The unit for information security and IT architecture at The Swedish Defence Research Agency (FOI) conducts work with a cyber range called CRATE (Cyber Range and Training Environment). Currently, simulation of user activity involves scripts inside the simulated network. This solution is not ideal because of the traces it leaves in the system and the general lack of standardised GUI API between different operating systems. FOI are interested in testing the use of artificial user agent located outside the virtual environment using computer vision and the virtualisation API to execute actions and extract information from the system. This paper focuses on analysing the reliability of template matching, a computer vision algorithm used to localise objects in images using already identified images of said object as templates. The analysis will evaluate both the reliability of localising objects and the algorithms ability to correctly identify if an object is present in the virtual environment. Analysis of template matching is performed by first creating a prototype of the agent's sensory system and then simulate scenarios which the agent might encounter. By simulating the environment, testing parameters can be manipulated and monitored in a reliable way. The parameters manipulated involves both the amount and type of image noise in the template and screenshot, the agent’s discrimination threshold for what constitutes a positive match, and information about the template such as template generality. This paper presents the performance and reliability of the agent in regards to what type of image noise affects the result, the amount of correctly identified objects given different discrimination thresholds, and computational time of template matching when different image filters are applied. Furthermore the best cases for each study are presented as comparison for the other results. In the end of the thesis we present how for screenshots with objects very similar to the templates used by the agent, template matching can result in a high degree of accuracy in both object localization and object identification and that a small reduction of similarity between template and screenshot to reduce the agent's ability to reliably identifying specific objects in the environment.

Philosophiae Doctor - PhD / The underlying theme of this thesis is identification, classification, detection and prediction of cellular network faults using state of the art technologies, methods and algorithms.

Indiana University-Purdue University Indianapolis (IUPUI) / Multi-agent systems consist of multiple agents that interact and coordinate with each other to work towards to certain goal. Multi-agent systems naturally arise in a variety of domains such as robotics, telecommunications, and economics. The dynamic and complex nature of these systems entails the agents to learn the optimal solutions on their own instead of following a pre-programmed strategy. Reinforcement learning provides a framework in which agents learn optimal behavior based on the response obtained from the environment. In this thesis, we propose various novel de- centralized, learning automaton based algorithms which can be employed by a group of interacting learning automata. We propose a completely decentralized version of the estimator algorithm. As compared to the completely centralized versions proposed before, this completely decentralized version proves to be a great improvement in terms of space complexity and convergence speed. The decentralized learning algorithm was applied; for the first time; to the domains of distributed object tracking and distributed watershed management. The results obtained by these experiments show the usefulness of the decentralized estimator algorithms to solve complex optimization problems. Taking inspiration from the completely decentralized learning algorithm, we propose the novel concept of partial decentralization. The partial decentralization bridges the gap between the completely decentralized and completely centralized algorithms and thus forms a comprehensive and continuous spectrum of multi-agent algorithms for the learning automata. To demonstrate the applicability of the partial decentralization, we employ a partially decentralized team of learning automata to control multi-agent Markov chains. More flexibility, expressiveness and flavor can be added to the partially decentralized framework by allowing different decentralized modules to engage in different types of games. We propose the novel framework of heterogeneous games of learning automata which allows the learning automata to engage in disparate games under the same formalism. We propose an algorithm to control the dynamic zero-sum games using heterogeneous games of learning automata.

This study leverages a critical provocation design approach to examine user interactions with intelligent agents, specifically focusing on how non-conventional agent behaviours impact user perceptions. By embedding playful, challenging, and mischievous elements into the user experience, the research aims to uncover insights that traditional methods might miss. The experimental design involved participants interacting with a trivia game agent named Trivia Friend, which intentionally provided false feedback to provoke reactions and gain insights on user perceptions. Key findings highlight the emotional spectrum elicited by the agent’s behaviour, ranging from frustration and mistrust to amusement and engagement. The study reveals that user perceptions of fairness and communication style are influenced by the agent’s provocations. Furthermore, the research underscores the importance of managing user expectations. A provocative design can stimulate engagement. However, real world implementations of intelligent agents must be designed with fairness and transparency to ensure positive user experiences. The study calls for incorporating efforts towards emotional understanding, clear communication, and ethical considerations when implementing socially capable intelligent agents. This research contributes to the development of adaptive, user-friendly, and ethically sound intelligent AI based agents by offering valuable insights into the complex dynamics of human-agent interactions.

Thesis (M. Tech.) -- Central University of Technology, Free State, 2010 / Traditional corrective maintenance is both costly and ineffective. In some situations it is more cost effective to replace a device than to maintain it; however it is far more likely that the cost of the device far outweighs the cost of performing routine maintenance. These device related costs coupled with the profit loss due to reduced production levels, makes this reactive maintenance approach unacceptably inefficient in many situations. Blind predictive maintenance without considering the actual physical state of the hardware is an improvement, but is still far from ideal. Simply maintaining devices on a schedule without taking into account the operational hours and workload can be a costly mistake. The inefficiencies associated with these approaches have contributed to the development of proactive maintenance strategies. These approaches take the device health state into account. For this reason, proactive maintenance strategies are inherently more efficient compared to the aforementioned traditional approaches. Predicting the health degradation of devices allows for easier anticipation of the required maintenance resources and costs. Maintenance can also be scheduled to accommodate production needs. This work represents the design and simulation of an intelligent maintenance management system that incorporates device health prognosis with maintenance schedule generation. The simulation scenario provided prognostic data to be used to schedule devices for maintenance. A production rule engine was provided with a feasible starting schedule. This schedule was then improved and the process was determined by adhering to a set of criteria. Benchmarks were conducted to show the benefit of optimising the starting schedule and the results were presented as proof. Improving on existing maintenance approaches will result in several benefits for an organisation. Eliminating the need to address unexpected failures or perform maintenance prematurely will ensure that the relevant resources are available when they are required. This will in turn reduce the expenditure related to wasted maintenance resources without compromising the health of devices or systems in the organisation.

This dissertation investigates high-level decision making for agents that are both goal and utility driven. We develop a partially observable Markov decision process (POMDP) planner which is an extension of an agent programming language called DTGolog, itself an extension of the Golog language. Golog is based on a logic for reasoning about action—the situation calculus. A POMDP planner on its own cannot cope well with dynamically changing environments and complicated goals. This is exactly a strength of the belief-desire-intention (BDI) model: BDI theory has been developed to design agents that can select goals intelligently, dynamically abandon and adopt new goals, and yet commit to intentions for achieving goals. The contribution of this research is twofold: (1) developing a relational POMDP planner for cognitive robotics, (2) specifying a preliminary BDI architecture that can deal with stochasticity in action and perception, by employing the planner. / Computing / M. Sc. (Computer Science)

The ability to change one's beliefs in a rational manner is one of many facets of the abilities of an intelligent agent. Central to any investigation of belief change is the notion of an epistemic state. This dissertation is mainly concerned with three issues involving epistemic states: 1. How should an epistemic state be represented? 2. How does an agent use an epistemic state to perform belief change? 3. How does an agent arrive at a particular epistemic state? With regard to the first question, note that there are many different methods for constructing belief change operations. We argue that semantic constructions involving ordered pairs, each consisting of a set of beliefs and an ordering on the set of "possible worlds" (or equivalently, on the set of basic independent bits of information) are, in an important sense, more fundamental. Our answer to the second question provides indirect support for the use of semantic structures. We show how well-known belief change operations and related structures can be modelled semantically. Furthermore, we introduce new forms of belief change related operations and structures which are all defined, and motivated, in terms of such semantic representational formalisms. These include a framework for unifying belief revision and nonmonotonic reasoning, new versions of entrenchment orderings on beliefs, novel approaches to withdrawal operations, and an expanded view of iterated belief change. The third question is. one which has not received much attention in the belief change literature. We propose to extract extra-logical information from the formal representation of an agent's set of beliefs, which can then be used in the construction of epistemic state. his proposal is just a first approximation, although it seems to have the potential for developing into a full-fledged theory. / Computing / D.Phil.(Computer Science)

Agentes inteligentes é um termo guarda-chuva que agrega diversas pesquisas no desenvolvimento de softwares autônomos que utilizam técnicas de Inteligência Artificial a fim de satisfazer metas estabelecidas por seus usuários. A construção de sistemas baseados em agentes inteligentes é uma tarefa complexa que envolve aspectos como comunicação entre agentes, planejamento, divisão de tarefas, coordenação, representação e manipulação de conhecimento e comportamentos, entre outras tarefas. Plataformas para agentes prevêem alguns serviços que permitem a desenvolvedores construir soluções sem a necessidade de se preocupar com todos detalhes da implementação. Um novo modelo para criação de agentes chamado 'agentes atômicos' é proposto com o objetivo de oferecer flexibilidade para o gerenciamento de conhecimento e implementação de comportamentos. A arquitetura Agentes Semânticos provê um framework para a implementação de tal modelo, oferecendo um conjunto de ferramentas para a criação de agentes inteligentes. Um protótipo de plataforma para agentes, baseado em tal arquitetura, foi desenvolvido em Java e permite a criação de aplicações capazes de processar linguagem natural restrita, manipular conhecimento e executar ações úteis. / Intelligent Agents is an umbrella term that aggregates different research on the development of autonomous software that uses Artificial Intelligence techniques in order to satisfy user requests. The construction of systems based on intelligent agents is a complex task that involves aspects such as agent communication, planning, work division, cooperation, epresentation and manipulation of knowledge,among other activities. Agent Platforms provide some services that allow developers to build solutions without the need of worrying about every implementation detail. A new model for creating agents, called 'atomic agents', is proposed with the goal of offering flexible knowledge management and behavior implementation for constructing software agents. The Semantic AgentArchitecture provides a framework for the implementation of such model, offering a set of tools for the creation of intelligent agents. A prototype Agent Platform, based on the architecture, was developed in Java and allows the creation of applications that are able to process restricted natural language, manipulate knowledge and execute useful actions.

L'experiència de l'autor en la temàtica d'agents intel·ligents i la seva aplicació als robots que emulen el joc de futbol han donat el bagatge suficient per poder encetar i proposar la temàtica plantejada en aquesta tesi: com fer que un complicat robot pugui treure el màxim suc de l'autoconeixement de l'estructura de control inclosa al seu propi cos físic, i així poder cooperar millor amb d'altres agents per optimitzar el rendiment a l'hora de resoldre problemes de cooperació. Per resoldre aquesta qüestió es proposa incorporar la dinàmica del cos físic en les decisions cooperatives dels agents físics unificant els móns de l'automàtica, la robòtica i la intel·ligència artificial a través de la noció de capacitat: la capacitat vista com a entitat on els enginyers de control dipositen el seu coneixement, i a la vegada la capacitat vista com la utilitat on un agent hi diposita el seu autoconeixement del seu cos físic que ha obtingut per introspecció. En aquesta tesi es presenta l'arquitectura DPAA que s'organitza seguint una jerarquia vertical en tres nivells d'abstracció o mòduls control, supervisor i agent, els quals presenten una estructura interna homogènia que facilita les tasques de disseny de l'agent. Aquests mòduls disposen d'un conjunt específic de capacitats que els permeten avaluar com seran les accions que s'executaran en un futur. En concret, al mòdul de control (baix nivell d'abstracció) les capacitats consisteixen en paràmetres que descriuen el comportament dinàmic i estàtic que resulta d'executar un controlador determinat, és a dir, encapsulen el coneixement de l'enginyer de control. Així, a través dels mecanismes de comunicació entre mòduls aquest coneixement pot anar introduint-se als mecanismes de decisió dels mòduls superiors (supervisor i agent) de forma que quan els paràmetres dinàmics i estàtics indiquin que pot haver-hi problemes a baix nivell, els mòduls superiors es poden responsabilitzar d'inhibir o no l'execució d'algunes accions. Aquest procés top-down intern d'avaluació de la viabilitat d'executar una acció determinada s'anomena procés d'introspecció. Es presenten diversos exemples per tal d'il·lustrar com es pot dissenyar un agent físic amb dinàmica pròpia utilitzant l'arquitectura DPAA com a referent. En concret, es mostra tot el procés a seguir per dissenyar un sistema real format per dos robots en formació de comboi, i es mostra com es pot resoldre el problema de la col·lisió utilitzant les capacitats a partir de les especificacions de disseny de l'arquitectura DPAA. Al cinquè capítol s'hi exposa el procés d'anàlisi i disseny en un domini més complex: un grup de robots que emulen el joc del futbol. Els resultats que s'hi mostren fan referència a l'avaluació de la validesa de l'arquitectura per resoldre el problema de la passada de la pilota. S'hi mostren diversos resultats on es veu que és possible avaluar si una passada de pilota és viable o no. Encara que aquesta possibilitat ja ha estat demostrada en altres treballs, l'aportació d'aquesta tesi està en el fet que és possible avaluar la viabilitat a partir de l'encapsulament de la dinàmica en unes capacitats específiques, és a dir, és possible saber quines seran les característiques de la passada: el temps del xut, la precisió o inclòs la geometria del moviment del robot xutador. Els resultats mostren que la negociació de les condicions de la passada de la pilota és possible a partir de capacitats atòmiques, les quals inclouen informació sobre les característiques de la dinàmica dels controladors. La complexitat del domini proposat fa difícil comparar els resultats amb els altres treballs. Cal tenir present que els resultats mostrats s'han obtingut utilitzant un simulador fet a mida que incorpora les dinàmiques dels motors dels robots i de la pilota. En aquest sentit cal comentar que no existeixen treballs publicats sobre el problema de la passada en què es tingui en compte la dinàmica dels robots.El present treball permet assegurar que la inclusió de paràmetres dinàmics en el conjunt de les capacitats de l'agent físic permet obtenir un millor comportament col·lectiu dels robots, i que aquesta millora es deu al fet que en les etapes de decisió els agents utilitzen informació relativa a la viabilitat sobre les seves accions: aquesta viabilitat es pot calcular a partir del comportament dinàmic dels controladors. De fet, la definició de capacitats a partir de paràmetres dinàmics permet treballar fàcilment amb sistemes autònoms heterogenis: l'agent físic pot ser conscient de les seves capacitats d'actuació a través de mecanismes interns d'introspecció, i això permet que pugui prendre compromisos amb altres agents físics.

The discourse that surrounds and constitutes the post-human emerged as a response to earlier claims of an essential or universal human or human nature. These discussions claim that the human is a discursive construct that emerges from various configurations of nature, embodiment, technology, and culture, configurations that have also been variously shaped by the forces of social history. And in the absence of an essential human figure, post-human discourses suggest that there are no restrictions or limitations on how the human can be reconfigured. This axiom has been extended in light of a plethora of technological reconfigurations and augmentations now potentially available to the human, and claims emerge from within this literature that these new technologies constitute a range of possibilities for future human biological evolution. This thesis questions the assumption contained within these discourses that technological incursions or reconfigurations of the biological human necessarily constitute human biological or human social evolution by discussing the role the evolution theories plays in our understanding of the human, the social, and technology. In this thesis I show that, in a reciprocal process, evolution theory draws metaphors from social institutions and ideologies, while social institutions and ideologies simultaneously draw on metaphors from evolution theory. Through this discussion, I propose a form of evolution literacy; a tool, I argue, is warranted in developing a sophisticated response to changes in both human shape and form. I argue that, as a whole, our understanding of evolution constitutes a metanarrative, a metaphor through which we understand the place of the human within the world; it follows that historical shifts in social paradigms will result in new definitions of evolution. I show that contemporary evolution theory reflects parts of the world as codified informatic systems of associated computational network logic through which the behaviour of participants is predefined according to an evolved or programmed structure. Working from within the discourse of contemporary evolution theory I develop a space through which a version of the post-human figure emerges. I promote this version of the post-human as an Artificial Intelligence computational programme or autonomous agent that, rather than seeking to replace, reduce or deny the human subject, is configured as an exosomatic supplement to and an extension of the biological human.