Submitted by Franciele Moreira (francielemoreyra@gmail.com) on 2017-09-11T13:42:06Z No. of bitstreams: 2 Dissertação - Jailton Alkimin Louzada - 2017.pdf: 5226619 bytes, checksum: bb291ad68acd8cf80fcb2c6c2392c596 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2017-09-15T15:43:54Z (GMT) No. of bitstreams: 2 Dissertação - Jailton Alkimin Louzada - 2017.pdf: 5226619 bytes, checksum: bb291ad68acd8cf80fcb2c6c2392c596 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2017-09-15T15:43:54Z (GMT). No. of bitstreams: 2 Dissertação - Jailton Alkimin Louzada - 2017.pdf: 5226619 bytes, checksum: bb291ad68acd8cf80fcb2c6c2392c596 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2017-08-11 / The Information Technology (I.T) market is growing. In the age of information, world economies have been investing more and more in the services market. Within this competitive scenario, software testing is an important component for raising the quality of software developed in Brazil and in the global competitiveness cenario. However, Small and medium-sized enterprises (SMEs) have limited resources for investments in software testing maturity processes, tools and models in their business. Faced with this, this work has as main proposal to produce an apparatus to improve the process of software testing for SMEs. Objective: As the main objectives, this study proposed a more update version of the FreeTest 1.0 process, as well as practical instructions on how to implement the activities suggested in the process, all this formatted in a new process and a wizard respectively. Methodology: In order to fulfill the objectives of this work, the FreeTest 2.0 process was created as an improvement of the FreeTest Method process, focused mainly on Agile techniques, DevOps and aligned with the SMEs ecosystem. And the FreeTest Wizard, which consists of a deployment guide that supports the implementation of the process in a didactic, dynamic and flexible way. Another contribution of this work was the creation of support tools to disseminate this knowledge and content management, in this case the creation of a web platform, distributed free of charge and in the format "as a Service". Finally, the results and conclusions can be seen in the final chapter of this work. / O mercado de Tecnologia da Informação (T.I.) é crescente. Na era da informação, as economias mundiais investem cada vez mais no mercado de Serviços. Dentro deste cenário competitivo, o teste de software é um importante componente para elevação da qualidade do software desenvolvido no Brasil e sua competitividade mundial. Contudo, as Micro e Pequenas Empresas (MPEs) possuem recursos limitados para investimentos em processos, ferramentas e modelos de maturidade de teste de software em seus negócios. Diante disto, este trabalho tem como proposta principal produzir um aparato para melhoria do processo de teste de software para MPEs. Objetivo: Como objetivos principais, este estudo propôs uma versão mais atualizada do processo FreeTest 1.0, bem como instruções práticas de como implantar as atividades sugeridas no processo, tudo isso formatado em um novo processo e um guia de implantação, respectivamente. Metodologia: Com o propósito de cumprir os objetivos almejados neste trabalho foram criados: 1) o processo FreeTest 2.0 como uma melhoria do processo do FreeTest 1.0, focado principalmente em técnicas Ágeis, DevOps e alinhado ao ecossistema das MPEs; e 2) o FreeTest Wizard, que consiste em um guia de implantação que apoia a implantação do processo de forma didática, dinâmica e flexível. Outra contribuição deste trabalho foi a criação de ferramentas de apoio para disseminação deste conhecimento e gestão dos conteúdos, neste caso, a criação de uma plataforma web distribuída de forma gratuita e no formato "as a Service". Por fim, os resultados e conclusões serão vistos no capítulo final deste trabalho.

L’électronique dans les voitures devient de plus en plus complexe et représente plus de 30% du coût global d’une voiture. Par exemple, dans une BMW série 5 modèle 2008, on peut trouver jusqu’à 80 calculateurs électroniques communiquant ensemble et représentant aux alentours de 10 millions de lignes de code logiciel. Face à cette montée en complexité, les constructeurs et équipementiers électroniques de l’automobile s’intéressent de plus en plus à des méthodes efficaces de développement, vérification et validation de modules électroniques. Plus précisément, ils focalisent leurs efforts sur la partie logicielle de ces modules puisqu’elle est à l’origine de plus de 80% des problèmes détectés sur ces produits. Dans ce contexte, nous avons mené un travail de recherche dont l’objectif est de proposer une approche globale d’amélioration de la qualité des logiciels embarqués dans les véhicules. Notre recherche part d’un audit des processus et outils actuellement utilisés dans l’industrie électronique automobile. Cet audit a permis d’identifier des leviers potentiels d’amélioration de la qualité logicielle. En se basant sur les résultats de l’audit et en tenant compte de la littérature dans le domaine de la qualité logicielle, nous avons proposé une approche globale de conception de cas de test pour les produits logiciels. En effet, nous avons développé une plateforme de génération automatique de tests pour un produit logiciel. Cette plateforme consiste à modéliser les spécifications du produit logiciel pour le simuler lors de tests, à se focaliser sur les tests critiques (ayant une forte probabilité de détecter des défauts) et à piloter la génération automatique des tests par des critères de qualité ; telles que la couverture du code et de la spécification mais aussi le coût des tests. La génération de tests critiques est rendue possible par la définition de profils d’utilisation réelle par produit logiciel, ainsi que par la réutilisation des défauts et des tests capitalisés sur des anciens projets. En plus des aspects algorithmiques du test logiciel, notre approche prend en compte des aspects organisationnels tels que la gestion des connaissances et des compétences et la gestion de projet logiciel. Notre approche a été mise en œuvre sur deux cas d’étude réels d’un équipementier électronique automobile, disposant de données de tests historiques. Les résultats de nos expérimentations révèlent des gains de qualité significatifs : plus de défauts sont trouvés plus tôt et en moins de temps. / Nowadays, car electronics become more and more complex and represents more than 30% of the total cost of a car. For instance, in a 2008 BMW 5 series model, one can find up to 80 electronic modules communicating together and representing 10 million lines of software code. Facing this growing complexity, carmakers and automotive electronic suppliers are looking for efficient methods to develop, verify and validate electronic modules. In fact, they focus on the software part of these modules since it accounts for more than 80% of the total number of problems detected on these modules. In this context, we achieved our research project with the aim of proposing a global approach able to improve the quality of automotive embedded software. We started with an audit of the software practices currently used in automotive industry and we pinpointed potential levers to improve the global software quality. Based on the results of the audit and the literature review related to software quality, we developed a global approach to improve the design of test cases for software products. In fact, we developed a test generation platform to automatically generate test cases for a software product. It is mainly based on modeling the software functional requirements in order to be simulated when testing the software, focusing on critical tests to be done (because of their higher probability to detect a bug) and monitoring the automatic generation of tests by quality indicators such as the structural and functional coverage but also the tests cost. The generation of critical tests is based on the definition of real use profiles by software product and on the reuse of bugs and test cases capitalized on previous projects. Besides the computational aspects of software testing, our approach takes into account organizational matters such as knowledge management, competency management and project management. Our approach have been implemented in a computer platform and experimented on two typical case studies of an automotive electronic supplier, with historical test data. The results of our experiments reveal significant improvement in software quality: more bugs are detected earlier and in less time.

Submitted by Natalia de Souza Gonçalves (natalia.goncalves@ufpe.br) on 2016-09-28T12:23:59Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Dissertacao_Cleydiane_FINAL_Imprimir.pdf: 3971527 bytes, checksum: f06a014d16a761f030bbfccee75d176f (MD5) / Made available in DSpace on 2016-09-28T12:23:59Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Dissertacao_Cleydiane_FINAL_Imprimir.pdf: 3971527 bytes, checksum: f06a014d16a761f030bbfccee75d176f (MD5) Previous issue date: 2016-02-19 / Dívida Técnica (DT) está relacionada a tarefas que devem ser executadas e são acumuladas ao longo de um projeto para serem realizadas posteriormente. A falta de planejamento de tempo e recursos em um projeto, dentre outros problemas, propiciam seu surgimento. DTs não solucionadas tendem a criar dependências e aumentar o grau de complexidade para correção, resultando em maior esforço, custo e retrabalho em um projeto. A sua identificação e gerenciamento ao longo de um projeto é essencial para minimizar seu impacto e consequências negativas. Estudos apontam teste de software como uma das principais áreas impactadas por DTs em projetos de software. Nesta dissertação foram mapeadas, a partir da literatura técnica, 22 possíveis DTs relacionadas ao processo de teste de software, assim como suas causas e indicadores. Elas foram avaliadas por meio de um survey com profissionais da área de teste de software. Como resultados do survey, são apresentados os níveis de concordância obtidos para cada DT, suas causas e indicadores sugeridos. Após essa validação da concordância quanto às DTs, foi identificada a necessidade de vincular soluções às causas de cada DT, para isto foi realizada uma nova revisão de literatura. A partir dos dados das DTs, com suas causas, indicadores e, agora, soluções, foi elaborado um mapa, com objetivo de apoiar profissionais da área de teste de software na gestão de DTs que podem ocorrer no decorrer da execução do processo de teste de software. Após a elaboração do mapa de gestão de dívida técnica, o mesmo foi avaliado a partir de um questionário (baseado na técnica de revisão em pares), por profissionais experientes em teste de software (líderes e consultores). Como resultado desta avaliação, são apresentadas as opiniões e considerações destes profissionais sobre a versão inicial do mapa elaborado. Por fim, é apresentado o resultado final do mapa, com as devidas alterações baseadas nas respostas obtidas no questionário de avaliação do mapa. / Technical Debit (TD) is related to the tasks which have to be executed but are accumulated during the development of a project to be implemented later. The lack of time and resource planning on a project, among other problems, facilitate its emergence. Unsolved TDs tend to create dependencies and increase the degree of complexity for corrections, resulting in bigger efforts, costs and rework on a project. Identifying and managing those issues are essential to minimize the impacts and negative consequences. Studies point out software testing as one of the most affected areas by TDs on a software project. At this dissertation, 22 (twenty-two) possible TDs related to testing software process were mapped from the technical lecture, along with its causes and indicators. They were evaluated through surveys with testing software experts. As result of the survey, it is presented the levels of accordance for each TD, also the suggested causes and indicators. After the accordance validation of the TDs, it was identified a demand to bind solutions to the causes of every TD, for this a new study was made. From the data of the TDs, along with its causes, indicators and solutions, a map was produced in order to support testing software professionals managing TDs that may appear during the testing software process execution. After producing the technical debit management map, the same was submitted for an evaluation, using a questionnaire (based on peer review technique), to testing software experts (leaders and consultants). As result of these evaluations, it is presented the opinions and remarks of those professionals about the final version of the produced map. Finally, it is presented the final result of the map, with the modifications based on the answers obtained through the map evaluation questionnaire.