Pour améliorer l’analyse des risques dans les projets et renforcer son efficacité, les gestionnaires de projet devraient réutiliser les expériences et les bonnes pratiques acquises au cours de projets antérieurs. Le retour d’expérience constitue une source de connaissances importante pour réduire les niveaux d’incertitude et donc les risques dans les projets. Les méthodologies de retour d’expériences sont ainsi devenues réellement reconnues dans de nombreuses entreprises. Toutefois, ces entreprises se sont souvent contentées de recueillir des informations en fin de projet en pensant que cela suffirait à générer les connaissances nécessaires pour améliorer leur performance. Malheureusement, la capitalisation de ces expériences est traditionnellement une étape statique à la clôture du projet qui permet simplement la capture de quelques événements mémorisés par les experts impliqués. En outre, l’information capitalisée est difficile à réutiliser directement dans une nouvelle analyse des risques. Nos efforts se sont concentrés sur l’élaboration d’une méthode ayant pour but d’améliorer le processus de gestion des risques dans les projets en utilisant un système de retour d’expérience et, ainsi, de contribuer à une logique d’amélioration continue. La méthode envisagée est basée sur un système d’exploitation des connaissances qui permet de développer des compétences clés comme : la résolution des problèmes, la capacité à prendre des décisions collectivement, la réflexion, l’apprentissage et la capacité prospective (envisager). Le modèle proposé est défini sur la base de l’étude de la contribution de trois domaines centraux : le Management de Projets, le Management des Risques et le retour d’expérience. La singularité de nos travaux réside dans le fait d’intégrer explicitement un mécanisme de retour d’expérience en continu afin d’améliorer la performance des processus de gestion des risques dans les projets. Nous proposons de mettre en oeuvre une démarche de modélisation des connaissances orientée retour d’expérience. Nous définissons un modèle permettant de caractériser les projets, les risques et les expériences en vue du repérage, de la capitalisation et de l’exploitation des connaissances. La modélisation de ces éléments permet également de disposer d’une structuration facilitant une lecture plus rapide du projet ou de ses processus. Nous avons donc développé un modèle qui permet de représenter les éléments clé utilisés lors du processus de management des risques et ainsi de faciliter la capitalisation les expériences et la recherche des expériences antérieures similaires. Cela permet ensuite de standardiser et d’améliorer la démarche de management des risques. Du point de vue de la recherche des expériences antérieures, nous définissons : D’une part, un algorithme de recherche des expériences similaires fondé sur les comparaisons entre des graphes étiquetés orientés. Cela consiste en l’établissement d’une correspondance entre deux graphes (le graphe 1 étant le contexte du projet actuel et le graphe 2 le contexte d’un projet passé enregistré sous forme d’expérience) par une factorisation par paire de la matrice d’affinité qui découple la structure des noeuds et des arcs similaires (Zhou et De la Torre, 2012). D’autre part, un algorithme pour trouver la correspondance optimale entre ces graphes, de sorte que la somme de la compatibilité des noeuds et des arcs soit maximisée. Pour ce faire, nous avons utilisé un algorithme génétique. Enfin, nous proposons une démarche d’exploitation des expériences passées similaires. De cette manière nous pouvons obtenir un ensemble de risques associés à ces objets similaires afin d’alimenter le système d’aide à la décision dans la gestion du projet. / In order to improve Project Risk Management and to reinforce its effectiveness, the managers of the project should reuse the experiences and the good practices acquired in previous projects. The Experience Management constitutes an important source of knowledge to reduce the levels of uncertainty and thus the risks in the projects. The methodologies for Experience Management have thus become truly recognized in many companies. However, these companies were often satisfied to collect information at the end of the project by thinking that would be enough to generate knowledge necessary to improve their performance. Unfortunately, the capitalization of these experiments is traditionally a static stage at the close of the project, which simply allows the capture of some events memorized by the implied experts. Moreover, capitalized information is difficult to reuse directly in a new-risk analysis. Our efforts focused on developing a methodology to improve the Risk Management Process in Projects using an Experience Management System and thus contribute to the logic of continuous improvement. The proposed approach is based on a knowledge-based system that enables the development of key competencies such as problem solving, collective decision-making, reflection, learning and prospective capacity. The proposed model is based on the study of the contribution of three central fields: Projects Management, Risks Management and the Experience Management. The singularity of our work lies in the fact of integrating a mechanism back from experiment explicitly uninterrupted in order to improve the performance of the processes of risk management in the projects. We propose to implement a knowledge-based modeling approach based on Experience management. We define a model to characterize projects, risks and experiences for the identification, capitalization and exploitation of knowledge. The modeling of these elements also makes it possible to have a structuring facilitating a faster reading of the project or its processes. Therefore, we have developed a model which allows us to represent the key elements used during the risk management process and thus to facilitate the capitalization of the experiments and the research of previous similar experiences. That then makes it possible to standardize and improve the risk management approach. From the point of view of the search for previous experiences, we define: On the one hand, an algorithm of research of similar experiments founded on the comparisons between directed labeled graphs. That consists of the establishment of a correspondence between two graphs (graph 1 being context of the current project and graph 2 context of the last project recorded in the form of experiment) by a pairwise factorization of the affinity matrix, which decouples the structure of nodes and similar arcs (Zhou and Of Torre, 2012). On the other hand, an algorithm to find the optimal correspondence between these graphs, so that the sum of the compatibility of the nodes and the arcs is maximized. With this intention, we used a genetic algorithm. Lastly, we propose an approach of exploitation of similar last experiences. In this manner, we can obtain a set of risks associated with these similar objects in order to feed the assistance system with the decision in the management of the project.

As metodologias tradicionais de gerenciamento de projetos são caracterizadas como sendo rígidas e adequadas apenas para ambientes de pouca incerteza. No entanto, atualmente cada vez mais projetos são desenvolvidos em ambientes de alta complexidade e maiores incertezas, o que requer formas diferentes de gerenciamento do projeto, menos rígidas e mais flexíveis. Assim, o objetivo deste trabalho é propor um quadro conceitual para o gerenciamento de incertezas em projetos complexos. A abordagem metodológica mescla revisão sistemática da literatura e estudos de caso. O quadro conceitual desenvolvido, baseado na teoria da contingência, sugere que as abordagens de gestão de projetos sob incertezas são, em parte, determinadas pelas características das incertezas existentes. As respostas às incertezas podem ser orientadas à causa ou ao efeito da incerteza, e essas são escolhidas de acordo com a habilidade de influência sobre a causa, que aumenta para incertezas internas e diminui para incertezas externas. Já a flexibilidade da abordagem de gestão de projetos é impactada pelo grau da incerteza. Foram realizados seis estudos de caso, e um caso piloto, em projetos complexos, que mostraram que existe relação entre a origem da incerteza e a habilidade de influência, e entre a habilidade de influência e a orientação das ações. Eles mostraram também como a flexibilidade é importante para o gerenciamento das incertezas do projeto, associando métodos de antecipação, de instrucionismo, como o gerenciamento de riscos e a resiliência, principalmente a capacidade da alta direção assumir a existência de incertezas. Os estudos de caso evidenciaram que quanto maior o nível de incerteza do projeto, maior é a flexibilidade necessária. A dissertação apresenta ainda as limitações do estudo e as sugestões para trabalhos futuros. / Traditional project management methodologies are considered rigid and suitable only for environments of little uncertainty. However nowadays increasing number of projects is being developed in high complexity and uncertainty environment, requiring different approaches for project management: less rigid and more flexible. Thus, the purpose of this paper is to present a framework for managing projects under uncertainties, through a systematic literature review. The developed framework, based on the contingency theory, suggests that approaches for project management under uncertainty are, in part, determined by the characteristics of the existing uncertainties. The responses for uncertainty can be driven by the cause or consequence of the uncertainties and those are chosen according to the ability to influence the cause, which is higher for internal uncertainties and lower for external uncertainties. The flexibility of the project management approach, in its turn, is impacted by the uncertainty degree. Six case studies and a pilot case study were conducted in projects with complexity. The cases showed that there is a relation between the source of the uncertainty and the ability to influence, and between the ability to influence and the orientation of the responses. They also pointed out the importance of flexibility to manage uncertainties, combining prediction methods, the instructionism, such as using risk management, and resilience, especially the ability of the management to admit the existence of uncertainties. The cases also indicated that the highest is the project uncertainty, the highest is the need for flexibility. The paper also presents the limitation of the research, and suggestions for further works.

As metodologias tradicionais de gerenciamento de projetos são caracterizadas como sendo rígidas e adequadas apenas para ambientes de pouca incerteza. No entanto, atualmente cada vez mais projetos são desenvolvidos em ambientes de alta complexidade e maiores incertezas, o que requer formas diferentes de gerenciamento do projeto, menos rígidas e mais flexíveis. Assim, o objetivo deste trabalho é propor um quadro conceitual para o gerenciamento de incertezas em projetos complexos. A abordagem metodológica mescla revisão sistemática da literatura e estudos de caso. O quadro conceitual desenvolvido, baseado na teoria da contingência, sugere que as abordagens de gestão de projetos sob incertezas são, em parte, determinadas pelas características das incertezas existentes. As respostas às incertezas podem ser orientadas à causa ou ao efeito da incerteza, e essas são escolhidas de acordo com a habilidade de influência sobre a causa, que aumenta para incertezas internas e diminui para incertezas externas. Já a flexibilidade da abordagem de gestão de projetos é impactada pelo grau da incerteza. Foram realizados seis estudos de caso, e um caso piloto, em projetos complexos, que mostraram que existe relação entre a origem da incerteza e a habilidade de influência, e entre a habilidade de influência e a orientação das ações. Eles mostraram também como a flexibilidade é importante para o gerenciamento das incertezas do projeto, associando métodos de antecipação, de instrucionismo, como o gerenciamento de riscos e a resiliência, principalmente a capacidade da alta direção assumir a existência de incertezas. Os estudos de caso evidenciaram que quanto maior o nível de incerteza do projeto, maior é a flexibilidade necessária. A dissertação apresenta ainda as limitações do estudo e as sugestões para trabalhos futuros. / Traditional project management methodologies are considered rigid and suitable only for environments of little uncertainty. However nowadays increasing number of projects is being developed in high complexity and uncertainty environment, requiring different approaches for project management: less rigid and more flexible. Thus, the purpose of this paper is to present a framework for managing projects under uncertainties, through a systematic literature review. The developed framework, based on the contingency theory, suggests that approaches for project management under uncertainty are, in part, determined by the characteristics of the existing uncertainties. The responses for uncertainty can be driven by the cause or consequence of the uncertainties and those are chosen according to the ability to influence the cause, which is higher for internal uncertainties and lower for external uncertainties. The flexibility of the project management approach, in its turn, is impacted by the uncertainty degree. Six case studies and a pilot case study were conducted in projects with complexity. The cases showed that there is a relation between the source of the uncertainty and the ability to influence, and between the ability to influence and the orientation of the responses. They also pointed out the importance of flexibility to manage uncertainties, combining prediction methods, the instructionism, such as using risk management, and resilience, especially the ability of the management to admit the existence of uncertainties. The cases also indicated that the highest is the project uncertainty, the highest is the need for flexibility. The paper also presents the limitation of the research, and suggestions for further works.

Le but de cette recherche est de décrire et de comprendre le contrôle du transfert de Bonnes Pratiques entre différentes agences réparties sur le territoire. Cette étude émerge de l’observation d’un problème terrain : à partir d’un jeu de pratiques environnementales et de maitrises des risques dites « idéales » et sélectionnées pour avoir déjà prouvées leur efficacité dans d’autres unités organisationnelles, les membres du Comité de Pilotage transforment ces pratiques, tout au long des réunions du projet, de façon à permettre leur implémentation. Mais ces pratiques dépassent la simple application de prescriptions règlementaires et sont souvent des innovations pour l’entreprise. Ces innovations engagent le Comité de pilotage dans un véritable processus de conception. La finalité du projet est d’autonomiser les opérateurs dans l’usage de ces pratiques qui doivent pouvoir appréhender le terrain, adapter leur réponse tout en se conformant à l’idéal porté par la pratique. Notre hypothèse de travail est que l’action du Comité est mue par une logique de contrôle. Dans ce travail, nous cherchons à caractériser la façon dont le Comité de pilotage de ce projet contrôle le déploiement de ces Bonnes Pratiques et à rendre compte de sa dynamique. Pour répondre à cette question, cette étude conduite au sein de la filiale régionale d’un groupe s’inscrit dans une approche qualitative. Notre méthode d’analyse s’appuie sur le recours à la théorie ancrée en se fondant sur les éléments empiriques du terrain. Pour caractériser le processus de contrôle, nous avons voulu échapper aux catégories usuelles des mécanismes de contrôle de la théorie (Ouchi, 1979 ; Simons, 2004) en expliquant comment l’acteur collectif Comité de pilotage, par des actions sur le processus de conception des pratiques, par des actions d’information et de formation des agents et par des actions sur la rédaction de fiches de procédures, cherchait à s’assurer que chaque membre du Comité de pilotage et les agents impliqués dans ce processus de conception se comportent dans le sens des objectifs fixés par la Direction générale. Ainsi, parmi nos contributions, le niveau de granularité plus fin dans l’analyse du contrôle nous a permis d’apporter une relecture des travaux traditionnels en contrôle. Pour rendre compte de la dynamique du contrôle, nous procédons également à une opérationnalisation du modèle de la réplication de Winter et Szulanski (2001). En complément, nos résultats illustrent la dynamique du contrôle fondée sur l’analyse de la convergence des projets par Midler (1993). / The purpose of this study is to describe and understand how the transfer of the Good Practices among the different agencies spread out in the territory is controlled. This research emerges from a problem observed in practice : based on a set of environmental and risk-management practices which were said to be « ideal » and which had proved to be efficient, the members of the Piloting Committee transform these practices throughout their project meetings so that they may be implemented. Those practices however, go beyond statutory requirements, and often turn out to be innovations for companies. Those innovations engage the Piloting Committee in a true conception process. The project aims at making the users/operators autonomous. They are to be able to apprehend the situation and adapt their response, all the while complying with the fundamental ideal of the Practice. Our working hypothesis is that the action of the Piloting Committee is driven by logic of control. In this work, we are seeking to characterize the way the Piloting Committee of this project controls the way in which the Good Practices are implemented and to report their dynamics. To investigate these research questions, a qualitative study was designed and conducted. We use grounded theory methodology. This intervention-research led in a subsidiary company. To characterize the control process, we chose to avoid the usual categories of mechanical control of the theory (Ouchi, 1979; Simons, 2004). Instead, we chose to explain how the collective actor Piloting Committee, by setting up information and training campaigns for all the Piloting Committee members and at the actors involved in the conception process, sought to ensure that each and every one was indeed acting consistently and in compliance with the targets set by the General Management. We are procceeding to the operationalization of the replication of Winter and Szulanski model (2001). Our results show Midler’s work (1993).

Este trabalho trata dos temas sustentabilidade, gestão de projetos e sucesso em projetos. Por abordá-los de forma conjunta, preenche uma lacuna de pesquisa que trata da conexão entre os três temas. O objetivo da tese, desenvolvida por artigos, é propor um modelo teórico e empírico de sustentabilidade aplicável no contexto de gestão de projetos e orientado para o sucesso em projetos. Como metodologia fez-se uso da abordagem combinada, envolvendo pesquisa qualitativa e quantitativa com aplicação de multi métodos. A pesquisa qualitativa foi utilizada para a sistematização da literatura dos temas chaves e para o desenvolvimento de estudos de casos múltiplos com aplicação de entrevistas com gestores de projetos e de sustentabilidade, bem como aplicação de questionário aos membros de equipes de projetos. A pesquisa quantitativa foi utilizada para a realização de levantamento survey, com 143 gestores de projetos em empresas brasileiras, e análise estatística por meio da modelagem de equações estruturais, bem como foi aplicado o método de análise fatorial exploratória. Os constructos e variáveis do modelo teórico proposto, bem como os protocolos de pesquisa survey, foram validados por meio de um fórum com 11 especialistas oriundos de 10 universidades, localizadas em seis países diferentes. Como contribuições deste estudo pode-se destacar a estruturação do referencial teórico de sustentabilidade em gestão de projetos e de sucesso em projetos; a apresentação e validação da estrutura de um modelo teórico e empírico de constructos e variáveis de sustentabilidade com enfoque Triple-Bottom Line - TBL (econômica, ambiental e social) em gestão de projetos orientado ao sucesso em projetos; e, a validação estatística do relacionamento entre sustentabilidade em gestão de projetos e sucesso em projetos. Esta última relação foi confirmada pela alta significância da relação causal entre os constructos, pois sustentabilidade em gestão de projetos contribui positivamente para o sucesso em projetos, explicando 43% dos efeitos sobre sucesso em projetos. Em decorrência, foram desenvolvidos seis artigos científicos, alguns publicados e outros submetidos para periódicos internacionais. / This research looks to study sustainability, Project management and project success. Merging these themes, this thesis works on a gap of research related to approaching both topics. For that matter, this study, developed by articles model, aims to propose a theoretical and empirical framework of sustainability, applicable into project management context and oriented to project success. The combined approach was used as methodology, engaging qualitative and quantitative research by applying multi methods. Qualitative research was used to systematize the literature of key themes and to develop multiple case studies by applying semi-structured interviews with people responsible for projects and sustainability areas and questionnaire application to projects team members. Quantitative research was used to develop a survey involving 143 project managers in Brazilian companies. Statistics analysis was carried out in order to treat the data through structural equation modeling. Exploratory factor analysis was also used. Constructs and variables of the conceptual framework proposed, as well as the survey research protocols, were validated by experts through and expert panel of eleven experts from ten universities and six different countries. Contributions of this study may highlight the development of the bibliographic review of sustainability in project management and project success topics; the presentation and validation of the conceptual and empirical model of constructs and variables of sustainability (with triple-bottom line perspective) in project management oriented to project success; and the statistical validation of the relationship between sustainability in project management and project success. This relationship was confirmed by the high significance of the casual relation between both constructs, explaining that the constructs of sustainability in project management contributes positively to obtain project success. It was found that 43% of these constructs affect project success. At the end, six scientific articles were developed, and some of them were already published and others submitted to international journals.

Este trabalho trata dos temas sustentabilidade, gestão de projetos e sucesso em projetos. Por abordá-los de forma conjunta, preenche uma lacuna de pesquisa que trata da conexão entre os três temas. O objetivo da tese, desenvolvida por artigos, é propor um modelo teórico e empírico de sustentabilidade aplicável no contexto de gestão de projetos e orientado para o sucesso em projetos. Como metodologia fez-se uso da abordagem combinada, envolvendo pesquisa qualitativa e quantitativa com aplicação de multi métodos. A pesquisa qualitativa foi utilizada para a sistematização da literatura dos temas chaves e para o desenvolvimento de estudos de casos múltiplos com aplicação de entrevistas com gestores de projetos e de sustentabilidade, bem como aplicação de questionário aos membros de equipes de projetos. A pesquisa quantitativa foi utilizada para a realização de levantamento survey, com 143 gestores de projetos em empresas brasileiras, e análise estatística por meio da modelagem de equações estruturais, bem como foi aplicado o método de análise fatorial exploratória. Os constructos e variáveis do modelo teórico proposto, bem como os protocolos de pesquisa survey, foram validados por meio de um fórum com 11 especialistas oriundos de 10 universidades, localizadas em seis países diferentes. Como contribuições deste estudo pode-se destacar a estruturação do referencial teórico de sustentabilidade em gestão de projetos e de sucesso em projetos; a apresentação e validação da estrutura de um modelo teórico e empírico de constructos e variáveis de sustentabilidade com enfoque Triple-Bottom Line - TBL (econômica, ambiental e social) em gestão de projetos orientado ao sucesso em projetos; e, a validação estatística do relacionamento entre sustentabilidade em gestão de projetos e sucesso em projetos. Esta última relação foi confirmada pela alta significância da relação causal entre os constructos, pois sustentabilidade em gestão de projetos contribui positivamente para o sucesso em projetos, explicando 43% dos efeitos sobre sucesso em projetos. Em decorrência, foram desenvolvidos seis artigos científicos, alguns publicados e outros submetidos para periódicos internacionais. / This research looks to study sustainability, Project management and project success. Merging these themes, this thesis works on a gap of research related to approaching both topics. For that matter, this study, developed by articles model, aims to propose a theoretical and empirical framework of sustainability, applicable into project management context and oriented to project success. The combined approach was used as methodology, engaging qualitative and quantitative research by applying multi methods. Qualitative research was used to systematize the literature of key themes and to develop multiple case studies by applying semi-structured interviews with people responsible for projects and sustainability areas and questionnaire application to projects team members. Quantitative research was used to develop a survey involving 143 project managers in Brazilian companies. Statistics analysis was carried out in order to treat the data through structural equation modeling. Exploratory factor analysis was also used. Constructs and variables of the conceptual framework proposed, as well as the survey research protocols, were validated by experts through and expert panel of eleven experts from ten universities and six different countries. Contributions of this study may highlight the development of the bibliographic review of sustainability in project management and project success topics; the presentation and validation of the conceptual and empirical model of constructs and variables of sustainability (with triple-bottom line perspective) in project management oriented to project success; and the statistical validation of the relationship between sustainability in project management and project success. This relationship was confirmed by the high significance of the casual relation between both constructs, explaining that the constructs of sustainability in project management contributes positively to obtain project success. It was found that 43% of these constructs affect project success. At the end, six scientific articles were developed, and some of them were already published and others submitted to international journals.