Esta dissertação aborda um ramo da computação que se encontra em crescente desenvolvimento: a computação em tempo real. Os sistemas de computação tempo real surgiram a partir da necessidade de substituição do controle humano, que muitas vezes é falho, em situações complexas ou críticas, onde máxima confiabilidade e disponibilidade são exigidas para garantir a segurança do sistema. A área de aplicação diferencia-se de outras convencionais por possuir diferentes tipos de restrições de tempo e operar em ambientes não-determinísticos. Entretanto, atualmente tais sistemas estão tornando-se grandes, complexos, distribuídos, adaptativos e cada vez mais presentes nas aplicações do dia-a-dia,o que tende a exigir soluções mais simples e generalizadas. Pelo fato de tais sistemas normalmente atuarem sobre aplicações críticas, importante salientar que, em algumas situações, pequenos erros no sistema podem levar a grandes catástrofes. Mesmo atrasos mínimos no tempo de resposta são problemáticos, podendo ocasionar degradações ou ações erradas no mundo físico controlado pelo sistema tempo real. Como nestes casos máxima confiabilidade e disponibilidade são exigidas para garantir a sua segurança, tornou-se importante a construção de sistemas tempo real tolerantes a falhas. Dessa forma, é visivelmente crescente a necessidade de utilização de mecanismos capazes de abordar os requisitos de tempo real e tolerância a falhas de forma integrada durante o desenvolvimento do sistema. Assim, o processo de desenvolvimento de sistemas tempo real confiáveis torna-se mais simples e mais eficiente. A necessidade de maior conhecimento do uso de tolerância a falhas para obter segurança no funcionamento de aplicações tempo real levou ao desenvolvimento deste trabalho, onde buscou-se um caminho de solução para a adequação das técnicas de tolerância a falhas a estas aplicações. Sabe-se que para produzir software confiável e, desta forma de maior qualidade, além do emprego de boas técnicas de engenharia de software, é necessário compreender os principais conceitos e técnicas de tolerância a falhas. Por outro lado, é importante ter-se conhecimento dos mecanismos oferecidos pelas diversas camadas de software de um sistema - protocolo de comunicação, sistema operacional e linguagem de programação - para apoiar estas atividades de tolerância a falhas. Este trabalho busca analisar os mecanismos e técnicas usados na implementação de software tolerante a falhas frente às situações mencionadas, uma vez que nem todas as técnicas conhecidas podem ser indistintamente aplicáveis a estas situações. Os resultados desta análise são organizados na forma de uma taxonomia, visando assim auxiliar projetistas de desenvolvimento de software a tomarem decisões importantes na construção de sistemas tempo real tolerantes a falhas. Os mecanismos são agrupados de acordo com o nível de implementação: sistemas operacionais, linguagens de programação e protocolos de comunicação, destacando suas características e aplicabilidade. Por fim uso da classificação é demonstrado com a análise de três casos-exemplo. / This dissertation is about a, computer science field which is in growing development, that is, real-time computation. Real-time computing systems have emerged from the necessity of substituting. human control which is sometimes failed in complex or critical situations. In these ones maximum availability and reliability are requested in order to guarantee the system dependability. The application area differs from the conventional ones because it has particular time constraints and operates in nondeterministic environments. Nevertheless, nowadays such systems are becoming large, complex, distributed and adaptive but tend to demand simpler and generalized solutions as they are more present in daily applications. Since such systems normally act on critical applications it is important to reinforce, that in some situations, subtle systems errors may generate big catastrophes. Even slight delays in response time are troublesome and they may cause degradation or wrong acts in physical world controlled by real-time systems. In these cases maximum reliability and availability are requested in order to guarantee system dependability. Thereby, the requirement of including mechanisms capable of achieving real-time and fault tolerance in an integrated way during the system design has been increased. Thus, the developing process of reliable real-time systems becomes simpler and more effective. The necessity of improving designers knowledge on using fault tolerance in order to obtain dependability on real-time applications has motivated this study. Our main goal has been to find an adequate way of using fault tolerance techniques to these applications. It is known that the development of reliable software not only requires appropriate software engineering techniques but also demands understanding of main politics and mechanisms used to implement fault tolerance techniques in these situations. Otherwise, it is very important to know the related support that is offered by the different software levels of a system - communication protocol, operating system and programming language. This study has as purpose analyzing the mechanisms and techniques used in implementation of fault-tolerant software applied to the previously mentioned situations. The basic supposition is that not all the known techniques may be applied indistinctly to these situations. The properties of the software are organized according to a taxonomy, where the mechanisms are bracketed in groups according to implementation level: operating systems, programming languages and communication protocols. In this presentation, the characteristics and applicability of the software tools are stood out in order to help developing-software designers to decide what is important to build faulttolerant software. Finally, the use of the classification is demonstrated by analyzing three case-examples.

When working with oceans in the visual effects industry, it is not always very practical or even possible to use real live footage, especially if extreme weather conditions are required. A common scenario is computer generated objects crashing into an ocean generating splashes and foam that should stick to and integrate well with the real ocean surface. Making a shot like that look realistic is very difficult, this is where a fully computer generated ocean surface comes in handy. Creating high resolution computer generated sequences of an ocean surface with interacting objects is difficult using today’s available commercial 3D packages. I have therefore implemented a VFX ocean toolkit, which is a system built for generating the ocean surface, Kelvin wakes and interaction with objects. The ocean toolkit was built with the artist in mind and the need for real time preview to produce results quick and easy in order for the system to remain cost effective. The ocean toolkit is tightly integrated directly into the procedural 3D animation package Houdini1 as several plug-ins and shaders that can be combined to create numerous ocean surface effects.

Typning av hla-b*27:En jämförelsestudie mellan två analysmetoder för att påvisa HLA-B*27 molekylen i ankyloserande spondylitCarolina BermudezBermudez, C. Typning av HLA-B*27. En jämförelsestudie mellan två analysmetoder för att påvisa HLA-B*27 molekylen i Ankyloserande spondylit. Examensarbete i Biomedicinsk vetenskap, 15 högskolepoäng. Malmö universitet: Fakulteten för hälsa och samhälle, institutionen för Biomedicinsk vetenskap, 2018.Human leukocyt antigen (HLA) är vävnadsantigener, belägna på våra vita blodkroppar. HLA-B*27 allelen är starkt kopplat till Ankyloserande spondylit (AS). Det är en kronisk inflammatorisk ledsjukdom, som främst attackerar ryggraden, bäckenet och bröstkorgen. Det finns idag ingen enskild laborativ metod som med full säkerhet kan fastställa diagnos av denna sjukdom, innan de kliniska symtomen uppträder. Typning av HLA-B*27 ger endast information om närvaro eller frånvaro av antigenet, vid utredning av AS. Vidare är HLA-B*27 en polymorf och de olika alleltyperna varierar kraftigt, bland skilda etniska grupper samt mellan geografiska områden. Genetiska- och miljöfaktorer påverkar också. Sjukdomsutveckling i samband med närvaro av HLA-B*27 allelen, varierar därför från individ till individ. Därmed fungerar metoden endast som ett komplement-verktyg, för att ytterligare bekräfta diagnos. Syftet med denna studie var att med realtids-polymerase chain reaction (PCR), utföra typning av HLA-B*27 med Linkseq kit samt jämföra analysresultaten med uthämtade resultat från intern sjukhusdatabas, där typning av HLA-B*27 hade utförts med PCR-SSP (sekvens-specifika primers). Samtliga resultat stämde överens till 100%, vilket indikerar att metoden fungerar bra. Det finns studier som visat att HLA-B*27 molekylens fria tunga kedjor (HLA-B*272) har en starkare benägenhet än andra HLA-molekyler att binda in till killer immunoglobine-like receptorer (KIRs). Inbindning till KIRs med efterföljande ökad stimulering av interleukiner (IL) främst IL-17 och IL-23 bidrar till sjukdomsutvecklingen av AS. Dock finns ingen HLA-B*272 specifik antikropp som kan bevisa detta och det behövs därför ytterligare undersökning för att hitta en sådan. Därefter skulle en ny laborativ metod kunna utvecklas för att fastställa diagnos av AS i ett tidigt skede, innan de kliniska symtomen uppvisas. Nyckelord: Allelvarianter, Ankyloserande spondylit, HLA-B*27, KIR, PCR-SSP, Realtids-PCR. / typing of hla-b*27:a comparison study between two analysing methods for the detection of the HLA-B*27 molecule in ankylosing spondylitisCarolina BermudezBermudez, C. Typing of HLA-B*27. A comparison study between two analysing methods for the detection of the HLA-B*27 molecule in Ankylosing spondylitis. Degree project in Biomedical Laboratory Science, 15 credit points. Malmö University: Faculty of Health and Society, Department of Biomedical science, 2018.Human leukocyte antigen (HLA) are tissue antigens located on our white blood cells. The HLA-B*27 allele is strongly related to Ankylosing spondylitis (AS). It is a chronical inflammatory rheumatic disease that primarily affects the spine, the pelvis and the chest. At present, there is no single laboratory method that with all certainty may determine diagnosis of this disease, before the clinical symptoms appear. Typing of HLA-B*27 only gives information about the presence or absence of the antigen, upon the investigation of AS. Furthermore, HLA-B*27 is a polymorph and the different types of alleles, strongly vary among different ethnic groups and also between geographic regions. Genetic- and environmental factors also affect. Development of disease in conjunction with the presence of the HLA-B*27 allele, therefore varies from one individual to another. So, the method only functions as a complementary tool, to further confirm diagnosis. The aim of this study was to perform HLA-B*27 typing with realtime-polymerase chain reaction (PCR) using Linkseq kit and compare the analysed results with those results that were retrieved from the internal database of the hospital, where typing of HLA-B*27 had been performed with PCR-SSP (sequence specific primers). All results agreed with 100%, which indicates that the method functions well. There are studies that show that the heavy chains (HLA-B*272) of the HLA-B*27 molecule have a stronger affinity than other HLA-molecules of binding in to killer immunoglobulin-like receptors (KIRs). Increased stimulation of interleukins (IL) primarily IL17 and IL23, following binding to KIRs, contributes to the pathogenesis of ankylosing spondylitis. However, there is no HLA-B*272 specific antibody that may prove this and therefore more investigation is needed, in order to find one. A new laboratory method could then be developed to determine diagnosis of AS at an early stage, before the clinical symptoms emerge. Keyword: Allelvariants, Ankylosing spondylitis, HLA-B*27, KIR, PCR-SSP, Realtime-PCR.

The ain of this thesis is to define and describe specific challenges occuring on the crossroard between project management and knowledge management with the focus on agile software development and agile modeling. Based on the found and verified problem it tries to find a existing solution. After that, it analyses, specifies and designs an own solution. Focusing on covering of three different perspectives makes this thesis unique. After process of design, there are technologies defined. For all used technologies there is described detailed implementation of the application. The third-party technologies are connected in this application.This connection creates the extra added value for the application and user in processes of agile software development and agile modeling.

This thesis addresses the topic of object detection on graphics processing units. As a part of it, a system for object detection using NVIDIA CUDA was designed and implemented, allowing for realtime video object detection and bulk processing. Its contribution is mainly to study the options of NVIDIA CUDA technology and current graphics processing units for object detection acceleration. Also parallel algorithms for object detection are discussed and suggested.

This work is dealing with detecting and recongnizing matrix codes. It is experimenting use of PCLines algorithm. PCLines is using Hough transform and parallel coordinates for fast detection of lines. Suggested algorithm with double use PCLines detects sets of parallels and problem with image distorted by with parallel projection is solved by cross-ratio equation. We did some optimizations for realtime running and created experimental implementation. Test results shows, that use PCLines one way to detect matrix codes.

The goal of master’s thesis was to demonstrate and benchmark peformance of Mantle and Vulkan APIs with different optimization methods. This thesis proposes a rendering toolkit with optimization methods based on parallel command buffer generating, persistent staging buffers mapping, minimal pipeline configuration and descriptor sets changing, device memory pre-allocating with managing and sharing between multiple resources. The result is reference implementation that could render dynamic scene with thousands of objects in real time.

Der Einsatz von Computern, zur unterstützenden Durchführung physikalischer Experimente, gewinnt immer mehr an Bedeutung. Computer, unter anderem auch Personal Computer (kurz PC), können durch ihre heutige Leistungsfähigkeit, immer mehr nützliche Funktionen während eines physikalischen Experimentes übernehmen. Diese Arbeit beschäftigt sich mit dem Vorbereiten eines solchen PC, welcher für die Umsetzung einer Regelschaltung für ein physikalisches Experiment genutzt werden kann. Dazu werden Gentoo-Linux, RTAI, Scilab/Scicos und Comedi verwendet. Besonderer Schwerpunkt der Arbeit ist dabei die Regelung in Echtzeit auszuführen. Im Verlauf der Arbeit werden Hardware- sowie Software-Ansprüche und deren Anpassungen besprochen, so dass alle Voraussetzungen des Experiment-PC, zur Umsetzung eines zuverlässigen Regelprozesses, erfüllt werden. Aus Gründen, welche in dieser Arbeit erläutert werden, wurde der Experiment-PC mit der Open-Source Linux-Distribution namens Gentoo ausgestattet. Zusätzlich verfügt der Experiment-PC über eine Mess- und Steuerkarte, welche es dem Experiment-PC erlaubt Signale aus dem Experiment aufzunehmen und eine entsprechende Regelung, durch die Ausgabe von Signalen, vorzunehmen.

A passive recirculation of hydrogen using ejectors is effective and efficient only in a limited operating window, which kept ejectors until recently from being integrated into automotive fuel cell systems, where a more dynamic operation is expected. Strategies like parallel setup or a PWM-drive employed to expand the operating window demand reliable control algorithms. Such algorithms can greatly benefit from the predictive power a mathematical model. In a fuel cell system, the anode and cathode side cannot be separated, and a mathematical description should encompass all components to a reasonable degree, to allow the model to run on a low power automotive-grade platform in real time. This contribution demonstrates an approach to reduce the computation expense of the mathematical model.

GNSS-Messungen werden neben systembedingten Fehlereinflüssen vor allem von den Auswirkungen der Mehrwegeausbreitung und Signalbeugung insbesondere in der Empfangsumgebung dominiert. Verschiedene Dienste z.B. der Landesvermessungsämter haben deshalb ein primäres Interesse daran, die Auswirkungen der Effekte möglichst gering zu halten oder aber genau bestimmen zu können, um Korrekturwerte zu generieren. Mehrwege- und Beugungseffekte lassen sich besonders innerhalb von Netzstrukturen gut bestimmen. Liegen Sollkoordinaten aller Beobachtungsstationen vor gelingt dies auch in Echtzeit. In der vorliegenden Arbeit werden neben einer detaillierten Beschreibung der jeweiligen Einflussgrößen auch Möglichkeiten aufgezeigt, die genannten Effekte zu erkennen und Maßnahmen zur Reduktion der Auswirkungen auf das Meßergebnis zu ergreifen. Kern der Untersuchungen ist ein zweistufiges Modell zur Reduzierung von Mehrwegeeffekten in Echtzeit innerhalb von (Referenz-) Stationsnetzen durch Bestimmung von Korrekturwerten für originale und abgeleitete Meßwerte pro Epoche, Station und Satellit.