for autonomouslyoperating systems and consequently crucial for many companies. Since DNNs do not behavein the same way as traditional deterministic systems, there are several challenges to cope withbefore being used in safety-critical applications. Both random and systematic failures must betaken care of, including permanent and transient faults, design faults in hardware andsoftware, and adversarial inputs. In this thesis, we will be constructing an architecture that isrobust and can detect misleading errors produced by a DNN to some extent. One way to copewith failures in DNNs is through architectural mitigation. By adding redundant and diversearchitectures, it can detect misclassification to a greater area. Convolutional neural networkarchitectures will be tested and trained using MATLAB and Simulink. The focus will be onfault-tolerant architectures. The method used in this thesis is experimental research. Theresults show that parallel architectures can detect misleading image classification better. Inaddition, and somewhat unexpected, combining three different networks gives worse resultsthan combining two networks.

Being taken into custody due to excessive intoxication, also called an LOB, is a big infringement of one’s personal liberty and integrity and takes place without the right to a fair trial. The number of such cases in Sweden average at around 60 000 per year, with the point of an action being to provide a safe space for the intoxicated individual to sober up.In this study, the judicial requirements, internal protocols and guidelines which police are to follow in a case of excessive intoxication are laid out and examined. By closely examining two decisions from the Parliamentary Ombudsmen (JO) we have identified situations in which police have arrested someone on subjective grounds, rather than following the internal guidelines. This survey, including 30 police officers from the local district Gothenburg City, was conducted based on the decisions from the Parliamentary Ombudsmen (JO). Officers were given two real world case studies to later describe how they would have acted in the situation. Answers were analysed via code and were thereafter compared to the legal code to make a judgement on officers’ knowledge of LOB. In the first case study, 80% of officers chose to arrest according to LOB, and the remainder chose to reject alternatively remove the person according to The Police Act 13 §In the second case study, where subjective judgement had to be used, 44% chose to arrest according to LOB, 53% chose to follow The Police Act 13 § and 3% answered with answers that were difficult to classify. In the Parliamentary Ombudsmen’s (JO) opinion regarding a similar scenario which was presented in the second case study, there has been critique aimed towards the police officer that decided to arrest the individual for intoxication. It is noticeable that 43% would have chosen to arrest in a similar situation to the one that the Parliamentary Ombudsman had earlier criticized.To examine if there are arrests which take place which lack an actual legal precedent, 71 so called LOB protocols have been looked at, where around 40% of cases lack a satisfactory reasoning as to why the arrest has taken place. This means that there is a legal pitfall when one doesn’t define specifically enough in the LOB protocol why one has made an arrest. From these results, there is an indication that there is a risk that arrests take place which lack legal weight. Finally, the study also highlights that there is a lack of necessities required to care for the arrested in Gothenburg.

During the last years, ethics has gained greater attention in the field of Artificial Intelligence (AI). Stakeholders engage in debates on how to apply ethical principles in various AI organisations with many differing viewpoints. The importance of ethics in AI continuously increases as this area grows rapidly, with a considerable amount of small organisations appearing every day. At the same time, AI solutions enter all the spheres of human activities. Algorithmic accountability can be seen as one of the ways to achieve AI ethics. This thesis will, therefore, explore how small AI organisations can ensure algorithmic accountability. The thesis is based on qualitative data collected from semi-structured interviews in four Swedish small AI organisations. Findings reveal that small AI organisations in Sweden do not apply or consider applying ethical principles when developing their products. The reasons for such decisions are mainly a lack of policies and regulations. The lack of legislation makes ethics of less priority. This thesis proposes governmental bodies facilitate the AI industry by establishing clear policies, standards, and certifications. Such initiatives can assist AI organisations, especially small ones, to address ethical implications and push them towards a high level of AI.  Keywords: Artificial Intelligence, Algorithmic accountability, Small organisations, AI governance

This qualitative essay was aimed to investigate the legal aspects regarding preliminary police investigation. Furthermore, the study also intended to research the decentralization of decision making within the Swedish police force. The study set its focus on the delegation of coercive methods used in certain crime investigations. This in a pursuit to find answers on how the decentralization of certain decisions are applied and function. A legal dogmatic analysis was applied and carried out to determine applicable law. The analysis was performed to gain a broader perspective on how individual police officers perceive the use of decentralization of certain coercive methods, and the subject of initiating a preliminary police investigation. A total of 20 interviews were carried out, consisting of 16 interviews with police officers and 4 interviews with preliminary investigations leaders. The interviews were after completion analyzed using a thematic analysis. Based on the end results of the study, one can see that there is a need for further studies within the subject of initiating a preliminary crime investigation. There is also a need for additional clarification regarding the use of decentralized decision making in certain crime investigations. Further studies within the research field should be aimed at creating greater clarification and uniformity within the Swedish police authority.

Within the Swedish Police Authority, prehospital emergency care is provided according to the method S-CABCDE. The basics of the method refer to prioritizing measures depending on the type of injury in order to save lives and structure the rescue work. Since patrolling police officers commonly find themself being the first responders in different settings where knowledge of prehospital emergency care is required, it is crucial they are qualified and feel confident using standardized methods.The aim of this study was to investigate whether patrolling police officers deemed current education sufficient and examine if knowledge differed depending on years of patrol duty or where the police were stationed. The study was performed as a survey sent to police personnel officiating either in the Police District (PO) of Uppsala or in the PO of Västerbotten. The survey accumulated answers as gender, age and placement of duty. Furthermore, the respondents evaluated their perceived confidence and knowledge in carrying out emergency care in alignment with S-CABCDE using a Likert scale ranging from 1 to 6. Within this thesis, the researchers have settled for 5-6 on the likert scale as acceptable bounds when measuring knowledge and self-confidence. The result was interpreted amidst bivariate analysis and t-test to search for variations between PO Uppsala and PO Västerbotten.Differences that depended on which Police District the respondent was employed at were detected. Moreover, a correlation between training and self-reported knowledge as well as confidence was observed. According to the requirements for acceptable levels of knowledge that this study relies on, there were also knowledge inconsistencies discovered among respondents. For instance, early stages of the method S-CABCDE rendered the lowest rate in the two participating Police Districts alike. Furthermore, respondents who had been employed as patrolling police officers for a period of 10

No description available.

The purpose of this research is to the investigate the effectiveness of the Local Ischemic Compressiontreatment given to the patients who suffered from a cerebrovascular event, and present themusculoskeletal disorders characterized by the presence of trigger points, specifically regardingelliptical motor performance.There exist studies that demonstrate the kinematic characteristics of linear and elliptical movement inpeople with trigger points without neurological injury, however, these characteristics have not beendescribed in patients who have suffered a cerebrovascular event. Thus, this study is novel whenanalyzing the improper motor functioning in non-paretic upper limb movement control. These typesof patients have been approached from the perspective of neurological injury, and the musculoskeletalinjury that could have an impact on movement control has been neglected.This study provides new evidence on the motor performance of the non-paretic upper limb in peoplewho are suffering from musculoskeletal pain. The obtained results helped us to understand the motorfunctioning in the study sample and offers new insights based on the kinematic characteristics andpower law analysis of the movement performed.Analysis being carried out by considering the 2/3 power law, an empirical law which helps in studyingthe relationship between the tangential hand speed and the curvature of its trajectory during curvedmotion, is widely acknowledged to be an invariant of upper-limb movement. Therefore, the study isfocused on this law to interpret the effectiveness of the Local Ischemic Compression treatment givento the patients suffering with musculoskeletal injury resulting myofascial trigger points. The obtainedresults shows that power law is not much altered by the treatment which means that the ischemiccompression has only a peripherical effect (i.e., musculoskeletal system) and does not affect the motorprogramme of the human being (i.e., central nervous system). The one possible reason of having suchresults could be small sample size. Further research is suggested with large study sample.

Additive manufacturing (AM) has made it possible for patients to have personalized healthcare treatment such as implants, surgical guides etc., which was not possible previously. Despite the massive opportunity of custom-made implants, some factors made their use in clinical practice more difficult. Keeping this in focus the purpose of our thesis is i) To identify stakeholders relevant to AM in healthcare ii) Identify barriers for AM in healthcare iii) Propose a business framework based on barriers and stakeholder identification.  A literature review delineated current barriers for AM in general as well as specific to healthcare. Additionally, different dimensions that AM impacts in healthcare and the current business frameworks used were explored. A case study design was applied using deductive reasoning. The Data was collected from semi-structured interviews with the help of snowball purposive sampling. Also, a part of the studies relied on archival data. CFIR framework was employed for data collection, and an interview guide was prepared using the same. The stakeholder analysis model and new business concepts were also utilized under the theoretical framework. Data from the interview was analysed by thematic analysis. The main themes were the Inner setting, Intervention characteristics, Network, Individual professional characteristics, and Outer setting.   The study results are categorized based on the technology used for printing, i.e., Metal printing and Polymer printing. Metal printing superseded polymer printing in terms of the number of barriers. Moreover, in terms of the ‘Inner setting,’ common barriers for both printing technologies were lack of training for medical professionals w.r.t AM machines and lack of skilled personal for handling the design software. In terms of ‘Intervention characteristics’, lack of infrastructure (not good enough scanner for scanning patients) and lack of user-friendly software were few common barriers for both printing technologies. The theme ‘Individual professional characteristics’ contained barriers like lack of entrepreneurial mindset (lacking in individuals who take initiatives and invest time), laggard’s attitude (individuals not willing to change the way they work), technological barrier due to generation gap, and lack of practical experience. Lack of demand and lack of legal framework were barriers for metal printing in the context of the ‘Outer setting’.  Lastly, with evidence of AM facing numerous barriers, it becomes paramount to examine those barriers at different levels critically during its implementation. Addressing and tackling them with a solution regarding business frameworks and knowledge of relevant stakeholders can alleviate some of the hurdles.

Mellanbegrepp inom juridiken utgör ett effektivt sätt att koppla rättsfakta (grunder) till rättsliga följder. Ett intressant mellanbegrepp inom redovisning är ‘väsentlig följd av avvikelse’ som återfinns i den allmänna väsentlighetsprincipen (2kap 3a§ Årsredovisningslagen (SFS 1995:1554) [ÅRL]) vilken ger företag vissa rättigheter att avvika från redovisningsbestämmelser när följden av avvikelsen inte är väsentlig. Syftet med detta examensarbete är att underlätta bedömning av när en avvikelse från en viss bestämmelse får göras genom att applicera teorin om grundöppna mellanbegrepp på den allmänna väsentlighetsprincipen i 2 kap 3a§ ÅRL för ett företag som tillämpar Bokföringsnämndens K3-regelverk vid upprättande av årsredovisning och/eller koncernredovisning. Uppsatsen visar ett förslag på hur en struktur för bedömning kan skapas för detta med hjälp av teorin om mellanbegrepp och en resonemangsmodell för flerdimensionellt beslutsfattande. Ett begreppsligt nätverk struktureras för mellanbegreppet ‘väsentlig följd av avvikelse x’ och möjliga grunder och följder analyseras. Följden är att ‘avvikelse x är inte tillåten’. Grundsidan utgörs av ytterligare mellanbegrepp, ‘rimligen förväntad påverkan på användares beslut’ med grund i ‘väsentlig påverkan på rättvisande bild’. En avvikelses förväntade påverkan på användares beslut enskilt såväl som med andra relaterade avvikelser struktureras med två aggregationsträd. Tre möjliga grunder för en avvikelses påverkan på rättvisande bild diskuteras; påverkan utifrån postens eller felets relativa värde, påverkan utifrån inkonsekvent hantering i förhållande till liknande uppgifter och poster samt påverkan utifrån postens karaktär. Bedömningen struktureras genom att tillämpa en resonemangsmodell för flerdimensionellt beslutsfattande. Med modellen visas hur en upprättare av årsredovisning kan resonera vid prövning av om en avvikelse från en viss bestämmelse är tillåten. Genom att steg för steg se om avvikelsen leder till högre påverkan än vad som är godtagbart på någon aspekt av relevans för väsentlighet hos följd av avvikelse kan upprättare utreda om avvikelsen inte uppfyller något villkor som är nödvändigt för att avvikelsen ska vara tillåten. / Intermediate concepts are used within the law as a means of effectively connecting legal grounds to legal consequences. A particularly interesting intermediate concept within accounting is ‘material consequence of deviation’ which is found in the general materiality principle (2nd Ch. 3a§ The Annual Accounts Act [AAA] of 1995) which gives companies some rights to under certain circumstances deviate from the law of accounting when the consequence of deviating is deemed to not be material. The purpose of this essay is to facilitate assessment of when this deviation is allowed by applying the theory of ground open intermediate concepts on the general materiality principle (2nd Ch. 3a§ AAA of 1995) for a company that applies The Swedish Accounting Standards Board´s K3 regulations for financial reports. The thesis shows an approach to how a structure for decision making can be set up using the theory of intermediate concepts and a model of reasoning for multicriteria decision making. A network of intermediate concepts surrounding ‘material consequence of deviation x’ is presented and possible grounds and consequences analyzed. The consequence is that ‘deviation x is not allowed’. The grounds are also intermediate concepts; ‘reasonably expected influence on users' decisions’ with the ground ‘material consequence on the true and fair view of the company’. A deviation’s expected influence on users’ decisions alone as well as with related deviations is structured in two aggregation trees. Three possible grounds are discussed for the influence of a deviation on a true and fair view of the company; influence based on the item’s or error’s relative value, influence based on inconsistent handling in relation to similar items and influence based on the characteristics of the item. The assessment is structured by applying a reasoning model for multi criteria decision making. With the model it is shown how a preparer of financial reports can reason when assessing if a deviation from a certain regulation is allowed. By step by step assessing if the deviation leads to higher influence than what is acceptable in any aspect of relevance for materiality of a consequence of deviation, preparers can assess if there are any criteria that the deviation does not fulfill which is necessary for the deviation to be allowed.