Cette thèse traite de l'analyse de nuages de points 3D désorganisé et s'appuie sur deux outils : un algorithme de partitionnement efficace inspiré des C-moyennes floues d'une part, et des outils de filtrage morphologique de représentation à base de triangulation de Delaunay d'autre part. Le cadre applicatif essentiel est la navigation autonome en robotique mobile en environnement inconnu, c'est-à-dire dans modèle. Mais la méthodologie générique développée a été appliquée à d'autres types d'environnements, notamment plus structurés.

The goal of this work is to describe a robotics platform called the Building Wide Intelligence Segbot (segbot). The segbot is a two wheeled robot that can robustly navigate our building, perform obstacle avoidance, and reason about the world. This work has two main goals. First we introduce the segbot platform to anyone that may use it in the future. We begin by examining off-the-shelf components we used and how to build a robot that is able to navigate in a complex multi-floor building environment with moving obstacles. Then we explain the software from a top down viewpoint, with a three layer abstraction model for segmenting code on any robotics platform. The second part of this document describes current work on the segbot platform, which is able to non-robustly take requests for coffee and navigate to a coffee shop while having to move across multiple floors in a building. My contribution to this work is building an infrastructure for multi-floor navigation. The multi-floor infrastructure built is non-robust but has helped identify several issues that will need to be tackled in future iterations of the segbot. / text

This dissertation details the design and development of a mobile autonomous platform from which to conduct a 3D ground-penetrating radar survey. The system uses a three-rover multiagent team to perform a site-selection activity during a lunar analog mission. The work took place beginning in 2008 and culminated in a final field test on Mauna Kea in Hawaii. This demonstration of autonomous acquisition of 3D ground-penetrating radar in a space robtic application is promising not only for in-situ resource utilization, but also for the concept of multiagent teaming.

This dissertation details the design and development of a mobile autonomous platform from which to conduct a 3D ground-penetrating radar survey. The system uses a three-rover multiagent team to perform a site-selection activity during a lunar analog mission. The work took place beginning in 2008 and culminated in a final field test on Mauna Kea in Hawaii. This demonstration of autonomous acquisition of 3D ground-penetrating radar in a space robtic application is promising not only for in-situ resource utilization, but also for the concept of multiagent teaming.

Conselho Nacional de Desenvolvimento Científico e Tecnológico / Considering path planning problem for autonomous robots, the objective is to find a list of steps to be applied to obtain a path between the initial point and the goal. This work aims the investigation and implementation of cellular automata (CA) based models to path-planning. In an initial phase, a comparative study was conducted among the cellular automata-based methods to path-planning published in the literature. Subsequently, two published works were chosen to be implemented in simulation environments to verify the actual applicability of the proposed methods. The first model starts from an image captured from the environment and it applies a CA to perform the calculation of distances between free cells and the goal. The second model uses robot sensors to identify its neighborhood and it applies CA transition rules to determine the next movements. Some limitations which prevented the robots obtain good results in simulation were identified and improvements to the original models were applied. At the end, both new models exhibited better behaviors than their precursors in several scenarios. Aiming to validate our results, two simulation environments were employed (V-REP and Webots) and some experiments with e-puck robots were performed. / No problema do planejamento de caminhos para robôs autônomos, o objetivo é encontrar uma lista de passos a serem aplicados para se obter um caminho entre o ponto inicial e a meta. Este trabalho visa a investigação e implementação de modelos baseados em autômatos celulares (ACs) para o planejamento de caminhos. Em uma fase inicial, foi realizado um estudo comparativo entre os métodos de planejamento de caminhos baseados em autômatos celulares publicados na literatura. Posteriormente, foram escolhidos dois trabalhos publicados que foram implementados em ambientes de simulação para se verificar a real aplicabilidade dos métodos propostos. O primeiro modelo parte de uma imagem capturada do ambiente de navegação e utiliza um AC para fazer o cálculo das distâncias entre as células livres e a meta. O segundo modelo utiliza os sensores do robô para identificar sua vizinhança a cada instante e utiliza regras de transição de ACs para determinar os próximos movimentos. Algumas limitações que impossibilitaram que os robôs obtivessem bons resultados em simulação foram identificadas e melhorias foram aplicadas aos modelos originais. Ao final, os dois novos modelos propostos exibiram um melhor desempenho do que seus precursores em diversos cenários. Para validar nossos resultados, dois ambientes de simulação foram empregados (V-REP e Webots), além da execução de alguns experimentos com robôs e-puck. / Mestre em Ciência da Computação