<p>This thesis investigates the use of Genetic Programming (GP) to evolve controllers for an autonomous robot.</p><p>GP is a type of Genetic Algorithm (GA) using the Darwinian idea of natural selection and genetic recombination, where the individuals most often is represented as a tree-structure. The GP is used to evolve a population of possible solutions over many generations to solve problems.</p><p>The most common approach used today, to develop controllers for autonomous robots, is to employ a GA to evolve an Artificial Neural Network (ANN). This approach is most often used in simulation only or in conjunction with online evolution; where simulation still covers the largest part of the process.</p><p>The GP has been largely neglected in Behaviour Based Robotics (BBR). The is primarily due to the problem of speed, which is the biggest curse of any standard GP. The main contribution of this thesis is the approach of using a linear representation of the GP in online evolution, and to establish whether or not the GP is feasible in this situation. Since this is not a comparison with other methods, only a demonstration of the possibilities with GP, there is no need for testing the particular test cases with other methods.</p><p>The work in this thesis builds upon the work by Wolfgang Banzhaf and Peter Nordin, and therefore a comparison with their work will be done.</p>

This thesis investigates the use of Genetic Programming (GP) to evolve controllers for an autonomous robot. GP is a type of Genetic Algorithm (GA) using the Darwinian idea of natural selection and genetic recombination, where the individuals most often is represented as a tree-structure. The GP is used to evolve a population of possible solutions over many generations to solve problems. The most common approach used today, to develop controllers for autonomous robots, is to employ a GA to evolve an Artificial Neural Network (ANN). This approach is most often used in simulation only or in conjunction with online evolution; where simulation still covers the largest part of the process. The GP has been largely neglected in Behaviour Based Robotics (BBR). The is primarily due to the problem of speed, which is the biggest curse of any standard GP. The main contribution of this thesis is the approach of using a linear representation of the GP in online evolution, and to establish whether or not the GP is feasible in this situation. Since this is not a comparison with other methods, only a demonstration of the possibilities with GP, there is no need for testing the particular test cases with other methods. The work in this thesis builds upon the work by Wolfgang Banzhaf and Peter Nordin, and therefore a comparison with their work will be done.

No description available.

No description available.

Giving robots the ability to move around autonomously in various real-world environments has long been a major challenge for Artificial Intelligence. New approaches to the design and control of autonomous robots have shown the value of drawing inspiration from the natural world. Animals navigate, perceive and interact with various uncontrolled environments with seemingly little effort. Flying insects, in particular, are quite adept at manoeuvring in complex, unpredictable and possibly hostile environments.Inspired by the miniature machine view of insects, this thesis contributes to the autonomous control of mobile robots through the application of insect-based visual cues and behaviours. The parsimonious, yet robust, solutions offered by insects are directly applicable to the computationally restrictive world of autonomous mobile robots. To this end, two main navigational domains are focussed on: corridor guidance and visual homing.Within a corridor environment, safe navigation is achieved through the application of simple and intuitive behaviours observed in insect, visual navigation. By observing and responding to observed apparent motions in a reactive, yet intelligent way, the robot is able to exhibit useful corridor guidance behaviours at modest expense. Through a combination of both simulation and real-world robot experiments, the feasibility of equipping a mobile robot with the ability to safely navigate in various environments, is demonstrated.It is further shown that the reactive nature of the robot can be augmented to incorporate a map building method that allows previously encountered corridors to be recognised, through the observation of landmarks en route. This allows for a more globally-directed navigational goal.Many animals, including insects such as bees and ants, successfully engage in visual homing. This is achieved through the association of ++ / visual landmarks with a specific location. In this way, the insect is able to 'home in' on a previously visited site by simply moving in such a way as to maximise the match between the currently observed environment and the memorised 'snapshot' of the panorama as seen from the goal. A mobile robot can exploit the very same strategy to simply and reliably return to a previously visited location.This thesis describes a system that allows a mobile robot to home successfully. Specifically, a simple, yet robust, homing scheme that relies only upon the observation of the bearings of visible landmarks, is proposed. It is also shown that this strategy can easily be extended to incorporate other visual cues which may improve overall performance.The homing algorithm described, allows a mobile robot to home incrementally by moving in such a way as to gradually reduce the discrepancy between the current view and the view obtained from the home position. Both simulation and mobile robot experiments are again used to demonstrate the feasibility of the approach.

This paper describes an exercise in building a complete robot aimed at being as small as possible but using off-the-shelf components exclusively. The result is an autonomous mobile robot slightly larger than one cubic inch which incorporates sensing, actuation, onboard computation, and onboard power supplies. Nicknamed Squirt, this robot acts as a 'bug', hiding in dark corners and venturing out in the direction of last heard noises, only moving after the noises are long gone.

No description available.

This dissertation proposes autonomous robotic strategies for urban search and rescue (USAR) which are map-based semi-autonomous robot navigation and fully-autonomous robotic search, tracking, localization and mapping (STLAM) using a team of robots. Since the prerequisite for these solutions is accurate robot localization in the environment, this dissertation first presents a novel grid-based scan-to-map matching technique for accurate simultaneous localization and mapping (SLAM). At every acquisition of a new scan and estimation of the robot pose, the proposed technique corrects the estimation error by matching the new scan to the globally defined grid map. To improve the accuracy of the correction, each grid cell of the map is represented by multiple normal distributions (NDs). The new scan to be matched to the map is also represented by NDs, which achieves the scan-to-map matching by the ND-to-ND matching. In the map-based semi-autonomous robot navigation strategy, a robot placed in an environment creates the map of the environment and sends it to the human operator at a distant location. The human operator then makes decisions based on the map and controls the robot via tele-operation. In case of communication loss, the robot semi-autonomously returns to the home position by inversely tracking its trajectory with additional optimal path planning. In the fully-autonomous robotic solution to USAR, multiple robots communicate one another while operating together as a team. The base station collects information from each robot and assigns tasks to the robots. Unlike the semi-autonomous strategy there is no control from the human operator. To further enhance the efficiency of their cooperation each member of the team specifically works on its own task. A series of numerical and experimental studies were conducted to demonstrate the applicability of the proposed solutions to USAR scenarios. The effectiveness of the scan-to-map matching with the multi-ND representation was confirmed by analyzing the error accumulation and by comparing with the single-ND representation. The applicability of the scan-to-map matching to the real SLAM problem was also verified in three different real environments. The results of the map-based semi-autonomous robot navigation showed the effectiveness of the approach as an immediately usable solution to USAR. The effectiveness of the proposed fully- autonomous solution was first confirmed by two real robots in a real environment. The cooperative performance of the strategy was further investigated using the developed platform- and hardware-in-the-loop simulator. The results showed significant potential as the future solution to USAR. / Ph. D.

This master thesis describes the design and implementation of control sys- tem for autonomous robot which is able to run through user defined points in unknown environment without colliding with obstacles. The work contains analysis of the available hardware and software solutions, modular design with control system implementation divided into separate subsystems (control, lo- calization, route planning, driving the robot using Hermit curves and low-level hardware control). The work also contains explanation of rework of the school robotic platform. The implementation was tested on a created robotic platform. Driving the robot along the Hermit curve allows smooth and in some cases quicker passage through defined points, than passage consisting of rotations on the spot and direct movements. 1

Pour des tâches simples ou dans un environnement contrôlé, la coordination des différents processus internes d’un robot est un problème relativement trivial, souvent implémenté de manière ad-hoc. Toutefois, avec le développement de robots plus complexes travaillant dans des environnements non contrôlés et dynamiques, le robot doit en permanence se reconfigurer afin de s’adapter aux conditions extérieures et à ses objectifs. La définition d’une architecture de contrôle efficace permettant de gérer ces reconfigurations devient alors primordiale pour l’autonomie de tels robots. Dans ces travaux, nous avons d’abord étudié les différentes architectures proposées dans la littérature, dont l’analyse a permis d’identifier les grandes problématiques qu’une architecture de contrôle doit résoudre. Cette analyse nous a mené à proposer une nouvelle architecture de contrôle décentralisée, générique et réutilisable, selon une démarche qui intègre une approche "intelligence artificielle" (utilisation de raisonneur logique, propagation dynamique de contraintes) et une approche "génie logiciel" (programmation par contrats, agents). Après une présentation des concepts qui sous-tendent cette architecture et une description approfondie de son fonctionnement, nous en décrivons une implémentation, qui est exploitée pour assurer le contrôle d’un robot terrestre d’extérieur dans le cadre de tâches de navigation, d’exploration ou de suivi. Des résultats sont présentés et analysés. Dans une seconde partie, nous nous sommes penchés sur la modélisation et la vérifiabilité d’une telle architecture de contrôle. Après avoir analysé différentes solutions, nous décrivons un modèle complet de l’architecture qui utilise la logique linéaire. Nous discutons ensuite des différentes approches possibles pour montrer des propriétés d’atteignabilité et de sûreté de fonctionnement en exploitant ce modèle. Enfin nous abordons différentes voies d’enrichissement de ces travaux. En particulier, nous discutons des extensions possibles pour le contrôle d’un ensemble de robots coopérants entre eux, mais aussi de la nécessité d’avoir des liens plus forts entre cette couche de contrôle, et les approches de modélisation des fonctionnalités sous-jacentes. / For simple tasks in a controlled environment, the coordination of the internal processes of a robot is a relatively trivial task, often implemented in an ad-hoc basis. However, with the development of more complex robots that must operate in uncontrolled and dynamic environments, the robot must constantly reconfigure itself to adapt to the external conditions and its own goals. The definition of a control architecture to manage these reconfigurations becomes of paramount importance for the autonomy of such robots. In this work, we first study the different architectures proposed in the literature, and analyse the major issues that a control architecture must address. This analysis led us to propose a new architecture, decentralized, generic and reusable, integrating an artificial intelligence approach (use of logical reasoning, dynamic propagation of constraints) and a software engineering approach (programming by contract, agents). After a presentation of the concepts underlying this architecture and an in-depth description of its operation, we describe an implementation which is used to control of a ground robot for navigation, exploration and monitoring tasks. Results are presented and analyzed. In a second part, we focus on the modeling and verifiability of such a control architecture. After analyzing different solutions, we present a comprehensive model of the proposed architecture that uses linear logic. We then discuss the different possible approaches to assess the properties of reachability and safety within this model. Finally we discuss different ways to enrich this work. In particular, we discuss possible extensions to the control of a multiple cooperating robots, but also the need for stronger links between the control layer and the modeling.