No description available.

Autonomous decision-making machines – ranging from autonomous vehicles to chatbots – are already able to make decisions that have ethical consequences. If these machines are eventually deployed on a large scale, members of society will have to be able to trust the decisions that are made by these machines. For these machines to be trustworthy, their decisions must be overseen by socially accepted ethical principles; moreover, these principles and their role in machine decision-making must be transparent and explainable: it must be possible to explain why machine decisions are made and such explanations require that the mechanisms involved for making them are transparent. Furthermore, manufacturing companies have a corporate social responsibility to design such robots in ways that make them not only safe but also trustworthy. Members of society will not trust a robot that works in mysterious, ambiguous, or inexplicable ways, particularly if this robot is required to make decisions based on ethical principles. The current literature on embedding ethics in robots is sparse. This thesis aims to partially fill this gap in order to help different stakeholders (including policy makers, the robot industry, robots designers, and the general public) to understand the many dimensions of machine- executable ethics. To this end, I provide a framework for understanding the relationships among different stakeholders who legislate, create, deploy, and use robots and their reasons for requiring transparency and explanations. This framework aims to provide an account of the relationships between the transparency of the decision-making process in ethical robots, explanations for their behaviour, and the individual and social trust that results. This thesis also presents a model that decomposes the stages of ethical decision-making into their elementary components with a view to enabling stakeholders to allocate the responsibility for such choices. In addition, I propose a model for transparency which demonstrates the importance of and relationships between disclosure, transparency, and explanation which are needed for societies to accept and trust robots. One of the important stakeholders of robotics is the general public and, in addition to providing an analytical framework with which to conceptualize ethical decision-making, this thesis also performs an analysis of opinions drawn from hundreds of written comments posted on public forums concerning the behaviour of socially autonomous robots. This analysis provides insights into the layperson’s responses to machines that make decisions and offers support for policy recommendations that should be considered by regulators in the future. This thesis contributes to the area of ethics and governance of artificial intelligence.

Abstract Intraoperative surgery has created a need to develop new kinds of surgical tools. Also, the development of imaging techniques and devices has precipitated the need. Robotics plays an increasingly important role in surgery. A robot can yield better accuracy, smaller movements and, as a result, a faster healing process than a normal operation would require for recovering and healing larger cuts in the human body. Magnetic resonance imaging, MRI, is one of the safest imaging techniques, and it has excellent soft tissue contrast. In the last few years, MRI has become a more frequently-used technique in the intraoperative surgery, such as the biopsy. Brain biopsies in particular are easier to perform by the help of MRI. When designing a robot, or any other mechatronic device, for an MR environment, it becomes vital to consider its appropriateness, i.e., electric and magnetic compatibility with MRI. The latter is a notion related to the surgical procedure and the magnetic field being applied. It implies that instrumentation has to be more compatible with MR in higher magnetic fields. In this study, an MR-compatible robot was developed to work inside open MRI equipment. The MR compatibility of the robot was evaluated, using the testing method evolved during the study. The method helps select the suitable material and parts for mechatronic devices operating under MRI. Most notably, this work also devised and introduced new types of sensors to achieve better MR compatibility of the equipment. As a result of this research, a suitable material was developed for the robot's body as well as for the sensors, actuators and tools. Furthermore, it was deduced that some sensors and the control system when powered can not be used in the MR scanner at all. A further result was that the movement of a robotic arm does not disturb image quality in any way. The testing method developed in this research helps address the compatibility issues arising from the use of any device that works in MRI. The testing method can be used for magnetic fields of different strengths. The robot and the control parts constructed in the research were tested under a 0.23-T open MRI scanner. The results show which materials and fibre optics provide a highly MR compatible solution for an MRI environment. The results also show that normal electric motors can not be driven close to the magnetic field while imaging.

The thesis focused on the development and analysis of a distributed parameter model that apply to a class of an autonomous hyper-redundant slender robotic systems interacting with the environment. The class of robotic devices that will be implemented based on the modelling in this thesis, is intended to be autonomously deployed in unknown, unstructured environments, in which it has to accomplish different missions by being able to robustly negotiate unknown obstacles and unpredictable and unmodelled irregularities. Therefore the mechanical models presented here are inspired by some features of a class of organisms - millipedes and centipedes - that possess many of these capabilities. Specifically, these organisms posses flexible slender bodies whose shape morphs according to the curvature of the terrain on which they operate, and possess a highly redundant system of legs that couple the body with the terrain providing propulsion for forward or backward motion, with the high number of legs ensuring a robust distributed contact even on very irregular substrates. The mechanical model that naturally captures the structure of millipede bodies is the Timoshenko beam, which is therefore adopted here. Moreover, the coupling with the environment is modeled by a system of compliant elements, that provides a distributed support analogous to the one exerted by millipedes' legs; such support provides a distributed force that in a control framework is treated as the actuation for shape morphing, so that the body of the system deforms according to the curvature of the substrate. By using a Lagrangian mechanics approach, the evolution of the system is described in a suitable product Hilbert space, in which rigid body degrees of freedom and deformations are coupled. This formulation allows to pose a distributed parameter control problem in which shape morphing and locomotion are dictated by the interaction with the substrate, which in this case is approximated as rigid (that is, the profile of the substrate is not affected by the interaction with the system). Additionally, by modeling the material response of the substrate with a simple linear viscoelastic model, we pose an estimation problem in which, by measuring deformations and/or stresses on the body represented by the beam, we can infer the material properties of the substrate. In this case, the overall coupled system is modelled as a beam on a multi-layer viscoelastic foundation. Predictions of this sensor model are in good agreement with published results, suggesting that the system can be used in a versatile way as an autonomous agent operating in a generic environment, and simultaneously as a sensor that could inform the action of the system itself, or that could be used to monitor the environment. The modeling work done in this study opens the possibility for the implementation in engineering systems applied to environmental monitoring and health applications, in which we envision the system to be used to estimate material properties of living tissues, that can be correlated to the diagnosis of classes of diseases.

The research developed force-motion strategies and subsequent force and position control algorithms, using a PUMA 560 robot arm and its original controller. A task decomposition methodology has been developed that enables a mechanical assembly task to be subdivided into a series of executable subtasks. By applying this methodology to the assembly of a hydraulic gear pump, a library of special purpose, task oriented, subtask programs were created. Most of these programs, though derived for a pump assembly task, are applicable (when used with appropriate parameters) to other assembly tasks. Most of the algorithms require force/torque sensory information that is supplied by a JR³ wrist force sensor. The force control algorithms use that data and system compliance in order to produce new position instructions that are transferred to the controller of the arm. The logic of the control law and system behaviour when contacting the environment, were checked, using the dynamics and compliance of a simplified structure of a robotic arm and its wrist sensor. A demonstration of the pump assembly task, using the arm, force sensor, controller and the derived library algorithms is an integral part of the thesis. / Applied Science, Faculty of / Mechanical Engineering, Department of / Graduate

The main subject of this thesis is use of inertial sensors for better motion and stability control of mobile robots. In background research are described the basic methods of mobile robots localization. Second part of background research is about mobile robot stability, in this part are described the methods of mobile robots stability control. In next part is description of MEMS accelerometers and gyroscopes and description of basic method of filtering and integration which are useful for input processing of these sensors. Thesis continues with inertial sensors analysis for application on control of wheeled and legged mobile robots. In end of thesis are specified the experiment results, which confirm the applicability of sensors for mobile robot control.

Parallel mechanisms are characterized by their kinematic structure presented a closed kinematic chain. The end effector of mechanism is then bound to a base of over one arm. This design provides an advantage especially at high stiffness of mechanism and related properties. The disadvantage of such a structure is then usually very limited working space. In terms of simulation modeling is the construction due to its complexity rather problematic. The aim of work is design comprehensive simulation model chosen parallel mechanism and its analysis in selected working modes. Expected to use Adams simulation environment, if necessary. Matlab - SimMechanics and Ansys.

The general public is being exposed to robots more often every day. This thesis focused on the advancement of research by analyzing whether or not the type of information provided by a robot determined the level of trust humans have for a robot. A study was conducted where the participants were asked to answer two different types of questions: mathematical/logical and ethical/social. The participants were divided into two different conditions: controlled and misinformed. A humanoid robot provided its own spoken answer after the participants said their answers. The participants then had the chance to select whose answers they would like to keep. During the misinformed condition, there were times when the robot purposely gave incorrect answers. The results of the study support the hypothesis that the participants were more likely to select the robot’s answers when the question type was mathematical/logical, whether the robot provided a correct or incorrect response.

The use of multi-robot teams in the Robot Operating System (ROS) has encountered difficulty in advancement because of a lack of effective ways for the robots to communicate. Several attempts towards solving this problem have been made, but these approaches have had trouble with either low fault tolerance or high network load. The Gray Transceiver is an interface and communication protocol for inter-robot communication using ROS. The Gray Transceiver leverages multicasting for reduced network load and increased fault tolerance. Results from simulations, high throughput testing, and live multi-robot evaluations are included. The live mult-robot and simulation evaluations show that it functions properly operating across multiple robots while tolerating faults. The high throughput test shows how the Gray Transceiver operates under high load across a several types of conditions.

No description available.