This thesis addresses the problem of intelligent control of autonomous mobile robots, particularly under circumstances unforeseen by the designer. As the range of applications for autonomous robots widens and increasingly includes operation in unknown environments (exploration) and tasks which are not clearly specifiable a priori (maintenance work), this question is becoming more and more important. It is argued that in order to achieve such flexibility in unforeseen situations it is necessary to equip a mobile robot with the ability to autonomously acquire the necessary task achieving competences, through interaction with the world. Using mobile robots equipped with self-organising, behaviour-based controllers,experiments in the autonomous acquisition of motor competences and navigational skills were conducted to investigate the viability of this approach. A controller architecture is presented that allows extremely fast acquisition of motor competence such as obstacle avoidance, wall and corridor following and deadend escape: these skills are obtained in less than five learning steps,performed in under one minute of real time. This is considerably faster than previous approaches. Because the effective wiring between sensors and actuators is determined autonomously by the robot, sensors and actuators may initially be wired up arbitrarily,which reduces the risk of human error during the setting up phase of the robot. For the first time it was demonstrated that robots also become able to autonomously recover from unforeseen situations such as changes in the robot's morphology, the environment or the task. Rule-based approaches to error recovery obviously cannot offer recovery from unforeseen errors,as error situations covered by such approaches have to be identified beforehand. A robust and fast map building architecture is presented that enables mobile robots to autonomously construct internal representations of their environment, using self-organising feature maps. After a short training time the robots are able to use these self-organising feature maps successfully for location recognition. For the first time the staged acquisition of multiple competences in mobile robots is presented. First obtaining fundamental motor competences such as wall following and deadend escape (primary skills), the robots use these in a second stage to learn higher levels of competence such as the navigational task of location recognition (secondary skills). Besides laying the foundation of autonomous, staged acquisition of high level competences, this approach has the interesting property of securely grounding secondary skills in the robot's own experience, as these secondary skills are defined in terms of the primary ones.

In this thesis, a variety of different approaches are proposed to study the energy restoration problem in wireless sensor networks by one or more robots. First, we introduce an on-demand decentralized strategy performed by a robot that visits the sensors in a predefined circular order. We study it both analytically and experimentally analyzing the impact of various network parameters on network coverage, disconnection time, and time sensors have to wait to be served. We then introduce an optimal centralized approach as a benchmark to assess how close to optimal our on-demand strategy is, and we discover that, for sufficiently large networks, the on-demand strategy is indeed optimal. We then propose an even simpler mechanism where the robot simply moves blindly along the circular order, which is experimentally shown to be as efficient as the other two. The results above apply to arbitrary sensor network; we then consider a common special topology: a linear arrangement of sensors, were we propose three restoring mechanisms. We compare them experimentally discovering, once again, that the simplest approach is also the best, in most cases. We finally consider the case of multiple robots. We propose two strategies where the network is portioned among the robots and each robot takes care of a portion, and we compare those with a collaborative strategy where all robots work on the global network. The main general result of this study is that simple solutions are often as good as more sophisticated ones. In fact, a totally blind strategy where a robot simply moves around restoring energy on its way turns out to be as efficient as the best possible centralized solution for most networks.

Thesis (M.S.M.E.)--University of Delaware, 2006. / Principal faculty advisor: Sunil K. Agrawal, Dept. of Mechanical Engineering. Includes bibliographical references.

The control of nonholonomic wheeled mobile robots (WMRs) has gained a lot of attention in the field of robotics over the past few decades as WMRs provide an increased range of motion resulting in a larger workspace. This research focuses on the application of Model Predictive Control (MPC) for real-time trajectory tracking of a nonholonomic WMR. MPC is a control strategy in which the control law is designed based on optimizing a cost function. The input and output constraints that may arise in practical situations can be directly incorporated into the control system using MPC. Computation time is the biggest hurdle in adapting MPC strategies for trajectory tracking. This research applies a non-feasible active set MPC algorithm developed in [1] which is faster than the traditional active set methods (ASMs). A discrete-time linear model of a general WMR is used for the simulation. MATLAB simulations are performed for tracking circular as well as square trajectories using the discretized WMR model and the non-feasible ASM (NF-ASM). The performance of NF-ASM is compared to two other well-known traditional algorithms, i.e. Fletcher’s ASM and MATLAB’s Quadratic Programming algorithm. It is shown that, although all these algorithms are capable of providing satisfactory trajectory tracking performance, NF-ASM is a better choice in terms of the simulation time and required number of iterations for realtime trajectory tracking of any type as long as the constraints on the inputs stay active for a long period during the simulation. / UOIT

The goal of this thesis is to develop an intuitive, adaptive, and flexible architecture for controlling intelligent mobile robots. This architecture is a hybrid architecture that combines deliberative planning, reactive control, finite state automata, behaviour trees and uses competition for behaviour selection. This behaviour selection is based on a task manager, which selects behaviours based on approximations of their applicability to the current situation and the expected reward value for performing that behaviour. One important feature of this architecture is that it makes important behavioural information explicit using Extensible Markup Language (XML). This explicit representation is an important part in making the architecture easy to debug and extend. The utility, intuitiveness and flexibility of this architecture is shown in an evaluation of this architecture against older control programs that lack such explicit behavioural representation. This evaluation was carried out by developing behaviours for several common robotic tasks and demonstrating common problems that arose during the course of this development. / February 2006

Mobile robots are used for a wide range of purposes such as mapping an environment and transporting material goods. Regardless of the specific application, the navigation of the mobile robot is usually divided into three separate parts: localization, path planning and path execution. Localization is the process of determining the location of the robot with respect to a reference coordinate system. There are many different approaches to localizing a mobile robot which employ a wide variety of sensors. The objective of my research is to develop a method for the localization of multiple mobile robots equipped with inexpensive range sensors in an indoor environment. Each mobile robot will be equipped with a rotating infrared sensor and a rotating CMOS camera. The multiple mobile robot system will be treated as a linked robot for localization. The proposed localization method is verified via both simulation and experiment. Through the use of the virtual link length and relative heading information, a system of mobile robots can be effectively localized using detected environmental features.

Autonomous mobile robots have the potential to be extremely beneficial to society due to their ability to perform tasks that are difficult or dangerous for humans. These robots will necessarily interact with their environment through the two fundamental processes of acting and sensing. Robots learn about the state of the world around them through their sensations, and they influence that state through their actions. However, in order to interact with their environment effectively, these robots must have accurate models of their sensors and actions: knowledge of what their sensations say about the state of the world and how their actions affect that state. A mobile robot’s action and sensor models are typically tuned manually, a brittle and laborious process. The robot’s actions and sensors may change either over time from wear or because of a novel environment’s terrain or lighting. It is therefore valuable for the robot to be able to autonomously learn these models. This dissertation presents a methodology that enables mobile robots to learn their action and sensor models starting without an accurate estimate of either model. This methodology is instantiated in three robotic scenarios. First, an algorithm is presented that enables an autonomous agent to learn its action and sensor models in a class of one-dimensional settings. Experimental tests are performed on a four-legged robot, the Sony Aibo ERS-7, walking forward and backward at different speeds while facing a fixed landmark. Second, a probabilistically motivated model learning algorithm is presented that operates on the same robot walking in two dimensions with arbitrary combinations of forward, sideways, and turning velocities. Finally, an algorithm is presented to learn the action and sensor models of a very different mobile robot, an autonomous car. / text

M. Tech. Engineering: Electrical. / Discusses the purpose of the study to develop a MFC scheme that can adapt and adjust to the changing workspace to enhance WMR operations.

This report summarizes the work performed for the design and development of the Proteus research robot. The Proteus design is motivated by the need for a modular, flexible, and usable autonomous robotic platform. To accomplish these goals, a modular hardware architecture coupled with low-power, high-computation processing is presented. The robot is subdivided into three layers: mobility, computation, and application. The interface between layers is characterized by well defined APIs and may be individually replaced to achieve different functionality. An efficient low-level event scheduler is described along with higher-level software algorithms for motion control and navigation. Experiments of Proteus robots are provided including field tests and collaboration with outside research institutions. / text

published_or_final_version / Physics / Master / Master of Philosophy