No description available.

This thesis studies the problem of intelligent localisation for an autonomous underwater vehicle (AUV). After an introduction about robot localisation and specific issues in the underwater domain, the thesis will focus on passive techniques for AUV localisation, highlighting experimental results and comparison among different techniques. Then, it will develop active techniques, which require intelligent decisions about the steps to undertake in order for the AUV to localise itself. The undertaken methodology consisted in three stages: theoretical analysis of the problem, tests with a simulation environment, integration in the robot architecture and field trials. The conclusions highlight applications and scenarios where the developed techniques have been successfully used or can be potentially used to enhance the results given by current techniques. The main contribution of this thesis is in the proposal of an active localisation module, which is able to determine the best set of action to be executed, in order to maximise the localisation results, in terms of time and efficiency.

Unmanned Aerial System (UAS) applications are growing day by day and this will lead Unmanned Aerial Vehicle (UAV) in the close future to share the same airspace of manned aircraft. This implies the need for UAS to define precise safety standards compatible with operations standards for manned aviation. Among these standards the need for a Sense And Avoid (S&A) system to support and, when necessary, substitute the pilot in the detection and avoidance of hazardous situations (e.g. midair collision, controlled flight into terrain, flight path obstacles, and clouds). This thesis presents the work come out in the development of a S&A system taking into account collision risks scenarios with multiple moving and fixed threats. The conflict prediction is based on a straight projection of the threats state in the future. The approximations introduced by this approach have the advantage of high update frequency (1 Hz) of the estimated conflict geometry. This solution allows the algorithm to capture the trajectory changes of the threat or ownship. The resolution manoeuvre evaluation is based on a optimisation approach considering step command applied to the heading and altitude autopilots. The optimisation problem takes into account the UAV performances and aims to keep a predefined minimum separation distance between UAV and threats during the resolution manouvre. The Human-Machine Interface (HMI) of this algorithm is then embedded in a partial Ground Control Station (GCS) mock-up with some original concepts for the indication of the flight condition parameters and the indication of the resolution manoeuvre constraints. Simulations of the S&A algorithm in different critical scenarios are moreover in-cluded to show the algorithm capabilities. Finally, methodology and results of the tests and interviews with pilots regarding the proposed GCS partial layout are covered.

This thesis describes a new approach for a vision-based positioning system for Un- manned Aerial Vehicles using a recognition method based on known, robust geo- graphic landmarks. Landmarks are used to calculate a position estimate in a global coordinate frame without requiring external signals, such as GPS. Absolute systems are of interest as they provide a redundant positioning system, allow UAVs to oper- ate when GPS-denied and can enable high-precision landings for spacecraft. The core challenge with vision-based absolute positioning is recognition of land- marks. Most abundant landmarks, such as buildings, are visually similar and dif- cult to distinguish. Previous research in the area tends to focus on matching raw aerial image data to a set of reference images. While these methods can achieve acceptable results in speci c conditions, they struggle with variations in lighting, seasonal changes and changing environments. This thesis presents a new multi- stage method that aims to solve this using a high-level matching framework where landmarks identi ed in an aerial image are matched to a reference database. This has led to the development of a geometric feature descriptor that encodes the topography of landmarks. The proposed system therefore matches the arrangement of features rather than the appearance, which lets it distinguish individual landmarks in large sets (20,000+ features). Since the arrangement of landmarks often is semi- structured and ambiguous, in particular when considering man-made landmarks, a matching stage has been developed that uses a number of strategies to enable matching of individual landmarks to a full database. The results have been evaluated for two conceptual vehicles with acceptable results, highlighting the strengths of the proposed system as well as areas for improve- ment.

The octorotor is an unmanned VTOL capable vehicle with eight motors with xed pitch rotors. It is controlled by varying the speeds of its eight motors which are placed around the vehicle. There is no need for a complex swashplate system, making the vehicle low cost and dynamically simple. The increase in the number of e ectors over the quadrotor allows for inbuilt hardware redundancy. It is this redundancy which is of particular interest as the capabilities and applications of VTOL capable UAVs increases and the payloads become more expensive and sensitive. It would be unacceptable for a hardware failure to result in the loss of the vehicle and payload, especially if operating in close proximity to people. An operational requirement is that the operator must be able to control the vehicle's position and yaw angle. Position reference commands are generated in an inertial frame and these must be related to the vehicle- xed frame through a rotation matrix. The downfall of this method is that trigonometric singularities exist for large body angles where gimbal lock can occur. For this reason the unit quaternion attitude representation is used. The octorotor is not open-loop stable so a PD controller is used to provide for singularity-free, almost global asymptotic stability which is capable of following ightpaths as well as recovering from an initial inverted attitude. The output of the controller is called the virtual control since this demand is passed to the control allocation subsystem where the overall forces and moments are generated. A suitable control allocation method is needed since there are more e ectors than actuated degrees of freedom. The e ectors are assumed to be linear and various methods are used to provide constrained control allocation. If the virtual control is constrained then the allocation problem is always the unconstrained allocation problem and is guaranteed to be successful. By applying the constraints directly to the e ectors it is not necessary to use complex face searching algorithms to calculate the constrained virtual control. An objective of this thesis is to ensure that e ector failures do not a ect the vehicle's ight performance. This is integral to the aim of demonstrating that the hardware redundancy is su cient to allow ights over populated areas. E ector failures are modelled as an instantaneous loss of thrust from an e ector. This causes an adverse roll, pitch, and yaw disturbance as well as a drop in altitude. The recovery is based on the fault hiding method where the virtual control remains invariant from the nominal case and the fault is hidden in the plant. If none of the remaining e ectors are saturated then the failure-free performance is maintained and the operator should not notice any change to the vehicle handling. Kalman controllability analysis is done to determine the combinations of e ector failures which result in a controllable vehicle. Flight testing has demonstrated the suitability of the controller to the task of stabilising the vehicle. The failure scenarios are initialised before the ight and the performance is invariant to the health of the e ectors. The reasons for di erences between the simulation data and ight data are explained. Future work will implement an online fault detection scheme.

Part of the testing process for athletic footwear is exposing the shoes to realistic wear conditions; this can be in the form of user trials or, as is becoming more common place, the use of mechanical test devices. However, current mechanical test devices tend to be somewhat simplistic and fail to expose the footwear to the realistic loading environment. Thus, the aim of this thesis was to investigate the potential of using an off the shelf 6 Degrees-of-Freedom industrial robot to emulate the ground contact phase of human gait. This was achieved through addressing four research questions. The first research question aimed to outline the biomechanical features that were to be emulated and what their typical values were. Kinematics and kinetics of the real human gait were then collected, for use in programming the robot and evaluating its outputted movements. This was complemented by a comprehensive review of relevant literature. Previous investigations had highlighted the need for understanding of the robot s capabilities. This was taken further and input parameters such as level of robotic smoothing, programme velocity and the number of three dimensional co-ordinate points used were found to have an effect on the output kinematics of the robot. These features were also found to be part of the accompanying programme software (RoboGuide). Despite this, the differences were not identical and it was concluded that the software could only have a limited use in supporting the wider thesis aim. Prior to emulation, there was a need for robot set-up and its environment to be optimised. A new robot end-effector, with improved biofidelity, was developed which incorporated a new way of generating the robot motion that intended to aid kinetic and kinematic emulation. Further to this, analysis on robot movements in various locations identified the optimal location for the ground contact phase to be achieved. Using all of the gathered knowledge the robot was programmed to complete a footstrike for human walking using two types of programming method. When the robot is programmed directly with the human kinematic data the emulation of the footstrike is relatively poor; ground contact time is too long with an increased footprint size and poor ground reaction force profiles replication. Using a rotation about a fixed point on the footform led to improved, although not complete, emulation of the human gait parameters. The developed system has been shown to improve on previous work at Loughborough University and is also comparable with what is being used in industry and developed within academia. The concept remains in the early phases but the current study indicates that future work can move the robot further towards being able to produce a more biofidelic emulation that can be used in the footwear testing industry.

Deterministic control communication is a backbone of many novel complex (robotic) engineering systems, e.g. HUBO uses CAN. The aim of this research , is to develop an approach for cheap and deterministic control communication using Ethernet. The objectives require a reliable data-transfer service between the various units for sensing and actuation. The key discriminator to assess reliability is determinism, (e.g. causality and well defined order of communication events in time and transaction). The strategy to support this at the network level, include investigations of approaches based on time-triggering, arbitration and guarantees of latency. A half-duplex Ethernet network populated with a small/medium number of Media Access Controllers (MACs) is used for timed real-time communication. A time-triggered approach is used, i.e. a synchronization signal triggers the sending of data from each Ethernet data transmitting unit. Moreover, data packets are sent at well defined times after each trigger instance to reduce collisions. Collisions mainly occur due to jitter of the transmission system, so that arbitration (similar to CANopen) is necessary. A Linear Backoff scheme is used in comparison to the Binary Exponential Backoff scheme. Matlab and Field Programmable Gate Array (FPGA) technology, i.e. Xilinx XC3S500E from the Spartan-3E family, are used to simulate and implement the practical test of the Ethernet communication strategy. Practical comparisons to a time division multiple access scheme for the Matlab simulation environment and the FPGA based experimental setup are carried out. Furthermore, a theoretical analysis of the communication approach is provided. In particular the advantage of . the communication unit specific minimal backoff time using a Nlarkov model is shown.

Legged systems are capable of locomotion in a far wider range of environments compared with wheeled or tracked vehicles. Man-made legged locomotion systems are currently economically inferior to wheeled vehicles, tracked vehicles, and legged locomotion systems which exist in nature. In previous research, trajectory optimisation has been applied to legged locomotion systems of a certain design to improve economy, however there has been little work in which the economy of legged locomotion systems is improved through design. This thesis describes the search for economically optimal designs of legged locomotion systems using parametric analysis. An abstract mathematical model of a legged locomotion system was constructed which included an imperfect spring element and electromagnetic actuator. Several actuation strategies were developed which allowed the model to be controlled to perform continuous locomotion. Six performance metrics were identified whose reduction in value implied an increase in economy. The model was simulated performing four locomotion exercises of increasing complexity, beginning with a system oscillating in the absence of gravity and ending with a planar locomotion system. Across these four locomotion exercises, the effect of system architecture, actuation strategy and key system parameters upon economy was determined. Two prototype legged locomotion systems were also constructed for comparison with simulation. It was not possible to simultaneously minimise all performance metrics because minimisation of some performance metrics required choices of system architecture, actuation strategy or system parameter which prevented minimisation of other performance metrics. The economically optimal design of a legged locomotion system was one which minimised the maximum force and peak power requirements of the locomotion system's actuator. Parallel compliance was economically superior to series compliance, which is particularly noteworthy given the comparative rarity of parallel compliance in current man-made legged locomotion systems.

A key environmental concern in the 21st century is polluted water, requiring monitoring or remediation, often in locations which are hazardous, expansive, or difficult to reach. Robots capable of long term autonomous operation with the ability to tackle these environmental challenges are in great need. Microbial fuel cells (MFCs) are an emerging technology for water decontamination and electricity generation which convert biodegradable matter found in waterways, including pollutants such as algae and petrochemicals, to usable electrical power. As such MFCs present a promising, bioinspired power source for remotely operating robots, particularly where the use of more established energy-scavenging technologies is limited. One of the greatest challenges in environmental robotics is to develop machines with the compliance and adaptability that equips natural organisms for unassisted survival in uncertain and changeable surroundings. The emergence of mechanisms that closely mimic biological organisms is prevalent in state of the art research and can advance MFC powered robots by enabling them to to forage for food and locomote biomimetically. Artificial muscles with low mass and high efficiency, including electro-active polymers, are well suited to the low voltage, relatively low power, output of MFCs. This thesis presents these complementary technologies in the design of a biomimetic, energy-autonomous artificial organism capable of long term, unassisted operation. We consider artificial muscles powered by artificial metabolism in an investigation that covers three objectives: • Design of systems comprising soft ionically and electronically active polymers that may be used for both power generation in MFCs and soft actuation. • Driving bio-inspired actuation within the energy budget defined by the output of a single MFC, thereby improving the effective fusion of MFCs and soft robotics. • Showing energy autonomy through the integration of these technologies in a swimming, artificial organism, powered by an artificial digestive system and exploiting soft robotic actuation. The presented artificial organism has demonstrated feasibly application in self-powered environmental monitoring and clean-up of polluted waterways. The study shows a crucial step in the development of bio-inspired autonomous robots capable of long term self-sustainability and presents significant scope for future development.

No description available.