Made available in DSpace on 2016-06-02T19:06:01Z (GMT). No. of bitstreams: 1 4760.pdf: 3586905 bytes, checksum: 50a6d8eb47edc80b5466b667cc1b31f4 (MD5) Previous issue date: 2010-08-23 / Universidade Federal de Sao Carlos / Background: The process quality model implantation on small enterprises requires an adjustment for this type of enterprise and the difficulty to reach this objective is harder due to the restrict amount of different type of resources. To support the process improvement in this context, concerning the software planning activity, the strategy PCU|PSP was defined. It is based on Use Case Points and Personal Software Process. Aim: The objective of this monograph is to present the PWPlanTe strategy, which is an evolution of the PCU|PSP strategy. PW-PlanTe generalizes the techniques used in the planning, applies well defined iterations aiming to be in conformance with the agile paradigm development, and defines a testing process. In addition, PW-PlanTe is supported by a set of free tools that facilitate its implantation. Methodology: Considering that PCU|PSP is naturally adherent to the agile methods, it was studied and generalized in relation to the planning issues aiming to make it even closer to the Scrum framework. After that, the testing process for small enterprises was defined and the free tools were identified aiming to establish a support for the strategy steps application. Results: PW-PlanTe was applied in three systems that were under development by the small enterprises which participated of this research. In two of the four systems where the planning activity support was explored, the results showed that the strategy facilitates the determination of the work to be done in each iteration. In three of the four systems where the testing activity support was explored, various defects were found by the application of the strategy and were not found without its application. Conclusion: Based on these results, evidences can be identified on the contribution of the strategy for the implantation of the software development process improvement in small enterprises, both for the planning activity as well as for the testing activity. Additional studies should be done aiming to make these results more evident. / Contexto: A implantação de modelos de qualidade de processo em empresas de pequeno porte requer uma adequação para esse tipo de empresa e enfrenta uma maior dificuldade devido às restrições dos diversos tipos de recursos. Para dar suporte à melhoria de processo nesse contexto, no que diz respeito à atividade de planejamento de software, foi definida a estratégia PCU|PSP, que está fundamentada na técnica Pontos de Casos de Uso e no Personal Software Process. Objetivo: O Objetivo deste trabalho é apresentar a estratégia PW-PlanTe, que é uma evolução da estratégia PCU|PSP. A PW-PlanTe generaliza as técnicas a serem adotadas no planejamento, utiliza iterações bem definidas de modo a atender o paradigma de desenvolvimento ágil, e define um processo de teste. Além disso, a PW-PlanTe é apoiada por um conjunto de ferramentas livres que facilitam sua implantação. Metodologia: Dada a aderência da PCU|PSP aos métodos ágeis, essa estratégia foi estudada e generalizada no que diz respeito à parte de planejamento, para se aproximar ainda mais do framework Scrum. Após isso, foi definido o processo de teste para o contexto de uma empresa de pequeno porte e foram identificadas ferramentas livres para dar suporte a todos os passos da estratégia. Resultados: A PW-PlanTe foi aplicada em três sistemas que estavam em desenvolvimento nas empresas de pequeno porte participantes do trabalho. Em dois sistemas, nos quais foi explorado o suporte à atividade de planejamento, os resultados mostraram que a estratégia facilita a determinação de trabalho a ser realizado em cada iteração. Em dois sistemas, nos quais foi explorado o suporte à atividade de teste, vários defeitos foram encontrados com o uso da estratégia e não foram revelados sem a sua aplicação. Conclusões: Com base nos resultados, existem evidências da contribuição da estratégia para a implantação da melhoria de processo de desenvolvimento de software em empresas de pequeno porte, tanto no que diz respeito à atividade de planejamento, como de teste.

This study explores stakeholder analysis and requirements engineering within an agile project at a system development company requesting a case management system. Stakeholder analysis is commonly explored in traditional project planning studies, such as the waterfall model. In agile projects, on the other hand, it is not common to have a project-planning phase or to carry out a stakeholder analysis, often due to that a stakeholder analysis entail a long and extensive documentation. Our study explores how Soft Systems Methodology can be used in an agile project-planning phase as well as how rich pictures from the SSM support requirements engineering. The study is based on an agile method in which the project-planning phase is expected to deliver a product log, and hence a requirements list. Requirements are collected, modelled and validated together with users in the form of user stories where role, goals and purpose are expressed in a sentence. Stakeholders and roles are identified and analyzed using methods in Soft Systems Methodology, like "finding-out" analyses. The result is then presented in so called rich pictures of the current situation and in a future desired situation. The data collection is carried out together with informants at the case company in the form of observations and semi-structured interviews. The result of the study indicates how rich pictures from SSM can support the start-up of an agile project. / Denna studie utforskar intressentanalys och kravinsamling inom ett agilt projekt på ett systemutvecklingsföretag som efterfrågar ett ärendehanteringssystem. Intressentanalyser är vanligt förekommande och utforskade i traditionella förstudier så som enligt vattenfallsmodellen. I agila projekt är det däremot vanligt att en förstudie inte utförs eller att förstudien inte innehåller en intressentanalys, ofta på grund av att intressentanalyser innebär lång och omfattande dokumentation. Vår studie utforskar hur Soft Systems Methodology kan användas i en agil förstudie samt hur rika bilder från SSM stöttar kravhantering. Studien utgår från an agil metod där förstudien förväntas leverera en produktlogg, där studiens fysiska bidrag blir en kravlista. Krav samlas in, modelleras samt valideras tillsammans med användare i formen av användarhistorier där roll, mål och syfte uttrycks i en mening. Intressenter och roller identifieras och analyseras med hjälp av metoder inom Soft Systems Methodology, som exempelvis ”finding out”-analys. Resultatet presenteras därefter i så kallade rika bilder över nuvarande situation samt i en framtida önskad situation. Datainsamlingen görs tillsammans med informanter på ett fallföretag i form av observationer och semistrukturerade intervjuer. Utfallet av studien visar hur den rika bilden från SSM kan stödja uppstarten av ett agilt projekt.

[pt] Grandes empresas de mídias digitais adotaram métodos ágeis para desenvolver produtos em ciclos mais curtos. Com isso, o tempo para testar as interfaces foi reduzido. Através de observações, verificou-se que nem sempre os testes curtos respondem a todas as perguntas que precisam de investigação junto aos usuários. Com o objetivo de propor uma metodologia de avaliação de usabilidade empírica, desenvolvida especificamente para uma melhor adequação às necessidades das equipes que desenvolvem produtos com métodos ágeis, realizou-se um estudo de caso, na Globo.com, com a equipe que projetou o site Receitas.com através da metodologia ágil Scrum. Com a realização de um grupo de foco, avaliou-se o cenário e definiram-se as questões de pesquisa. A partir daí, foi constatado que aplicação de somente uma técnica não iria responder à maior parte dessas questões. Além disso, caso fossem aplicadas de uma forma mais tradicional, poderia levar muito tempo. Como hipótese, acreditava-se que a criação e aplicação de uma metodologia de avaliação de usabilidade empírica, especificamente desenvolvida para melhor adequação às necessidades dessas equipes, seria capaz de resolver o problema, através da otimização e aplicação de técnicas em conjunto, como entrevista, eye tracking e teste de usabilidade. A metodologia foi validada através de entrevistas com cada integrante da equipe de desenvolvimento do Receitas.com. Os resultados obtidos comprovaram a hipótese da pesquisa. / [en] Big digital media companies have embraced agile methods to develop products in shorter cycles. With this, the time to test the interfaces was reduced. Through observations, we found that short tests do not always answer to all questions that need investigation with users. In order to propose a methodology for empirical usability evaluation, designed specifically to better fit the needs of teams who develop products with agile methods, we performed a case study in Globo.com, with the team that designed the site Receitas.com through the Scrum agile methodology. With the completion of a focus group, we assessed the scene and found the research questions. From there, it was found that application of only one technique would not answer most of these issues. Moreover, if applied in a more traditional way, could take a long time. As a hypothesis, it was believed that the creation and application of a methodology for empirical evaluation of usability, specifically designed to better match the needs of these teams, would be able to solve the problem through the optimization and application of techniques together, such as interview, eye tracking and usability testing. The methodology was validated through interviews with each team member s development Receitas.com. The results confirmed the hypothesis.

Ces travaux portent sur la conception multidisciplinaire de systèmes intégrés. Ces systèmes sont soumis à un nombre d’exigences toujours croissant, entraînant des besoins en termes d’intégration fonctionnelle et spatiale. Ces différents types d’intégration relative au produit sont également la source d’une complexité organisationnelle, provenant à la fois de la multitude d’acteurs réalisant différentes activités d’ingénierie, mais également de la diversité des domaines impliqués, désignée dans ce manuscrit par « intégration multidisciplinaire ». Pour favoriser cette intégration multidisciplinaire, les phases de « conception préliminaire » et de « conception détaillée » ont été identifiées comme déterminantes, notamment car elles se caractérisent par la collaboration de nombreux experts, manipulant un grand nombre de données techniques de définition. Les systèmes conçus lors de conceptions multidisciplinaires restent faiblement intégrés. Cela est en partie dû au cloisonnement entre les disciplines et à un mode d’organisation projet basé sur une planification prédominante, caractérisé notamment par une diffusion de l’information principalement descendante (top-down). Afin d’assurer une meilleure collaboration entre ces différentes disciplines, de permettre des prises de décision éclairées par des indicateurs opérationnels et de pouvoir analyser et mieux comprendre les phénomènes d’intégration des expertises, l’introduction d’une méthode inspirée des principes fondateurs des méthodes agiles est proposée pour la conception collaborative de systèmes intégrés.La contribution de ces travaux s’appuie sur trois concepts complémentaires. Le premier, intitulé Collaborative Actions Framework correspond à un cadre de collaboration opérationnelle autour d’actions. Un des objectifs de ce framework est de faciliter la collaboration des acteurs des projets de conception, quelle que soit leur origine disciplinaire, mais également d’assurer une traçabilité entre les prises de décision et les corrections/modifications apportées sur les données techniques. Cette traçabilité est rendue possible grâce aux liens existants avec le second concept intitulé Workspace. Apportant un nouvel éclairage sur les possibilités offertes par la collaboration autour de ces espaces de collaboration, ce concept offre un certain nombre de possibilités,notamment la mise en commun continue des travaux, l’intégration multidisciplinaire et la validation des modifications. Les échanges de données techniques entre les workspaces, ou le travail simultané sur les mêmes données techniques, s’appuient quant à eux sur la possibilité de pouvoir gérer de façon parallèle différentes versions d’une même donnée technique. Ces possibilités sont proposées par le troisième concept, intitulé branch & merge, qui permet également à différents acteurs de travailler simultanément sur les mêmes données. Enfin, ces trois concepts sont ensuite illustrés par l’intermédiaire d’un démonstrateur composé d’un scénario et d’un prototype informatique. Un produit mécatronique, combinaison synergique et systémique de la mécanique, de l'électronique et de l'informatique temps réel, est utilisé afin d’illustrer les possibilités offertes par nos travaux en termes d'intégration multidisciplinaire lors de la conception collaborative. / This work focuses on the multidisciplinary and collaborative design of integrated systems. These systems are subject to an ever increasing number of requirements, leading to the need for more comprehensive functional and spatial integration. These different types of product integration are also at the origin of organizational complexity. This complexity arises not only from the great number of actors performing various engineering activities but also from the diversity of disciplines involved (designated in this manuscript as “multidisciplinary integration”). To encourage this multidisciplinary integration, “preliminary design” and “detailed design” have been identified as the most significant steps, especially since they are characterized by the collaboration of multiple experts handling a large number of product definition’ technical data. Systems that have been designed thanks to multidisciplinary approaches are generally poorly integrated. This is partially due to the compartmentalization of disciplines, as well as to the “project-planned” method, where project planning is predominant and information is mainly spread out “top-down”. To ensure better cooperation between the various disciplines, to enable decision making based on operational indicators and to analyze and understand the multidisciplinary integration processes, a method inspired by the founding principles of agile methods (the agile manifesto) is proposed for the collaborative design of integrated systems. This work is based on three complementary concepts. The first is, the Collaborative Actions Framework, an operational framework for collaboration around actions. One objective of this framework is to improve the collaboration among designers, whatever their disciplinary origin. It also ensures traceability between decision making and corrections/changes made to technical data. This traceability is made possible by the useof the second concept, called Workspace. Even if this term is already well known, we propose a new definition/usage to transform it into collaboration spaces. This concept offers great possibilities, including the continuous delivering/sharing of experts’ contributions, multidisciplinary integration and change validation. The exchange of technical data between workspaces, or simultaneous work on the same data, relies on the ability to manage several parallel versions of the same item into a single datamanagement system. These opportunities are offered by the third concept, called Branch & Merge. Finally, these three concepts are illustrated through a scenario and a computer prototype. A mechatronic product, “the synergistic combination of mechanical and electrical engineering, computer science, and information technology” (Harashima et al., 1996), is used to illustrate the opportunities offered by our work in terms of multidisciplinary integration during collaborative design.