This thesis presents a novel distributed control method which allows a collection of independently mobile robotic units, with two or three dimensional movement, to self-assemble into self-repairing hierarchical structures. The proposed method utilises a simple model of the cellular adhesion mechanisms observed in biological cells, allowing the robotic units to form virtually bonded aggregates which behave as predicted by Steinberg’s differential adhesion hypothesis. Simulated robotic units based on the design of the subaquatic HYDRON module are introduced as a possible platform on which the model can be implemented. The units are used to carry out a detailed investigation of the model behaviour and parameter space focusing on the two main tasks of rounding and sorting in both two and three dimensions. These tasks assess the model’s ability to reach a thermodynamically stable configuration when the aggregates consist of either a single population of units or multiple populations of units with differing adhesive properties. The results are analysed in detail with particular attention given to the role of random movements in determining the overall performance, and demonstrate that this model provides a very robust solution to these complex tasks. Finally, a possible extension of this work is presented in which the original model is combined with a genetic regulatory network controller. The performance of this composite is evaluated, and the benefits of this hybrid approach, in which a powerful control system manipulates a robust self-organising behaviour, are discussed.

The task of developing a framework for which agents can communicate reliably and flexibly in open systems is not trivial. This thesis addresses the dichotomy between reliable communication and facilitation of the autonomy of agents to create more flexible and emergent interactions. By the introduction of adaptations to a distributed protocol language, agents benefit from the ability to communicate interaction protocols to elucidate the social norms (thus creating more reliable communication). Yet, this approach also provides the functionality for the agent to unilaterally introduce new paths for the conversation to explore unforeseen opportunities and options (thus restoring more autonomy than possible with static protocols). The foundation of this work is Lightweight Coordination Calculus (LCC). LCC is a distributed protocol language and framework in which agents coordinate their own interactions by their message passing activities. In order to ensure that adaptations to the protocols are done in a reasonable way, we examine the use of two models of communication to guide any transformations to the protocols. We describe the use of FIPA's ACL and ultimately its unsuitability for this approach as well as the more fecund task of implementing dialogue games, an model of argumentation, as dynamic protocols. The existing attempts to develop a model that can encompass the gulf between reliability and autonomy in communication have had varying degrees of success. It is the purpose and the result of the research described in this thesis to develop an alloy of the various models, by the introduction of dynamic and distributed protocols, to develop a framework stronger than its constituents. Though this is successful, the derivations of the protocols can be dificult to reconstruct. To this end, this thesis also describes a method of protocol synthesis inspired by models of human communication that can express the dialogues created by the previous approaches but also have a fully accountable path of construction. Not only does this thesis explore a unique and novel approach to agent communication, it is tested through a practical implementation.

Abstract Organic semiconductors (OSCs) and graphene are two classes of conjugated materials that hold promise to create flexible electronic displays, high speed transistors, and low-cost solar cells. Crucial to understanding the behavior of these materials is understanding the effects metallic contacts have on the local charge environment. Additionally, characterizing the charge carrier transport behavior within these materials sheds light on the physical mechanisms behind transport. The first part of this thesis examines the origin of the low-temperature, high electric field transport behavior of OSCs. Two chemically distinct OSCs are used, poly-3(hexylthiophene) (P3HT) and 6,13- bis(triisopropyl-silylethynyl) (TIPS) pentacene. Several models explaining the low-temperature behavior are presented, with one using the Tomonaga-Luttinger liquid (TLL) insulator-to-metal transition model and one using a field-emission hopping model. While the TLL model is only valid for 1-dimensional systems, it is shown to work for both P3HT (1D) and TIPS-pentacene (2D), suggesting the TLL model is not an appropriate description of these systems. Instead, a cross-over from thermally-activated hopping to field-emission hopping is shown to explain the data well. The second part of this thesis focuses on the interaction between gold and platinum contacts and graphene using suspended graphene over sub-100 nanometer channels. Contacts to graphene can strongly dominate charge transport and mobility as well as significantly modify the charge environment local to the contacts. Platinum electrodes are discovered to be strong dopants to graphene at short length scales while gold electrodes do not have the same effect. By increasing the separation distance between the electrodes, this discrepancy is shown to disappear, suggesting an upper limit on charge diffusion from the contacts. Finally, this thesis will discuss a novel technique to observe the high-frequency behavior in OSCs using two microwave sources and an organic transistor as a mixer. A theoretical model motivating this technique is presented which suggests the possibility of retrieving gigahertz charge transport phenomena at kilohertz detection frequencies. The current state of the project is presented and discrepancies between devices made with gold and platinum electrodes measured in the GHz regime are discussed.

The majority of automatic speech recognition (ASR) systems rely on hidden Markov models (HMM), in which the output distribution associated with each state is modelled by a mixture of diagonal covariance Gaussians. Dynamic information is typically included by appending time-derivatives to feature vectors. This approach, whilst successful, makes the false assumption of framewise independence of the augmented feature vectors and ignores the spatial correlations in the parametrised speech signal. This dissertation seeks to address these shortcomings by exploring acoustic modelling for ASR with an application of a form of state-space model, the linear dynamic model (LDM). Rather than modelling individual frames of data, LDMs characterize entire segments of speech. An auto-regressive state evolution through a continuous space gives a Markovian model of the underlying dynamics, and spatial correlations between feature dimensions are absorbed into the structure of the observation process. LDMs have been applied to speech recognition before, however a smoothed Gauss-Markov form was used which ignored the potential for subspace modelling. The continuous dynamical state means that information is passed along the length of each segment. Furthermore, if the state is allowed to be continuous across segment boundaries, long range dependencies are built into the system and the assumption of independence of successive segments is loosened. The state provides an explicit model of temporal correlation which sets this approach apart from frame-based and some segment-based models where the ordering of the data is unimportant. The benefits of such a model are examined both within and between segments. LDMs are well suited to modelling smoothly varying, continuous, yet noisy trajectories such as found in measured articulatory data. Using speaker-dependent data from the MOCHA corpus, the performance of systems which model acoustic, articulatory, and combined acoustic-articulatory features are compared. As well as measured articulatory parameters, experiments use the output of neural networks trained to perform an articulatory inversion mapping. The speaker-independent TIMIT corpus provides the basis for larger scale acoustic-only experiments. Classification tasks provide an ideal means to compare modelling choices without the confounding influence of recognition search errors, and are used to explore issues such as choice of state dimension, front-end acoustic parametrization and parameter initialization. Recognition for segment models is typically more computationally expensive than for frame-based models. Unlike frame-level models, it is not always possible to share likelihood calculations for observation sequences which occur within hypothesized segments that have different start and end times. Furthermore, the Viterbi criterion is not necessarily applicable at the frame level. This work introduces a novel approach to decoding for segment models in the form of a stack decoder with A* search. Such a scheme allows flexibility in the choice of acoustic and language models since the Viterbi criterion is not integral to the search, and hypothesis generation is independent of the particular language model. Furthermore, the time-asynchronous ordering of the search means that only likely paths are extended, and so a minimum number of models are evaluated. The decoder is used to give full recognition results for feature-sets derived from the MOCHA and TIMIT corpora. Conventional train/test divisions and choice of language model are used so that results can be directly compared to those in other studies. The decoder is also used to implement Viterbi training, in which model parameters are alternately updated and then used to re-align the training data.

An intelligent agent must be capable of using its past experience to develop an understanding of how its actions affect the world in which it is situated. Given some objective, the agent must be able to effectively use its understanding of the world to produce a plan that is robust to the uncertainty present in the world. This thesis presents a novel computational framework called the Adaptive Modelling and Planning System (AMPS) that aims to meet these requirements for intelligence. The challenge of the agent is to use its experience in the world to generate a model. In problems with large state and action spaces, the agent can generalise from limited experience by grouping together similar states and actions, effectively partitioning the state and action spaces into finite sets of regions. This process is called abstraction. Several different abstraction approaches have been proposed in the literature, but the existing algorithms have many limitations. They generally only increase resolution, require a large amount of data before changing the abstraction, do not generalise over actions, and are computationally expensive. AMPS aims to solve these problems using a new kind of approach. AMPS splits and merges existing regions in its abstraction according to a set of heuristics. The system introduces splits using a mechanism related to supervised learning and is defined in a general way, allowing AMPS to leverage a wide variety of representations. The system merges existing regions when an analysis of the current plan indicates that doing so could be useful. Because several different regions may require revision at any given time, AMPS prioritises revision to best utilise whatever computational resources are available. Changes in the abstraction lead to changes in the model, requiring changes to the plan. AMPS prioritises the planning process, and when the agent has time, it replans in high-priority regions. This thesis demonstrates the flexibility and strength of this approach in learning intelligent behaviour from limited experience.

As AI has begun to reach out beyond its symbolic, objectivist roots into the embodied, experientialist realm, many projects are exploring different aspects of creating machines which interact with and respond to the world as humans do. Techniques for visual processing, object recognition, emotional response, gesture production and recognition, etc., are necessary components of a complete humanoid robot. However, most projects invariably concentrate on developing a few of these individual components, neglecting the issue of how all of these pieces would eventually fit together. The focus of the work in this dissertation is on creating a framework into which such specific competencies can be embedded, in a way that they can interact with each other and build layers of new functionality. To be of any practical value, such a framework must satisfy the real-world constraints of functioning in real-time with noisy sensors and actuators. The humanoid robot Cog provides an unapologetically adequate platform from which to take on such a challenge. This work makes three contributions to embodied AI. First, it offers a general-purpose architecture for developing behavior-based systems distributed over networks of PC's. Second, it provides a motor-control system that simulates several biological features which impact the development of motor behavior. Third, it develops a framework for a system which enables a robot to learn new behaviors via interacting with itself and the outside world. A few basic functional modules are built into this framework, enough to demonstrate the robot learning some very simple behaviors taught by a human trainer. A primary motivation for this project is the notion that it is practically impossible to build an "intelligent" machine unless it is designed partly to build itself. This work is a proof-of-concept of such an approach to integrating multiple perceptual and motor systems into a complete learning agent.

An ecosytem-based spatial conservation plan for the South African sandy beaches. Sandy beaches are valuable ecosystems. They support a collection of species that is unique, comprising many endemic species, and provide a number of key ecosystem goods and services, including scenic vistas for human recreation, nesting sites for turtles and birds, and important areas for biogeochemical recycling, water filtration and purification. However, sandy beaches have not been well understood or appreciated as ecosystems, and consequently have a legacy of poor coastal management. In many instances this has lead to a "tyranny of small decisions", where multiple, seemingly insignificant management decisions and actions have resulted in complete transformation and degradation of the shoreline in several places. In addition to inappropriate management strategies, beaches are also poorly represented in conservation areas. Further, where they are recognised as being "conserved" in marine protected areas, this often is a false sense of protection because the far more sensitive dune portion of the littoral active zone is invariably not included in the reserve. In short, there is a need for a new way to approach sandy beach conservation and management that includes the system (dunes, intertidal beaches and surf zones) as a whole. On one hand, the approach should make provision for use of the abundant natural resources and opportunities associated with sandy shores in ways that are sustainable and contribute to biodiversity stewardship - through ecosystem-based management and marine spatial planning. But, on the other hand, it must simultaneously contribute to securing a sufficient amount of the key ecological attributes of beaches (habitats, biodiversity and processes) in a network of reserves, to ensure that the ecosystem, natural resources, and services all persist in perpetuity - through systematic conservation planning. The aim of this Thesis is to integrate these into a single approach, which I call ecosystem-based spatial conservation planning for sandy beaches, using the South African sandy shores as a case study. To achieve this broad aim, the Thesis is divided into three parts. Part 1 deals with establishing baseline information by quantifying spatial patterns in sandy beach habitats (Chapter 1), biodiversity, key assemblages and processes, and outstanding physical features (Chapter 2). First, mapping sandy beach habitats is a challenge given the vast, linear extent of shorelines and significant resources required to complete the project. Therefore, a novel approach was derived using statistical techniques (conditional inference trees) to identify physical features of beaches that can be observed on Google Earth (or similar) imagery, and that can provide good predictions of beach morphodynamic (habitat) types. Based on the results of this analysis, sandy beaches (and all other coastal habitat types) were mapped digitally in ArcGIS. Second, spatial patterns in sandy beach biodiversity (vertebrates, macrofauna, microflora and foredune plants) were mapped by compiling existing data on the distributions of key species that have been well studied or mapped previously (vertebrates and foredune plants), and by niche modelling (macrofauna and microflora). For the latter, data from all previous sandy-beach sampling events in South Africa were compiled from published and unpublished sources, and supplemented with additional sampling of 23 beaches along the national shoreline, targeting macrofauna and phytoplankton. Altogether, the macrofauna database comprised data from 135 sites and 186 sampling events, and the microflora (phytoplankton and microphytobenthos) database comprised data from 73 sites and 510 samples. The probabilistic distribution of each "resident" species (present at 10 or more sites) was modelled in MaxEnt version 3.3.3k, probability thresholds were determined statistically (to convert the data into predicted presence-absence), and displayed as a digital map. A composite biodiversity map was compiled, and key trends in species richness and endemism along the national shoreline were quantified. To supplement biodiversity proper, additional valued-features of sandy beaches were mapped, including: important assemblages; unique habitat features; and sites associated with key ecological processes. Part 2 considers threats to sandy beaches in the context of deriving an appropriate management strategy that seeks to provide for use of the coast, but in a way that has least overall impact to the ecosystem. A method for assessing cumulative threats to sandy beaches is adapted from an existing framework (Chapter 4). This entailed compiling a list of threats to beaches, and scoring these (out of 10) in terms of the severity of their respective impacts to beaches, and how long it would take the ecosystem to recover should the threat be removed. The scoring was based on the collective expert opinion of the scientific community working on sandy beaches, at a workshop during the VIth International Sandy Beach Symposium 2012. To standardize the scores and ensure broad applicability, a base case scenario of a pristine beach was established, and maximum theoretical scores were provided for this context. The method for integrating these scores into a spatial, cumulative threat assessment was then determined. In Chapter 5, the maximum theoretical scores (from Chapter 4) were down-scaled to suit the current threat regime to the South African sandy beaches, and the cumulative threat assessment methodology was applied. From this analysis, the most threatened beaches in South Africa, and the most important threats were highlighted. A decision-support tool for managers was derived from the site-specific cumulative threat-impact scores, based first on the degree of permanent habitat transformation, and second on the cumulative impact of other stressors where the impacts these stressors have could potentially be mitigated or ameliorated. Part 3 concerns conservation of beaches explicitly. It addresses how much of which valued features of beaches is required to ensure their long-term persistence, and the design of a network of beaches in South Africa that are of ecological importance and should be set aside as reserves. Conservation targets are set in Chapter 6, using species-area curves to determine a baseline percentage-area required to protect sandy beach habitats, which is modified using heuristic principles based on habitat rarity and threat status (from a recent national assessment). A fixed target was applied to all species, also modified by heuristic principles, and another fixed target was applied to key assemblages and processes.

Direct fitness is measured as the number of surviving offspring. Thus adult males may try to produce as many offspring as possible or to increase the survival of their offspring. Recent findings have shown the many potential benefits of fathers’ presence and support on infants’ development and survival. However, little is known about the influence of socio-ecological factors on male-infant interactions. The main aim of this thesis was therefore to investigate male-infant interactions in wild crested macaques (Macaca nigra). In particular, we aimed to examine the affiliative and agonistic interactions taking place in this species, along with the factors influencing these interactions and offspring survival. Data collection for this thesis took place in the Tangkoko-Duasudara Reserve in Sulawesi, Indonesia, on 3 wild groups of crested macaques. For the first study, data were collected on migrations, births, disappearances, and encounters between groups over 5 years. We analyzed the influence of socio-ecological factors (e.g. rainfall, alpha-male position takeover, and male hierarchy stability) on pre- and post-natal loss. The results showed that high infant mortality was mainly associated to male alpha-position takeover, which suggests that infanticide may indeed occur in this species. In addition, we found that female within-group competition for food sources and between-group resource defense influenced fetal and infant loss. Based on these findings, we were interested to see whether fathers protected their own offspring against male attacks. Thus, in the second study, we investigated the social determinants and characteristics of male-infant affiliations. Our results indicate that adult males and infants form preferential association, and that infants initiate the majority of male-infant affiliations. Infants initiated affiliations mainly towards a high ranking male or a male in a close relationship with their mother. In addition, infants affiliated mainly with adult males in the absence of their mother, while males affiliated mostly with infants when the infants‘ mother was present in proximity. Furthermore, males initiated affiliations towards an infant when they held a high rank or when they had a strong bond with the infant‘s mother. Interestingly, paternity did not affect male-infant affiliations. In conclusion, these studies provide insights in the specifics of both infant survival strategies and male reproductive strategies. In addition, we show that infants are active agents in establishing and maintaining preferential relationships with males. This thesis, thus, confirm that male-infant interactions, although rare, have a strong influence both on males’ and infants’ direct fitness.:Contents ........................................................................................................................................ 5 List of Figures .............................................................................................................................. 7 List of Tables ................................................................................................................................ 8 Summary ....................................................................................................................................... 9 Zusammenfassung .................................................................................................................. 13 1 General Introduction .......................................................................................................... 17 1.1 Infants and adult males in mammals ...................................................................... 18 1.2 Primate males’ use and abuse of infants ................................................................ 19 1.3 Male care and paternal care in primates ............................................................... 20 1.4 Crested macaques as study species ......................................................................... 21 1.5 Aims of this thesis ........................................................................................................... 23 2 Social and ecological factors influencing offspring survival in wild macaques ................................................................................................................................... 25 2.1 Abstract ............................................................................................................................... 26 2.2 Introduction ...................................................................................................................... 26 2.3 Methods .............................................................................................................................. 29 2.4 Results ................................................................................................................................. 34 2.5 Discussion .......................................................................................................................... 37 3 Mother-male bond, but not paternity, influences male-infant affiliation in wild crested macaques .......................................................................................................... 45 3.1 Abstract ............................................................................................................................... 46 3.2 Introduction ...................................................................................................................... 46 3.3 Methods .............................................................................................................................. 50 3.4 Results ................................................................................................................................. 58 3.5 Discussion .......................................................................................................................... 61 4 Thesis Conclusion ................................................................................................................ 71 Appendices ................................................................................................................................ 75 Supplementary figure and tables for Chapter 2 ......................................................... 76 Supplementary methods and tables for Chapter 3 ................................................... 79 Bibliography ............................................................................................................................. 83 Acknowledgements ................................................................................................................ 99 Contributions of co-authors .............................................................................................. 101 Curriculum vitae .................................................................................................................... 105 Publications and conference contributions ................................................................. 107 Selbstständigkeitserklärung ............................................................................................. 109

The document is the PhD thesis of Stanislav Nikitin. It summarizes results of the comprehensive investigation of the magnetic fluctuations in several quantum magnets by means of elastic and inelastic neutron scattering.

This thesis examines the parasite dynamics and the mechanisms affecting parasite load and transmission focalising on the role played by host and habitat heterogeneities. This study is based on the gastrointestinal nematode Heligmosomoides polygyrus and the small mammal yellow necked mouse and uses data gathered from experimental field manipulations of parasites intensities and data gathered from trapping monitoring. Initially the parasite community of yellow-necked mouse (Apodemus flavicollis) was explored in North-Eastern Italian Alps with the aim to describe the major patterns and identify the factors affecting parasite community structure. Despite the observed spatial variability it has been found that differences within the host population such age and secondly sex and breeding conditions, were the major factors acting on parasite occurrence and intensity. Habitat differences had a less apparent effect on parasite community structure. The consequences of H. polygyrus infection on other parasite species infections have been analysed, in specific the infestation of the tick Ixodes ricinus in populations of A. flavicollis. H. polygyrus load and tick infestation were monitored as well as were carried out field manipulations of H. polygyrus intensity and were monitored changes in tick infestation. It has been found that H. polygyrus load was negatively related to I. ricinus infestations. Host factors mediated the H. polygyrus-I. ricinus interaction such that young and non-breeding mice exhibited higher I. ricinus to H. polygyrus intensity respect breeding adults. The role of host sex on parasite abundance was then investigated carrying out a field experiment where the H. polygyrus intensity were manipulated in relation to mice gender. In specific, H. polygyrus was removed alternately from either sexes and the parasite load was analysed in the untreated sex. It was found that males mice were responsible to drive parasite transmission in the host population and this was observed in absence of sex-bias in parasite infection, suggesting that this pattern was not a mere consequence of quantitative differences in parasite loads between sexes. To disentangle the possible mechanism causing this sex bias in parasite transmission mathematical simulations based on parameters obtained for the field experiment were used. Two non mutually exclusive hypotheses causing sex bias in parasite transmission were tested: a- males immune response is less efficient and this causes the development of more successful parasite infective stages or b-males behaviours allow them to be more efficient is spreading in more exposed areas parasite infective stages. Multi-host models were developed and simulations were compared with field results. While it was not disentangled the most dominant mechanism causing sex bias in parasite transmission this study underlined the importance of host sexes in affecting parasite dynamics and host-parasite interaction. In conclusion this thesis highlighted the importance of considering host and environmental differences when investigating host parasite interactions. This finding could be extremely important when planning measured of disease control or to avoid disease outbreak. Controlling target group of individuals host could avoid economical losses and a more effective measure of intervention.