A aplicação de técnicas baseadas em amostragem em algoritmos que envolvem o planejamento de trajetórias de robôs tem se tornado cada vez mais difundida. Deste grupo, um dos algoritmos mais utilizados é chamado Rapidly-exploring Random Tree (RRT), que se baseia na amostragem incremental para calcular de forma eficiente os planos de trajetória do robô evitando colisões com obstáculos. Vários esforços tem sido realizados a fim de reduzir o custo computacional do algoritmo RRT, visando aplicações que necessitem de respostas mais rápidas do algoritmo, como, por exemplo, em ambientes dinâmicos. Um dos dilemas relacionados ao RRT está na etapa de geração de primitivas de movimento. Se várias primitivas são geradas, permitindo o robô executar vários movimentos básicos diferentes, um grande custo computacional é gasto. Por outro lado, quando poucas primitivas são geradas e, consequentemente, poucos movimentos básicos são permitidos, o robô pode não ser capaz de encontrar uma solução para o problema, mesmo que esta exista. Motivados por este problema, um método de geração de primitivas de movimento foi proposto. Tal método é comparado com os métodos tradicional e aleatório de geração de primitivas, considerando não apenas o custo computacional de cada um, mas também a qualidade da solução obtida. O método proposto é aplicado ao algoritmo RRT, que depois é aplicado em um caso de estudo em um ambiente dinâmico. No estudo de caso, o algoritmo RRT otimizado é avaliado em termos de seus custos computacionais durante planejamentos e replanejamento de trajetória. As simulações são realizadas em dois simuladores: um desenvolvido em linguagem Python e outro em Matlab. / The application of sample-based techniques in path-planning algorithms has become year-by-year more widespread. In this group, one of the most widely used algorithms is the Rapidly-exploring Random Tree (RRT), which is based on an incremental sampling of configurations to efficiently compute the robot\'s path while avoiding obstacles. Many efforts have been made to reduce RRT computational costs, targeting, in particular, applications in which quick responses are required, e.g., in dynamic environments. One of the dilemmas posed by the RRT arises from its motion primitives generation. If many primitives are generated to enable the robot to perform a broad range of basic movements, a signicant computational cost is required. On the other hand, when only a few primitives are generated, thus, enabling a limited number of basic movements, the robot may be unable to find a solution to the problem, even if one exists. To address this quandary, an optimized method for primitive generation is proposed. This method is compared with the traditional and random primitive generation methods, considering not only computational cost, but also the quality of local and global solutions that may be attained. The optimized method is applied to the RRT algorithm, which is then used in a case study in dynamic environments. In the study, the modied RRT is evaluated in terms of the computational costs of its planning and replanning. The simulations were developed to access the effectiveness and efficiency of the proposed algorithm.

Na área de robótica móvel autônoma é importante que o robô siga uma trajetória livre de obstáculos. Estes podem ser desde obstáculos estáticos, como paredes e cadeiras em um ambiente interno, ou mesmo obstáculos móveis, como pessoas caminhando na calçada e carros passando pela rua, quando consideramos ambientes externos. No caso de um ambiente estático, o problema pode ser resolvido planejando uma trajetória livre de colisões, sendo que não é necessário um replanejamento se todos os obstáculos estáticos foram considerados. Para ambientes onde os obstáculos estão em constante movimento, é necessário um constante replanejamento da trajetória para que se evite colisões. Alternativamente, pode ser verificada a possibilidade de se manter na rota planejada, alterando apenas a velocidade de cruzeiro do robô para que este desvie dos obstáculos móveis. Este trabalho propõe uma metodologia para calcular uma velocidade de cruzeiro para o robô de forma a minimizar a probabilidade de colisão com os obstáculos detectados pelos seus sensores. A escolha da variação de velocidade para o robô considera a sua velocidade atual, e as velocidades estimadas para os obstáculos. A metodologia para resolução deste problema é apresentada considerando incertezas na posição do robô e obstáculos. São apresentados resultados de simulação que exemplificam a aplicação da metodologia. / Following a free path is an important issue in the area of autonomous mobile robotics. The obstacles can be anything from walls and chairs in an indoor environment, or they can also be people walking on the sidewalk and cars moving through the street. In the case of a static environment, the problem can be solved by planning a path free from collisions, thus it is not essential another path planning as all static obstacles were considered. For an environment were the obstacles are constantly moving, it is necessary an unceasing path replanning to avoid possible collisions. Alternatively, keeping the robot on the previously calculated path can be verified modifying the robot\'s traffic velocity to avoid moving obstacles. Our proposal is to calculate a velocity for the robot which minimizes its collision probability with moving obstacles detected by its sensors. Varying the robot\'s velocity takes into account its current velocity and the estimated velocities of obstacles. The methodology for solving this problem is presented regarding uncertainties in robots and obstacles\' positions. Results from simulations that exemplifies an application for the methodology are presented.

Na área de robótica móvel autônoma é importante que o robô siga uma trajetória livre de obstáculos. Estes podem ser desde obstáculos estáticos, como paredes e cadeiras em um ambiente interno, ou mesmo obstáculos móveis, como pessoas caminhando na calçada e carros passando pela rua, quando consideramos ambientes externos. No caso de um ambiente estático, o problema pode ser resolvido planejando uma trajetória livre de colisões, sendo que não é necessário um replanejamento se todos os obstáculos estáticos foram considerados. Para ambientes onde os obstáculos estão em constante movimento, é necessário um constante replanejamento da trajetória para que se evite colisões. Alternativamente, pode ser verificada a possibilidade de se manter na rota planejada, alterando apenas a velocidade de cruzeiro do robô para que este desvie dos obstáculos móveis. Este trabalho propõe uma metodologia para calcular uma velocidade de cruzeiro para o robô de forma a minimizar a probabilidade de colisão com os obstáculos detectados pelos seus sensores. A escolha da variação de velocidade para o robô considera a sua velocidade atual, e as velocidades estimadas para os obstáculos. A metodologia para resolução deste problema é apresentada considerando incertezas na posição do robô e obstáculos. São apresentados resultados de simulação que exemplificam a aplicação da metodologia. / Following a free path is an important issue in the area of autonomous mobile robotics. The obstacles can be anything from walls and chairs in an indoor environment, or they can also be people walking on the sidewalk and cars moving through the street. In the case of a static environment, the problem can be solved by planning a path free from collisions, thus it is not essential another path planning as all static obstacles were considered. For an environment were the obstacles are constantly moving, it is necessary an unceasing path replanning to avoid possible collisions. Alternatively, keeping the robot on the previously calculated path can be verified modifying the robot\'s traffic velocity to avoid moving obstacles. Our proposal is to calculate a velocity for the robot which minimizes its collision probability with moving obstacles detected by its sensors. Varying the robot\'s velocity takes into account its current velocity and the estimated velocities of obstacles. The methodology for solving this problem is presented regarding uncertainties in robots and obstacles\' positions. Results from simulations that exemplifies an application for the methodology are presented.

A aplicação de técnicas baseadas em amostragem em algoritmos que envolvem o planejamento de trajetórias de robôs tem se tornado cada vez mais difundida. Deste grupo, um dos algoritmos mais utilizados é chamado Rapidly-exploring Random Tree (RRT), que se baseia na amostragem incremental para calcular de forma eficiente os planos de trajetória do robô evitando colisões com obstáculos. Vários esforços tem sido realizados a fim de reduzir o custo computacional do algoritmo RRT, visando aplicações que necessitem de respostas mais rápidas do algoritmo, como, por exemplo, em ambientes dinâmicos. Um dos dilemas relacionados ao RRT está na etapa de geração de primitivas de movimento. Se várias primitivas são geradas, permitindo o robô executar vários movimentos básicos diferentes, um grande custo computacional é gasto. Por outro lado, quando poucas primitivas são geradas e, consequentemente, poucos movimentos básicos são permitidos, o robô pode não ser capaz de encontrar uma solução para o problema, mesmo que esta exista. Motivados por este problema, um método de geração de primitivas de movimento foi proposto. Tal método é comparado com os métodos tradicional e aleatório de geração de primitivas, considerando não apenas o custo computacional de cada um, mas também a qualidade da solução obtida. O método proposto é aplicado ao algoritmo RRT, que depois é aplicado em um caso de estudo em um ambiente dinâmico. No estudo de caso, o algoritmo RRT otimizado é avaliado em termos de seus custos computacionais durante planejamentos e replanejamento de trajetória. As simulações são realizadas em dois simuladores: um desenvolvido em linguagem Python e outro em Matlab. / The application of sample-based techniques in path-planning algorithms has become year-by-year more widespread. In this group, one of the most widely used algorithms is the Rapidly-exploring Random Tree (RRT), which is based on an incremental sampling of configurations to efficiently compute the robot\'s path while avoiding obstacles. Many efforts have been made to reduce RRT computational costs, targeting, in particular, applications in which quick responses are required, e.g., in dynamic environments. One of the dilemmas posed by the RRT arises from its motion primitives generation. If many primitives are generated to enable the robot to perform a broad range of basic movements, a signicant computational cost is required. On the other hand, when only a few primitives are generated, thus, enabling a limited number of basic movements, the robot may be unable to find a solution to the problem, even if one exists. To address this quandary, an optimized method for primitive generation is proposed. This method is compared with the traditional and random primitive generation methods, considering not only computational cost, but also the quality of local and global solutions that may be attained. The optimized method is applied to the RRT algorithm, which is then used in a case study in dynamic environments. In the study, the modied RRT is evaluated in terms of the computational costs of its planning and replanning. The simulations were developed to access the effectiveness and efficiency of the proposed algorithm.

Purpose – The presented research aims to explain, describe and analyze the process of dynamic strategy development in high growth firms. Accordingly the research seeks to investigate how dynamic strategies are used within high growth firms and how strategic learning affects the process. Design/methodology/approach – The authors presents a review of theoretically relevant studies of high growth related to strategy, and two original studies examining the impact of dynamic strategy on high growth. A theoretical framework for the study of dynamic strategy processes is developed. The study comprehends eight interviews divided over five high growth firms, where high growth is defined by the OECD (2008) standard. Findings – In the study, the researchers finds significant evidence for the active and deliberate use of dynamic strategy in the high growth firms of the study. The implication of strategic learning on the dynamic strategies is found to be substantial. The findings shows that dynamic strategy development are used to a large extent and considered vital for achieving growth within in the high growth firms of the study. Research/theoretical implications/limitations – The findings demonstrate that dynamic strategy development is actively used in high growth phases of the firms studied. This has implications on the extension of previous research, as it shows the actual use of dynamic strategy and further emphasizes the importance of strategic learning within this process. With the important limitation that the study is considered too small to generalize over a larger population, which implies that further research on the subject is needed. Managerial implications – The findings provide guidelines for managers of how to handle strategy development in high growth, however due to the previous limitation this is presented as the way the high growth firms within this study handles this development. The guidelines could be used by anyone in managerial positions, thus increasing the understanding of how high growth firms handle strategy.

Cette thèse explore la façon dont les managers suscitent l’engagement des membres au sein d’équipes ouvertes, internationales et virtuelles. Ce type d’équipe, qu’on nomme une équipe « ouverte » est caractérisée par : des interactions régulières entre les membres, pour atteindre des objectifs organisationnels spécifiques ; une appartenance floue et instable des membres ; une « pluri-appartenance » des membres : chaque membre est simultanément impliqué dans plusieurs équipes ; des managers disposant d’une responsabilité formelle, sans pour autant disposer d’une autorité formelle. Les équipes ouvertes sont un phénomène fréquent dans les organisations commerciales internationales et les entreprises de conseil, mais peu de recherches s’y sont encore intéressées. Mobilisant la théorie enracinée, cette recherche étudie les équipes ouvertes d’une grande entreprise multinationale. Sur une période de deux ans, 70 entretiens formels ont été conduits auprès de 66 employés et 28 jours d’observation ont été réalisés dans neuf pays. Cette étude enrichit la littérature sur les équipes du concept d’équipe ouverte et apporte des compléments à trois cadres théoriques : la théorie de l’engagement, du « boundary spanning » et du leadership fonctionnel. Elle prolonge la théorie de l’engagement de Kahn, en montrant comment les leaders améliorent les conditions psychologiques de l’engagement en donnant du sens et en agissant sur le sentiment de sécurité psychologique. La recherche enrichit le concept de « boundary work » en identifiant de manière empirique et théorique deux pratiques de « boundary spanning » supplémentaires : la création de frontières et la coordination des frontières. L’étude des équipes ouvertes a fait émerger six fonctions additionnelles à la théorie du leadership fonctionnel et montré que six autres fonctions sont modifiées du fait de la structure même de ces équipes ouvertes. / The objective of this empirical study is to explore how managers engage team members in global virtual ‘open’ teams. Open teams are a collection of individuals with varying levels of interdependence, who contribute toward shared organizational objectives, who are seen by others as an intact social entity within one or more larger systems, whose membership is neither bounded nor stable, and whose members participate in multi-team systems, and whose managers have formal responsibility for group outcomes but little to no formal authority. Open teams are a phenomenon found in multinationals and large consulting firms, but little research is focused on this type of team. Using grounded theory, this research examines the open teams at a large multinational. Over a period of two years, 70 formal interviews with 66 employees were conducted and 28 days of observations in nine countries were logged. The study introduces the concept of ‘open team’ and contributes to three theoretical frameworks: engagement theory, boundary spanning theory and functional leadership theory. It extends Kahn’s theory of engagement by showing how leaders influence meaningfulness and psychological safety to improve the psychological conditions for engagement. In boundary spanning theory, the boundary work concept is extended by providing empirical evidence for boundary coordination and boundary creation as well as introducing two new boundary spanning practices. In function leadership theory, six leadership functions emerge from the study of open teams and show how another six changed due to the open team structure.

La robotique mobile autonome est un axe de recherche qui vise à donner à une machine la capacité de se mouvoir dans un environnement sans assistance ni intervention humaine. Cette thèse s’intéresse à la partie décisionnelle de la navigation robotique à savoir la planification de mouvement pour un robot mobile non-holonome, pour lequel, la prise en compte des contraintes cinématiques et non-holonomes est primordiale. Aussi, la nécessité de considérer la géométrie propre du robot et la bonne maîtrise de l’environnement dans lequel il évolue constituent des contraintes à assurer. En effet la planification de mouvement consiste à calculer un mouvement réalisable que doit accomplir le robot entre une position initiale et une position finale données. Selon la nature de l’environnement, notamment les obstacles qui s’y présentent, deux instances du problème se distinguent : la planification de chemin et la planification de trajectoire. L’objectif de cette thèse est de proposer de nouveaux algorithmes pour contribuer aux deux instances du problème de planification de mouvement. La méthodologie suivie repose sur des solutions génériques qui s’appliquent à une classe de systèmes robotiques plutôt qu’à une architecture particulière. Les approches proposées intègrent les B-splines Rationnelles non uniformes (NURBS) dans le processus de modélisation des solutions générées tout en s’appuyant sur la propriété de contrôle local, et utilisent les algorithmes génétiques pour une meilleure exploration de l’espace de recherche. / The mobile robotics is an area of research that aims to give a machine the ability to move in an environment without assistance or human intervention. This thesis focuses on the decisional part of robotic navigation, namely motion planning for a non-holonomic mobile robot, for which, the consideration of kinematic and non-holonomic constraints is paramount. Also, the need to consider the specific geometry of the robot and the good control of the environment in which it operates are constraints to insure. Indeed, motion planning is to calculate a feasible movement to be performed by the robot between an initial and a final given position. Depending on the nature of the environment, two instances of the problem stand out: the path planning and the trajectory planning. The objective of this thesis is to propose new algorithms to contribute to the two instances of motion planning problem. The followed methodology is based on generic solutions that are applicable to a class of robotic systems rather than a particular architecture. The proposed approaches include the Non-Uniform Rational B-Spline (NURBS) in the modeling process of the generated solutions while relying on the local control property. Also, they use genetic algorithms for better exploration of the search space.

Classical value estimation reinforcement learning algorithms do not perform very well in dynamic environments. On the other hand, the reinforcement learning of animals is quite flexible: they can adapt to dynamic environments very quickly and deal with noisy inputs very effectively. One feature that may contribute to animals' good performance in dynamic environments is that they learn and perceive the time to reward. In this research, we attempt to learn and perceive the time to reward and explore situations where the learned time information can be used to improve the performance of the learning agent in dynamic environments. The type of dynamic environments that we are interested in is that type of switching environment which stays the same for a long time, then changes abruptly, and then holds for a long time before another change. The type of dynamics that we mainly focus on is the time to reward, though we also extend the ideas to learning and perceiving other criteria of optimality, e.g. the discounted return, so that they can still work even when the amount of reward may also change. Specifically, both the mean and variance of the time to reward are learned and then used to detect changes in the environment and to decide whether the agent should give up a suboptimal action. When a change in the environment is detected, the learning agent responds specifically to the change in order to recover quickly from it. When it is found that the current action is still worse than the optimal one, the agent gives up this time's exploration of the action and then remakes its decision in order to avoid longer than necessary exploration. The results of our experiments using two real-world problems show that they have effectively sped up learning, reduced the time taken to recover from environmental changes, and improved the performance of the agent after the learning converges in most of the test cases compared with classical value estimation reinforcement learning algorithms. In addition, we have successfully used spiking neurons to implement various phenomena of classical conditioning, the simplest form of animal reinforcement learning in dynamic environments, and also pointed out a possible implementation of instrumental conditioning and general reinforcement learning using similar models.

Recent developments in the computer vision field achieve state-of-the-art performance by utilizing large-scale training datasets and in the absence of that, generating synthetic datasets of said magnitude. Yet, for certain applications, it is not feasible to synthesize high fidelity training data (e.g., biomedical computer vision domain), or to achieve detailed explainability for the program's decisions. Formulating a part-based approach can help alleviate the aforementioned challenges as (i) a scene can naturally be decomposed into a hierarchical part-based structure, and (ii) using domain knowledge by incorporating the object parts' topological and geometrical constraints reduces the complexity of learning and inference, benefiting methods in terms of data efficiency and computational resources. This dissertation investigates multiple applications that benefit from a part-based solution regarding the applications' performance metrics and/or computational efficiency. We develop part-based methods for registration, segmentation, unsupervised object discovery in large-scale image collections, and unsupervised unknown foreground discovery in streaming scenarios.

[pt] Robôs móveis são cada dia mais importantes na sociedade moderna, realizando tarefas consideradas tediosas ou muito repetitivas para humanos, como limpeza ou patrulhamento. A maioria dessas tarefas requer um certo nível de autonomia do robô. Para que o robô seja considerado autônomo, ele precisa de um mapa do ambiente, e de sua posição e orientação nesse mapa. O problema de localização e mapeamento simultâneos (SLAM) é a tarefa de estimar tanto o mapa quanto a posição e orientação simultaneamente, usando somente informações dos sensores, sem ajuda externa. O problema de SLAM visual consiste na tarefa de realizar SLAM usando somente câmeras para o sensoriamento. A maior vantagem de usar câmeras é a possibilidade de resolver problemas de visão computacional que provêm informações de alto nível sobre a cena, como detecção de objetos. Porém a maioria dos sistemas de SLAM visual assume um ambiente estático, o que impõe limitações para a sua aplicabilidade em cenários reais. Esta tese apresenta soluções para o problema de SLAM visual em ambientes dinâmicos e em mudança. Especificamente, a tese propõe um método para ambientes com multidões, junto com um detector de pessoas customizado baseado em aprendizado profundo. Além disso, também é proposto um método de SLAM visual para ambientes altamente dinâmicos contendo objetos em movimento, combinando um rastreador de objetos robusto com um algoritmo de filtragem de pontos. Além disso, esta tese propõe um método de SLAM visual para ambientes em mudança, isto é, em cenas onde os objetos podem mudar de lugar após o robô já os ter mapeado. Todos os métodos propostos são testados com dados públicos e experimentos, e comparados com diversos métodos da literatura, alcançando um bom desempenho em tempo real. / [en] Mobile robots have become increasingly important in modern society, as they can perform tasks that are tedious or too repetitive for humans, such as cleaning and patrolling. Most of these tasks require a certain level of autonomy of the robot. To be fully autonomous and perform navigation, the robot needs a map of the environment and its pose within this map. The Simultaneous Localization and Mapping (SLAM) problem is the task of estimating both map and localization, simultaneously, only using sensor measurements. The visual SLAM problem is the task of performing SLAM only using cameras for sensing. The main advantage of using cameras is the possibility of solving computer vision problems that provide high-level information about the scene, such as object detection. However, most visual SLAM systems assume a static environment, which imposes a limitation on their applicability in real-world scenarios. This thesis presents solutions to the visual SLAM problem in dynamic and changing environments. A custom deep learning-based people detector allows our solution to deal with crowded environments. Also, a combination of a robust object tracker and a filtering algorithm enables our visual SLAM system to perform well in highly dynamic environments containing moving objects. Furthermore, this thesis proposes a visual SLAM method for changing environments, i.e., in scenes where the objects are moved after the robot has already mapped them. All proposed methods are tested in datasets and experiments and compared with several state-of-the-art methods, achieving high accuracy in real time.