This dissertation demonstrates that effective control of autonomous mobile robots in real-world environments can be achieved by combining reactive and deliberative components into an integrated architecture. The reactive component allows the robot to respond to contingencies in real time. Deliberation allows the robot to make effective predictions about the world. By using different computational mechanisms for the reactive and deliberative components, much existing deliberative technology can be effectively incorporated into a mobile robot control system. The dissertation describes the design and implementation of a reactive control system for an autonomous mobile robot which is explicitly designed to interface to a deliberative component A programming language called ALF A is developed to program this system. The design of a control architecture which incorporates this reactive system is also described. The architecture is heterogeneous and asynchronous, that is, it consists of components which are structured differently from one another, and which operate in parallel. This prevents slow deliberative computations from adversely affecting the response time of the overall system. The architecture produces behavior which is reliable and goal-directed, yet reactive to contingencies, in the face of noise, limited computational resources, and an unpredictable environment. The system described in this dissertation has been used to control three real robots and a simulated robot performing a variety of tasks in real-world and simulated real-world environments. A general design methodology based upon bottom-up hierarchical decomposition is demonstrated. The methodology is based on the principle of cognizant failure, that is, that low-level activities should be designed in a way as to detect failures and state transitions at high levels of abstraction. Furthermore, the results of deliberative computations should be used to guide the robot's actions, but not to control those actions directly. / Ph. D.

This project focuses on the design and modeling of a two degree-of-freedom dissipative passive haptic display. Haptic displays are man-machine interfaces that transmit forces to the human operator. A dissipative passive haptic display is one that may only remove energy from the system using actuators such as brakes and dampers, thus ensuring the safety of the human operator. These devices may be used to implement virtual constraints such as desired paths and obstacles. Traditional friction brakes have previously been used as dissipative and coupling elements in a two degree-of-freedom parallel manipulator, resulting in undesired effects such as vibration, stiction, and slow response times. Alternatively, the new robot is actuated by rheological brakes, which feature fast response times and smooth application of torque. This approach aims to improve upon the accuracy and feel of the previous design. A commercial magnetorheological (MR) fluid brake was selected and put through an extensive series of tests. The data was used to develop a model that characterizes MR fluid behavior in low speed braking applications. A parallel five bar linkage was designed and built that has separate configurations corresponding to 3-brake and 4-brake operation. The length of each arm was chosen by means of a geometrical optimization that weighs the size and area of the workspace and actuator effects. A simulation was then developed by incorporating the brake model into the equations of motion of the robot. Next, two forms of path following velocity control were devised and tested in simulation. Finally, the accuracy, workload, and smoothness of both controllers and both configurations were examined in preliminary tests with human operators.

The primary objective of this research is the development a wind and solar-powered autonomous surface vehicle (WASP) for oceanographic measurements. This thesis presents the general design scheme, detailed aerodynamic and hydrodynamic aspects, sailing performance theory, and dynamic performance validation measurements obtained from a series of experiments. The WASP consists of a 4.2 meter long sailboat hull, a low-Reynolds number composite wing, a 2000 Watt-hour battery reservoir, a system of control actuators, a control system running on an embedded microprocessor, a suite of oceanographic sensors, and power regeneration from solar energy. The vehicle has a maximum speed of five knots and weighs approximately 350 kilograms. Results from four oceanographic missions that were conducted in the Port Everglades Intracoastal Waterway in Dania Beach [sic] Florida are presented. Water temperature, salinity and oxidation-reduction measurements recorded during these missions are also discussed. The combination of a mono-hull and solid wing in an autonomous system is a viable design for a long-range ocean observation platform. The results of four near-shore ocean observation missions illustrate the initial capabilities of the design. Future work aimed to further reduce both the mass of the wing design and the power requirements of the system will increase performance in all operating conditions and should be considered. Furthermore, the progression of the legal framework related to ocean vehicles must be pursued with respect to unmanned autonomous systems. / by Patrick Forde Rynne. / Thesis (M.S.C.S.)--Florida Atlantic University, 2008. / Includes bibliography. / Electronic reproduction. Boca Raton, Fla., 2008. Mode of access: World Wide Web.

Thomas Kin Fong Lei. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2000. / Includes bibliographical references (leaves 64-66). / Abstracts in English and Chinese. / LIST OF FIGURES --- p.IV / Chapter 1 --- Introduction --- p.1 / Chapter 2 --- Model-based Control of SMA Wires --- p.3 / Chapter 2.1 --- Model Identification of SMA Wires --- p.3 / Chapter 2.1.1 --- Temperature-Current Relationship --- p.3 / Chapter 2.1.2 --- Stress-Strain Relationship --- p.5 / Chapter 2.1.3 --- Martensite Fraction-Temperature Relationship --- p.8 / Chapter 2.2 --- Model-based Position Control of Two Linking SMA Wires --- p.9 / Chapter 2.3 --- Summary --- p.12 / Chapter 3 --- Neural-fuzzy-based Control of SMA Wires --- p.13 / Chapter 3.1 --- Adaptive Neuro-fuzzy Inference System (ANFIS) --- p.13 / Chapter 3.1.1 --- ANFIS Architecture --- p.13 / Chapter 3.1.2 --- Hybrid Learning Algorithm --- p.16 / Chapter 3.2 --- Generalized Neural Network (GNN) --- p.20 / Chapter 3.2.1 --- GNN Architecture --- p.20 / Chapter 3.2.2 --- Approximation of the GNN --- p.22 / Chapter 3.2.3 --- Backpropagation Training Algorithm --- p.24 / Chapter 3.2.4 --- Complexity Reduction of the GNN --- p.25 / Chapter 3.2.5 --- Error Bound of In-exact Reduction of the GNN --- p.29 / Chapter 3.3 --- Neural-fuzzy-based Position Control of Four Linking SMA Wires --- p.32 / Chapter 3.3.1 --- ANFIS-based Position Control of Four Linking SMA Wires --- p.32 / Chapter 3.3.2 --- GNN-based Position Control of Four Linking SMA Wires --- p.35 / Chapter 3.3.3 --- Performance Comparison of ANFIS and GNN Algorithms --- p.37 / Chapter 3.4 --- Summary --- p.39 / Chapter 4 --- SMA Actuated Artificial Face --- p.40 / Chapter 4.1 --- Muscles of the Human Face --- p.40 / Chapter 4.2 --- The Software Part: facial model --- p.41 / Chapter 4.3 --- The Hardware Part: artificial face and peripheral interface --- p.43 / Chapter 4.3.1 --- SMA Actuated Artificial Face --- p.43 / Chapter 4.3.2 --- Peripheral Interface --- p.45 / Chapter 4.4 --- Position Control on the Artificial Face --- p.47 / Chapter 4.4.1 --- Model-based Position Control on Artificial Face --- p.48 / Chapter 4.4.2 --- Neural-fuzzy-based Position Control on Artificial Face --- p.49 / Chapter 4.4.3 --- Comparison of the Model-based and Reduced GNN Control of Artificial Face --- p.49 / Chapter 4.5 --- Experimental Result --- p.50 / Chapter 5 --- Conclusion --- p.52 / Appendix1 --- p.53 / Appendix2 --- p.55 / Appendix3 --- p.56 / Appendix4 --- p.58 / Bibliography --- p.64

With current state-of-the-art approaches, development of a single mobile manipulation capability can be a labor-intensive process that presents an impediment to the creation of general purpose household robots. At the same time, we expect that involving a larger community of non-roboticists can accelerate the creation of new novel behaviors. We introduce the use of a software authoring environment called ROS Commander (ROSCo) allowing end-users to create, refine, and reuse robot behaviors with complexity similar to those currently created by roboticists. Akin to Photoshop, which provides end-users with interfaces for advanced computer vision algorithms, our environment provides interfaces to mobile manipulation algorithmic building blocks that can be combined and configured to suit the demands of new tasks and their variations. As our system can be more demanding of users than alternatives such as using kinesthetic guidance or learning from demonstration, we performed a user study with 11 able-bodied participants and one person with quadriplegia to determine whether computer literate non-roboticists will be able to learn to use our tool. In our study, all participants were able to successfully construct functional behaviors after being trained. Furthermore, participants were able to produce behaviors that demonstrated a variety of creative manipulation strategies, showing the power of enabling end-users to author robot behaviors. Additionally, we introduce how using autonomous robot learning, where the robot captures its own training data, can complement human authoring of behaviors by freeing users from the repetitive task of capturing data for learning. By taking advantage of the robot's embodiment, our method creates classifiers that predict using visual appearances 3D locations on home mechanisms where user constructed behaviors will succeed. With active learning, we show that such classifiers can be learned using a small number of examples. We also show that this learning system works with behaviors constructed by non-roboticists in our user study. As far as we know, this is the first instance of perception learning with behaviors not hand-crafted by roboticists.

Thesis (MScEng)-- Stellenbosch University, 2013. / ENGLISH ABSTRACT: This thesis presents the development of a flight control system (FCS) for an unmanned, unpowered parafoil and the integration with an existing parafoil system in collaboration with a team at the University of Cape Town (UCT). The main goal of the FCS is to autonomously guide the parafoil from an arbitrary deployment position to a desired landing target. A nonlinear 8 degrees of freedom (8-DOF) parafoil model by C. Redelinghuys is incorporated into a MATLAB Simulink simulation environment. The non-linear model is numerically linearised and modal decomposition techniques are used to analyse the natural modes of motion. All modes are determined to be stable but a poorly damped lateral payload relative twist mode is present which causes large payload yaw oscillations. The FCS is divided into stability augmentation, control and guidance subcomponents. Stability augmentation is proposed in the form of a yaw rate damper which provides artificial damping for the oscillatory payload twist mode. For control, a yaw rate controller is designed with the aim of a fast response while not exciting the payload twist oscillation. Subsequently, an existing guidance method is implemented for path following. Autonomous path planning and mission control logic is created, including an energy management (EM) method to eliminate excess height and a terminal guidance (TG) phase. The TG phase is the final turn before landing and is the last chance to influence landing accuracy. A TG algorithm is implemented which generates an optimal final turn and can be replanned en route to compensate for unknown wind and other disturbances. The FCS is implemented on existing avionics, integrated with the parafoil system and verified with hardware in the loop (HIL) simulations. Flight tests are presented but are limited to remote control (RC) tests that verify the integration of the avionics and the parafoil system and test preliminary FCS components. / AFRIKAANSE OPSOMMING: Hierdie tesis dra die ontwikkeling voor van ‘n vlug-beheerstelsel (VBS) vir ’n onbemande, onaangedrewe valskerm-sweeftuig (parafoil), asook die integrasie daarvan met ’n bestaande stelsel. Die projek is in samewerking met ’n span van die Universiteit van Kaapstad (UCT) uitgevoer. Die VBS se hoof doel is om die sweeftuig outonoom vanaf ’n arbitrêre beginpunt na ’n gewensde landingsteiken te lei. ’n Nie-lineêre 8 grade van vryheid sweeftuig model deur C. Redelinghuys is in die MATLAB Simulink omgewing geïnkorporeer. Die nie-lineêre model is numeries gelineariseer om ’n lineêre model te verkry, waarna die natuurlike gedrag van die tuig geanaliseer is. ’n Swak gedempte laterale draai ossillasie van die loonvrag is geïdentifiseer. Die VBS is opgedeel in stabiliteitstoevoeging, beheer en leiding. ’n Giertempo-demper (yaw rate damper) is as stabiliteitstoevoeging om die loonvrag ossillasie kunsmatig te demp, voorgestel. ’n Giertempo-beheerder is ontwerp met die klem op ’n vinnige reaksie terwyl die opwekking van die loonvrag ossillasie terselfdetyd verhoed word. Daarna is ’n bestaande metode vir trajekvolging geïmplementeer. Outonome padbeplanning en oorhoofse vlugplan logika is ontwikkel, insluitend ’n energie-bestuur (EB) metode, om van oortollige hoogte ontslae te raak, asook ’n terminale leiding (TL) metode. Die TL fase verwys na die finale draai voor landing en is die laaste kans om die landingsakkuraatheid te beïnvloed. ’n Bestaande TL algoritme is geïmplementeer wat ’n optimale trajek genereer en in staat is om vir wind en ander versteurings te kompenseer deur die trajek deurgaans te herbeplan. Die VBS is op bestaande avionika geïmplementeer, met die sweeftuigstelsel geïntegreer en met behulp van hardeware in die lus (HIL) simulasies geverifieer. Vlugtoetse is voorgedra, maar is egter beperk tot radio beheer vlugte wat die korrekte integrasie van die avionika en die voertuig toets, asook ’n beperkte aantal voormalige VBS toetse.

Thesis (M. Tech. (Engineering Electrical)) -- Central University of Technology, Free State, [2014] / This dissertation evaluates the software and hardware components used to develop a Reconfigurable Assembly System with an Integrated Information Management System. The assembly system consists of a modular Cartesian robot and vision system. The research focuses on the reconfigurability, modularity, scalability and flexibility that can be achieved in terms of the software and hardware components used within the system. The assembly system can be divided into high-level control and low-level control components. All information related to the product, Cartesian positioning and processes to follow resides in the Information Management System. The Information Management System is the high-level component and consists of a database, web services and low-levelcontrol drivers. The high-level system responds to the data received from the low-level systems and determines the next process to take place. The low-level systems consist of the PLC (Programmable Logic Controller) and the vision system. The PLC controls the Cartesian robot’s motor controllers and handles all events raised by field devices (e g. sensors or push buttons). The vision system contains a number of pre-loaded inspections used to identify barcodes and parts, obtain positioning data and verify the products’ build quality. The Cartesian robot’s positioning data and the vision system’s inspections are controlled by the Information Management System. The results showed that the high-level control software components are able to add more modularity and reconfigurability to the system, as it can easily adapt to changes in the product. The high-level control components also have the ability to be reconfigured while the assembly system is online without affecting the assembly system. The low-level control system is better suited to handling the control of motor controllers, field devices and vision inspections over an industrial network.

Thesis (M. Tech.) -- Central University of Technology, Free State, 2009 / Automated Guided Vehicles (AGVs) have many different design specifications, although they all have certain design features in common, for instance they are designed to follow predetermined paths, and they need to be aware of their surroundings and changes to their surroundings. They are designed to house sensors for navigation and obstacle avoidance. In this study an AGV platform was developed by modifying an electric wheelchair. A serial port interface was developed between a computer and the control unit of the electric wheelchair, which enables the computer to control the movements of the platform. Different sensors were investigated to determine which would be best suited and most effective to avoid collisions. The sensors chosen were mounted on the AGV and a programme was developed to enable the sensors to assist in avoiding obstacles. An imaging device as an additional sensor system for the AGV was investigated. The image produced by a camera and dome mirror was processed into a panoramic image representing an entire 360o view of the AGV‟s surroundings. The reason for this part of the research was to enable the user to make corrections to the AGV‟s path if it became stuck along the track it was following. The entire system was also made completely wireless to improve the flexibility of the AGV‟s applications.

Thesis (M.Tech.) - Central University of Technology, Free State, 2007 / The latest technological trend worldwide, is automation. Reducing human labour and introducing robots to do the work is a pure business decision. The reason for automating a plant can be some, or all, of the following: Improve productivity Reduce labour and equipment costs Reduce product damage System reliability can be monitored Improves plant safety When the automation process is started, Automatic Guided Vehicles (AGVs) will be one of the first commodities that can be used. The reason for this is that they are so versatile. They can be programmed to follow specific paths when moving material from one point to another and the biggest advantage of all is that they can operate for twenty four hours a day. Automatic Guided Vehicles are developed for many different applications and therefore many different types of AGVs are available. All AGVs are equipped with sensors so that they are able to “see” what is happening around them. Since the AGV must be able to function without any human help or control, it must be able to navigate through the work environment. In this study a remote control car was converted to an AGV and thorough research was done on the different types of sensors that can be used to make the AGV more intelligent when it comes to navigating in an unknown environment.

Thesis (M.Tech.) - Central University of Technology, Free State, 2006 / The ultimate goal regarding this research was to make an intelligent learning machine, thus a new method had to be developed. This was to be made possible by creating a programme that generates another programme. By constantly changing the generated programme to improve itself, the machines are given the ability to adapt to there surroundings and, thus, learn from experience. This generated programme had to perform a specific task. For this experiment the programme was generated for a simulated PIC microcontroller aboard a simulated robot. The goal was to get the robot as close to a specific position inside a simulated maze as possible. The robot therefore had to show the ability to avoid obstacles, although only the distance to the destination was given as an indication of how well the generated programme was performing. The programme performed experiments by randomly changing a number of instructions in the current generated programme. The generated programme was evaluated by simulating the reactions of the robot. If the change to the generated programme resulted in getting the robot closer to the destination, then the changed generated programme was kept for future use. If the change resulted in a less desired reaction, then the newly generated programme was removed and the unchanged programme was kept for future use. This process was repeated for a total of one hundred thousand times before the generated program was considered valid. Because there was a very slim chance that the instruction chosen will be advantageous to the programme, it will take many changes to get the desired instruction and, thus, the desired result. After each change an evaluation was made through simulation. The amount of necessary changes to the programme is greatly reduced by giving seemingly desirable instructions a higher chance of being chosen than the other seemingly unsatisfactory instructions. Due to the extensive use of the random function in this experiment, the results differ from one another. To overcome this barrier, many individual programmes had to be generated by simulating and changing an instruction in the generated programme a hundred thousand times. This method was compared against Genetic Algorithms, which were used to generate a programme for the same simulated robot. The new method made the robot adapt much faster to its surroundings than the Genetic Algorithms. A physical robot, similar to the virtual one, was build to prove that the programmes generated could be used on a physical robot. There were quite a number of differences between the generated programmes and the way in which a human would generally construct the programme. Therefore, this method not only gives programmers a new perspective, but could also possibly do what human programmers have not been able to achieve in the past.