This thesis is concerned with the problem of place recognition for a mobile robot using an omnidirectional camera as its sole sensor modality. The problems we are faced with range from orientation estimation to loop closure detection, in the absence of any prior knowledge of position. In order to resolve the challenging issues encountered by any appearance-based place recognition system - specifically, perceptual aliasing and variability - we first develop a quadtree-based image comparison method. In contrast to most existing methods, this method does not involve the computationally expensive step of feature or keypoint detection and description, which utilises the spatial structure property of an image to provide robustness against dynamic changes in scenes. Ouralgorithm is experimentally evaluated on one public dataset, and two datasets collected by ourselves in different environments, thereby demonstrating its effectiveness in handling perceptual aliasing and environment variability. For many tasks in mobile robotics, it is crucial accurately to determine the orientation of the robot, relying on a single vision sensor. For this purpose, we propose an evaluation methodology that focuses on the ability of different image-based algorithms to establish the heading of the robot when capturing two images. Critical analysis of performance is also provided. In addition, a quadtree-based loop closure detection method is proposed, with the intention of increasing the number of correctly-recognized revisited locations (high recall) at low false positives (high precision). The loop closure detection is performed by pairwise image comparison. The performance of the proposed method is evaluated using our collected dataset, which contains highly aliased images and drastic perceptual changes. The experimental results show that our method can achieve a high recall at 100% precision, and outperform other related algorithms in term of closeness to ground truth.

Underwater exploration is an important area of research, whether to expand our knowledge base of underwater environments, or to monitor and preserve structures both natural and man-made. Biomimetic fish-like robots are being developed to incorporate some of the aspects of live fish in the hope that more efficient machines can be produced by exploiting the energy available within flow phenomena such as vortices. However, in order for a fully autonomous robot to be able to swim in natural environments, a navigational strategy is needed that distinguishes between steady and unsteady flows. To develop an effective strategy, the link between the local flow fields measured by sensors on the robot and the global flow phenomena needs to be determined. Therefore, the objective of this study was to link the local and global flow fields around a fish-like robot with an on-board pressure sensor array in both steady and unsteady flows. Two hypotheses were put forward: that the signal from pressure sensors towards the front of a fusiform-shape robot could be used to detect the relative position of vortices; and that the changing aspects of the Karman vortex street could be identified by changing aspects of the pressure signals detected. These hypotheses were tested by placing the fish-like robot in steady flows and in Karman vortex streets, both whilst held stationary and whilst moving. The global flow field was measured using DPIV and the local flow field was measured using the pressure sensor array on-board the robot. The results showed characteristic patterns within the pressure signals when the robot was within the Karman vortex street. By combining different characteristics of the pressure signals, detection of the Karman vortex street was possible, with the relative position and characteristics of the vortices able to be determined. It is hoped that the results of this study could be used within navigational strategies for fish-like robots in order that they can identify and successfully navigate unsteady flows such as the Karman vortex street, with the potential for this to reduce the energy usage of the robot in these flows.

In the emerging world of human-robot interaction, social robotics has become ever more important. Social robotics is a fundamental area in such domains as healthcare and medical robotics, consumer robotics or service robotics. Social robots working among humans should be able to communicate naturally with people using not only verbal but also non-verbal signals. Some cues of non-verbal body language, associated with affect and emotions, have an evolutionary root in humans that allows them to signal their unobservable internal state and intentions to others around them. An ability to interpret an internal state and intentions of team-members or counterparts is important not only in human-human but also in human-robot teams. One possible way to contribute to understandability is for robots to make their otherwise unobservable internal state interpretable to people through the use of emotionally expressive body language. This makes robots more predictable, acceptable and likeable, thus, in the end, having a potential to make them more effective team-players. This thesis addresses the problem of enabling humans to better understand machines by examining the role of artificial emotions synthesized and expressed by robots in the process of human-robot interaction. In our first study, we probe whether it is possible to signal a wanted emotional meaning through bodily expressions of a non-humanoid robot. The results provide strong support for the potential utility of bodily expressions in robots for communicating emotional meaning to people. A set of design parameters was developed from an analysis of research on non-verbal expression of emotion in the animal world. In the next two studies, we explore how this set of design parameters impacts how people perceive the emotional meaning of a robot expression, and investigate the nature and dynamics of peoples' perception of emotion expressed in a robot through its bodily movements. The results provide the basis for a mapping between the different design parameters of a robot's bodily expression and emotional interpretations. In addition, the results of the study show that people perceive emotionally expressive robots as more anthropomorphic, more animate, more likeable, more responsible and even more intelligent. In two next studies we investigate two major factors that may have influence on the perception of robot emotions. In one study, we investigate how the particular situational context in which expressions are used by the robot influences how they are perceived and interpreted by people. Another major factor to investigate is how the morphology of a robot performing emotional expressions influences how these expressions are interpreted and whether people are consistent in the emotional meaning they perceive. Finally, having a coherent design scheme to produce meaningful emotional expressions through robot body movements, we investigate in our last study the impact of such expressions on people's attitudes towards a robot. The results of the work provide evidence of the impact of the robot's emotional expressiveness on the perception of their anthropomorphism, animacy, likeability and intelligence. Results of our work are discussed in terms of the utility of expressive behaviour for facilitating human understanding of robot intentions and the directions for the future development in the design of cues for emotionally expressive robot behaviour.

This thesis examines a balance between designer effort required in biasing a robot's learning of a task, and the effort required from an experienced agent in influencing the learning using social interactions, and the effect of this balance on learning performance. In order to characterise this balance, a two dimensional design space is identified, where the dimensions represent the effort from the designer, who abstracts the robot's raw sensorimotor data according to the salient parts of the task to increasing degrees, and the effort from the experienced agent, who interacts with the learner robot using increasing degrees of complexities to actively accentuate the salient parts of the task and explicitly communicate about them. While the influence from the designer must be imposed at design time, the influence from the experienced agent can be tailored during the social interactions because this agent is situated in the environment while the robot is learning. The design space is proposed as a general characterisation of robotic systems that learn from social interactions. The usefulness of the design space is shown firstly by organising the related work into the space, secondly by providing empirical investigations of the effect of the various influences on the robot's experiences and how learning performance varies as a function of these influences, and finally by identifying how the conclusions from these investigations apply to the related work for improving learning performance. The empirical investigations implement different learning approaches, and are conducted with simulated and physical mobile robots learning wall-following and phototaxis tasks from an experienced simulated robot or an experienced human, and with a simulated humanoid robot learning an object-interaction task from an experienced simulated robot. The design space is used not only to characterise these investigations and related work, but also to characterise a typical performance surface that can be used to guide the design of new and existing systems. The characterisation shows that a particular level of performance can be maintained by compensating one source of influence for the other, and that performance can generally be improved by increasing the influence from any of these sources. It also shows that the best performance depends on various factors that affect the robot's overall learning potential, such as the available learning resources. The thesis argues that characterising the balance between designer effort and social interactions and how learning performance is affected is crucial for addressing a difficult trade-off: increasing designer effort for biasing the learning of a particular task in a particular environment and thus providing more reliability, versus increasing the influence from the social interactions thus providing more generality.

The operational performance criteria of fluidic oscillators are described in relation to the requirements of a domestic water meter. The problems associated with developing a novel water meter using fluidic oscillatory technology are discussed and the performance enhancements required to develop a fluidic oscillator capable of meeting the BS5728 (1979) domestic water meter specifications are presented. A sensing configuration is described which provides adequate sensitivity over the range required for a water meter with a nominal flowrate of one cubic meter per hour. The nozzled imensions are investigated to reduce the pressure drop across the fluidic water meter whilst still maintaining the required turndown range and adequate sensitivity at low flowrates. The development of a novel fluidic oscillator flow conditioning device is described which radically improves the linearity of the fluidic oscillator and helps to reduce susceptibility to upstream disturbances. The device allows debris to pass through the meter without causing blockage and has an acceptable low pressure drop. Modifications to the fluidic oscillator transducer geometry are investigated which reduce the minimum point of oscillation, thus increasing the turndown range of the water meter. Also geometry modifications are developed which improve the strength of the jet oscillation at low flowrates and thereby significantly increase the strength of signal. The effects of geometry modification on meter linearity and meter factor response are investigated and a transducer design with enhanced range performance and improved linearity is described. Transducer designs are presented which are capable of meeting the BS5728 (1979) specification for both Class C and Class D QNLO domestic water meters.

Swann robotics is a new approach to coordinate the behaviours of large number of relatively simple robots in decentralised manner. As the robots in the swann have only local perception and very limited local communication abilities, one of the challenges in designing swann robotic systems with desired collective behaviour is to understand the effect of individual behaviour on the group performance. This thesis dedicates the research on design and optimisation of interaction rules for a group of foraging robots that try to achieve energy efficiency collectively. The investigation starts with designing a set of interaction rules for the individual robots, inspired from the widely observed self-organisation phenomenon in biological system, so as improve the energy efficiency at the group level. A threshold-based controller, using two internal time thresholds - resting time and searching time threshold, is introduced to regulate the behaviours for the robot in order to improve the energy efficiency. Three cues: internal cues, social cues and environmental cues are then proposed to adjust the internal time thresholds in a self-organised manner. A number of strategies have been developed by combining these three cues and applied to the collective foraging task. Although the simulation results show that the robot swarm with adaptation mechanisms has the ability to guide the system towards energy optimisation collectively, there are difficulties in manually finding a set of parameters for the adaptation algorithm which can lead to the best energy efficiency under certain environmental conditions. This thesis focuses most of its effort into developing a macroscopic probabilistic model to understand the effect of individual parameters (internal time thresholds) on the performance of the system and therefore help to design the adaptation algorithm more efficiently. The modelling work is divided in two stages: A simplified situation for a swarm of homogeneous foraging robots without adaptation mechanism is considered first, then the macroscopic probabilistic model is extended for a robot swann with full adaptation ability. 3 The essential idea of the probabilistic modelling approach is to treat the interactions among robots, or between robots and environment, as stochastic events. First, a probabilistic finite state machine (PFSM), adapted from the robot controller, is used to describe the foraging task at the group level. A number of difference equations are then developed to capture the change of number of robot in each state. The state transition probabilities and other parameters used in the model are obtained through a novel geometrical approach, which makes sure that no free parameters exist in the model. In addition, the adaptation rules are encoded into the difference equations by introducing the concept of private resting/searching time thresholds and public resting/searching time thresholds. The proposed macroscopic model has been validated using simulation. The results show that the model achieves very good accuracy in predicting the net energy of the swarm, not only in the final stage but also in the instantaneous level. Finally, with the extended macroscopic model, a real-coded steady-state genetic algorithm (GA) is introduced to simplify the process of parameters selection for the adaptation algorithms. Experiments are carried out using the the best set of parameters found by the GA. It shows that the robot swarm with selected parameters can achieve a near-optimal energy efficiency under different environmental conditions.

This thesis is concerned with the control system analysis of steel plate inspection robots and the design and development of co-operative mobile micro-robots and for the intelligent coordinated control over inspection tasks. This involves miniaturization of the existing steel plate inspection system and the introduction of intelligent control over a novel path-planning algorithm for single or several similar cooperative micro inspection robots to perform multitasking inspection. To automate steel plate inspection, the Centre for Automated & Robotic – NDT (CART) department of London South Bank University successfully developed a self navigating steel plate inspection robot [1]. However the resulting system has complex hardware arrangements for inspection. Therefore, this research is concerned with reducing the size as well the complexity of the inspection system and the reduction of the inspection time with larger area of coverage. Accordingly, cooperative micro-robots are realized for inspection. Part (1) of the thesis includes related and researched work for the control system analysis. The research findings lead to the identification of the directed parallel method (DPM) for the inspection robot using a derived cost equation. Consequently, research identified the work area sensory environment (SE) having a camera fixed permanently at a height above the work area. Research developed a DPM algorithm to find any coordinate location of the multitasking inspection robots on a (planar) rectangular inspection area. Research developed an efficient intelligent control to automate steel plate inspection which is required to enable reliable control operations over the micro-robots. Part (2) reports simulation of cooperative robots, the development of two micro-robot prototypes and their real time movements on the work area. The robot’s path and speed are complied with the inspection standards. With the two cooperative micro-robot prototypes, the results show that for a larger area of coverage the inspection time can be reduced to more than half compared to that of the CART’s previous single complex robot.

The deterioration observed in many industrial systems may be modelled in two phases. In the first phase, a period during which the system operates fault free ends with entry into a worn state. In the second phase, the system spends time in the worn state prior to failure. Should the system be found to be in the worn state upon inspection, failure can be pre-empted by preventive maintenance. The first goal of analysis is the design of cost effective policies for the inspection, repair, and renewal of such systems. The thesis extends previous work by offering a choice between a (cheap) repair and a (more expensive) renewal of the system, should it be found to be in the worn state upon inspection. The decision-maker may also renew the system at any time without inspection. Simple, cost effective heuristic policies are proposed, whose design avoids the computational complexities of a full dynamic programming (DP) solution. The second goal of analysis is to determine when deployment of an error-prone sensor may be beneficial to the operation of such systems. It is supposed that a system is monitored continuously by such a sensor, which returns a positive result should entry into the worn state be detected. The sensor may produce errors of both kinds, false-positive and false-negative. Extending the earlier work of the thesis, simple, cost effective heuristics are developed for use with the sensor. In se- -i 11 lected cases, numerical investigation identifies operational characteristics for which use ofa sensor is (i) cost indifferent, (ii) beneficial, and (iii) not beneficial. The question of how sensor quality impacts upon heuristic design is also investigated. To the author's knowledge, the model proposed in this section of the thesis is new to the literature.

Autonomous systems are playing increasingly important roles in today's world. Technological advancements have allowed autonomous applications in many areas such as ground robotics (including factory robots), unmanned aerial vehicles (UAVs), unmanned underwater vehicles, unmanned spacecrafts etc. UAVs are relatively a new inclusion into the broader field of autonomous systems. The possibility of using UAVs for a diverse range of beneficial applications such as fire fighting, search and rescue missions, combating crime, etc. is a major motivational factor in this research program. Thus mission related work presented in this thesis, although generic In nature, relates heavily to UAVs. Mission planning involves many dimensions In general and the fundamental aspects explored in this research program i.e. route planning, resource allocation optimization, vehicle selection and mission collaboration, in particular. At the end of this research program a novel graph theoretic routing algorithm that offers deconfliction and efficient route computation has been developed and its performance experimentally examined using computer simulations. In addition, novel exhaustive as well as sub-optimal resource allocation mission planning schemes have been developed and applied to both deterministic and stochastic input variables. Furthermore the significant benefits obtained during a mission due to collaboration between acting UAVs have been examined and experimentally demonstrated. This collaborative behaviour has been achieved via the inclusion of data communication between UAVs and the mission Control Centre and in a way that allows for UAVs to coordinate their time of arrival to destination and thus maximize mission success.

This thesis presents novel contributions to the design, modelling and physical implementation of soft biomimetic actuators for operation in liquid environments. Single DOF and multi DOF monolithic actuators are designed by exploiting kirigami and electrical multi-segmentation techniques. Single degree of freedom actuators, Burstbot and Vonibot, are designed capable of generating complex biologically-inspired actuation profiles resembling the flexion of the mammalian cuspid valve and the coiled contractions of the Vorticella respectively_ The symmetric and asymmetric fluid interact ions of the Burstbot are investigated and the effectiveness in fluid transport applications is demonstrated. The Vortibot actuator is geometrically optimized as a camera positioner capable of 360 degree scanning. An artificial cilium actuator is developed based on quantitatively mimicking the structural design and stroke planar kinematics of the natural cilium. This actuator is modelled on the cilia movement of the alga Volvox, and represents the cilium as a piecewise constant-curvature robotic actuator that enables the subsequent direct translation of natural articulation into a multi-segment ionic polymer metal composite actuator. It is demonstrated how the combination of optimal segmentation pattern and biologically derived per-segment driving signals reproduce natural ciliary motion. The amenability of the artificial cilia to scaling is also demonstrated through the comparison of the Reynolds number achieved with that of natural cilia.