Debugging multi-agent systems, which are concurrent, distributed, and consist of complex components is difficult, yet crucial. The development of these complex systems is supported by agent-oriented software engineering methodologies which utilise agents as the central design metaphor. The systems that are developed are inherently complex since the components of these systems may interact in flexible and sophisticated ways and traditional debugging techniques are not appropriate. Despite this, very little effort has been applied to developing appropriate debugging tools and techniques. Debugging multi-agent systems without good debugging tools is highly impractical and without suitable debugging support developing and maintaining multi-agent systems will be more difficult than it need be. In this thesis we propose that the debugging process can be supported by following an agent-oriented design methodology, and then using the developed design artifacts in the debugging phase. We propose a domain independent debugging framework which comprises the developed processes and components that are necessary in using design artifacts as debugging artifacts. Our approach is to take a non-formal design artifact, such as an AUML protocol design, and encode it in a machine interpretable manner such that the design can be used as a model of correct system behaviour. These models are used by a run-time debugging system to compare observed behaviour against specified behaviour. We provide details for transforming two design artifact types into equivalent debugging artifacts and show how these can be used to detect bugs. During a debugging episode in which a bug has been identified our debugging approach can provide detailed information about the possible reason for the bug occurring. To determine if this information was useful in helping to debug programs we undertook a thorough empirical study and identified that use of the debugging tool translated to an improvement in debugging performance. We conclude that the debugging techniques developed in this thesis provide effective debugging support for multi-agent systems and by having an extensible framework new design artifacts can be explored and as translations are developed they can be added to the debugging system.

Doctor of Philosophy / Department of Computing and Information Sciences / Scott A. DeLoach / Organization-based Multiagent Systems (OMAS) have been viewed as an effective paradigm for addressing the design challenges posed by today’s complex systems. In those systems, the organizational perspective is the main abstraction, which provides a clear separation between agents and systems, allowing a reduction in the complexity of the overall system. To ease the development of OMAS, several methodologies have been proposed. Unfortunately, those methodologies typically require the designer to handle system complexity alone, which tends to lead to ad-hoc designs that are not scalable and are difficult to maintain. Moreover, designing organizations for large multiagent systems is a complex and time-consuming task; design models quickly become unwieldy and thus hard to develop. To cope with theses issues, a framework for organization-based multiagent system designs based on separation of concerns and composition principles is proposed. The framework uses category theory tools to construct a formal composition framework using core models from the Organization-based Multiagent Software Engineering (O-MASE) framework. I propose a formalization of these models that are then used to establish a reusable design approach for OMAS. This approach allows designers to design large multiagent organizations by reusing smaller composable organizations that are developed separately, thus providing them with a scalable approach for designing large and complex OMAS. In this dissertation, the process of formalizing and composing multiagent organizations is discussed. In addition, I propose a service-oriented approach for building autonomous, adaptive multiagent systems. Finally, as a proof of concept, I develop two real world examples from the domain of cooperative robotics and wireless sensor networks.

This thesis investigates existing game development methodologies, through the process of researching game and system development models. The results indicate that these methodologies are engineered to solve specific problems, and most are suitable only for specific game genres. Different approaches to building games have been proposed in recent years. However, most of these methodologies focus on the design and implementation phase. This research aims to enhance game development methodologies by proposing a novel game development methodology, with the ability to function in generic game genres, thereby guiding game developers and designers from the start of the game development phase to the end of the implementation and testing phase. On a positive note, aligning development practice with universal standards makes it far easier to incorporate extra team members at short notice. This increased the confidence when working in the same environment as super developers. In the gaming industry, most game development proceeds directly from game design to the implementation phase, and the researcher observes that this is the only industry in which this occurs. It is a consequence of the game industry’s failure to integrate with modern development techniques. The ultimate aim of this research to apply a new game development methodology using most game elements to enhance success. This development model will align with different game genres, and resolve the gap between industry and research area, so that game developers can focus on the important business of creating games. The primary aim of Agent Oriented Agile Base (AOAB) game development methodology is to present game development techniques in sequential steps to facilitate game creation and close the gap in the existing game development methodologies. Agent technology is used in complex domains such as e-commerce, health, manufacturing, games, etc. In this thesis we are interested in the game domain, which comprises a unique set of characteristics such as automata, collaboration etc. Our AOAB will be based on a predictive approach after adaptation of MaSE methodology, and an adaptive approach using Agile methodology. To ensure proof of concept, AOAB game development methodology will be evaluated against industry principles, providing an industry case study to create a driving test game, which was the problem motivating this research. Furthermore, we conducted two workshops to introduce our methodology to both academic and industry participants. Finally, we prepared an academic experiment to use AOAB in the academic sector. We have analyzed the feedbacks and comments and concluded the strengths and weakness of the AOAB methodology. The research achievements are summarized and proposals for future work outlined.

A tecnologia de agentes surge como uma solução que fornece flexibilidade e robustez para lidar com domínios dinâmicos e complexos. Tal flexibilidade pode ser alcançada através da adoção de abordagens já existentes baseadas em agentes, como a arquitetura BDI, que provê agentes com características mentais de crenças, desejos e intenções. Essa arquitetura é altamente personalizável, deixando lacunas a serem preenchidas de acordo com aplicações específicas. Uma dessas lacunas é o algoritmo de seleção de planos, responsável por selecionar um plano para ser executado pelo agente buscando atingir um objetivo, e tendo grande influência no desempenho geral do agente. Grande parte das abordagens existentes requerem considerável esforço para personalização e ajuste a fim de serem utilizadas em aplicações específicas. Nessa dissertação, propomos uma abordagem para seleção de planos apta a aprender quais planos possivelmente terão os melhores resultados, baseando-se no contexto atual e nas preferências do agente. Nossa abordagem é composta por um meta-modelo, que deve ser instanciado a fim de especificar metadados de planos, e uma técnica que usa tais metadados para aprender e predizer resultados da execução destes planos. Avaliamos nossa abordagem experimentalmente e os resultados indicam que ela é efetiva. Adicionalmente, fornecemos uma ferramenta para apoiar o processo de desenvolvimento de agentes de software baseados em nosso trabalho. Essa ferramenta permite que desenvolvedores modelem e gerem código-fonte para agentes BDI com capacidades de aprendizado. Um estudo com usuários foi realizado para avaliar os benefícios de um método de desenvolvimento baseado em agentes BDI auxiliado por ferramenta. Evidências sugerem que nossa ferramenta pode auxiliar desenvolvedores que não sejam especialistas ou que não estejam familiarizados com a tecnologia de agentes. / Agent technology arises as a solution that provides flexibility and robustness to deal with dynamic and complex domains. Such flexibility can be achieved by the adoption of existing agent-based approaches, such as the BDI architecture, which provides agents with the mental attitudes of beliefs, desires and intentions. This architecture is highly customisable, leaving gaps to be fulfilled in particular applications. One of these gaps is the plan selection algorithm that is responsible for selecting a plan to be executed by an agent to achieve a goal, having an important influence on the overall agent performance. Most existing approaches require considerable effort for customisation and adjustment to be used in particular applications. In this dissertation, we propose a plan selection approach that is able to learn plans that provide possibly best outcomes, based on current context and agent’s preferences. Our approach is composed of a meta-model, which must be instantiated to specify plan metadata, and a technique that uses such metadata to learn and predict plan outcomes. We evaluated our approach experimentally, and results indicate it is effective. Additionally, we provide a tool to support the development process of software agents based on our work. This tool allows developers to model and generate source code for BDI agents with learning capabilities. A user study was performed to assess the improvements of a tool-supported BDI-agent-based development method, and evidences suggest that our tool can help developers that are not experts or are unfamiliar with the agent technology.

Industrial software systems are becoming more complex with a large number of interacting parts distributed over networks. Due to the inherent complexity in the problem domains, most such systems are modified over time to incorporate emerging requirements, making incremental development a suitable approach for building complex systems. In domain specific systems it is the domain experts as end users who identify improvements that better suit their needs. Examples include meteorologists who use weather modeling software, engineers who use control systems and business analysts in business process modeling. Most domain experts are not fluent in systems programming and changes are realised through software engineers. This process hinders the evolution of the system, making it time consuming and costly. We hypothesise that if domain experts are empowered to make some of the system changes, it would greatly ease the evolutionary process, thereby making the systems more effective. Agent Oriented Software Engineering (AOSE) is seen as a natural fit for modeling and implementing distributed complex systems. With concepts such as goals and plans, agent systems support easy extension of functionality that facilitates incremental development. Further agents provide an intuitive metaphor that works at a higher level of abstraction compared to the object oriented model. However agent programming is not at a level accessible to domain experts to capitalise on its intuitiveness and appropriateness in building complex systems. We propose a model driven development approach for domain experts that uses visual modeling and automated code generation to simplify the development and evolution of agent systems. Our approach is called the Component Agent Framework for domain-Experts (CAFnE), which builds upon the concepts from Model Driven Development and the Prometheus agent software engineering methodology. CAFnE enables domain experts to work with a graphical representation of the system , which is easier to understand and work with than textual code. The model of the system, updated by domain experts, is then transformed to executable code using a transformation function. CAFnE is supported by a proof-of-concept toolkit that implements the visual modeling, model driven development and code generation. We used the CAFnE toolkit in a user study where five domain experts (weather forecasters) with no prior experience in agent programming were asked to make changes to an existing weather alerting system. Participants were able to rapidly become familiar with CAFnE concepts, comprehend the system's design, make design changes and implement them using the CAFnE toolkit.

Software maintenance and evolution is arguably a lengthy and expensive phase in the life cycle of a software system. A critical issue at this phase is change propagation: given a set of primary changes that have been made to software, what additional secondary changes are needed to maintain consistency between software artefacts? Although many approaches have been proposed, automated change propagation is still a significant technical challenge in software maintenance and evolution. Our objective is to provide tool support for assisting designers in propagating changes during the process of maintaining and evolving models. We propose a novel, agent-oriented, approach that works by repairing violations of desired consistency rules in a design model. Such consistency constraints are specified using the Object Constraint Language (OCL) and the Unified Modelling Language (UML) metamodel, which form the key inputs to our change propagation framework. The underlying change propagation mechanism of our framework is based on the well-known Belief-Desire-Intention (BDI) agent architecture. Our approach represents change options for repairing inconsistencies using event-triggered plans, as is done in BDI agent platforms. This naturally reflects the cascading nature of change propagation, where each change (primary or secondary) can require further changes to be made. We also propose a new method for generating repair plans from OCL consistency constraints. Furthermore, a given inconsistency will typically have a number of repair plans that could be used to restore consistency, and we propose a mechanism for semi-automatically selecting between alternative repair plans. This mechanism, which is based on a notion of cost, takes into account cascades (where fixing the violation of a constraint breaks another constraint), and synergies between constraints (where fixing the violation of a constraint also fixes another violated constraint). Finally, we report on an evaluation of the approach, covering both effectiveness and efficiency.

La principal contribución de esta Tesis es la propuesta de un modelo de agente BDI graduado (g-BDI) que permita especificar una arquitetura de agente capaz de representar y razonar con actitudes mentales graduadas. Consideramos que una arquitectura BDI más exible permitirá desarrollar agentes que alcancen mejor performance en entornos inciertos y dinámicos, al servicio de otros agentes (humanos o no) que puedan tener un conjunto de motivaciones graduadas. En el modelo g-BDI, las actitudes graduadas del agente tienen una representación explícita y adecuada. Los grados en las creencias representan la medida en que el agente cree que una fórmula es verdadera, en los deseos positivos o negativos permiten al agente establecer respectivamente, diferentes niveles de preferencias o de rechazo. Las graduaciones en las intenciones también dan una medida de preferencia pero en este caso, modelan el costo/beneficio que le trae al agente alcanzar una meta. Luego, a partir de la representación e interacción de estas actitudes graduadas, pueden ser modelados agentes que muestren diferentes tipos de comportamiento. La formalización del modelo g-BDI está basada en los sistemas multi-contextos. Diferentes lógicas modales multivaluadas se han propuesto para representar y razonarsobre las creencias, deseos e intenciones, presentando en cada caso una axiomática completa y consistente. Para tratar con la semántica operacional del modelo de agente, primero se definió un calculus para la ejecución de sistemas multi-contextos, denominado Multi-context calculus. Luego, mediante este calculus se le ha dado al modelo g-BDI semántica computacional. Por otra parte, se ha presentado una metodología para la ingeniería de agentes g-BDI en un escenario multiagente. El objeto de esta propuesta es guiar el diseño de sistemas multiagentes, a partir de un problema del mundo real. Por medio del desarrollo de un sistema recomendador en turismo como caso de estudio, donde el agente recomendador tiene una arquitectura g-BDI, se ha mostrado que este modelo es valioso para diseñar e implementar agentes concretos. Finalmente, usando este caso de estudio se ha realizado una experimentación sobre la flexibilidad y performance del modelo de agente g-BDI, demostrando que es útil para desarrollar agentes que manifiesten conductas diversas. También se ha mostrado que los resultados obtenidos con estos agentes recomendadores modelizados con actitudes graduadas, son mejores que aquellos alcanzados por los agentes con actitudes no-graduadas. / The central contribution of this dissertation is the proposal of a graded BDI agent model (g-BDI), specifying an architecture capable of representing and reasoning with graded mental attitudes. We consider that making the BDI architecture more exible will allow us to design and develop agents capable of improved performance in uncertain and dynamic environments, serving other agents (human or not) that may have a set of graded motivations.In the g-BDI model, the agent graded attitudes have an explicit and suitable representation. Belief degrees represent the extent to which the agent believes a formula to be true. Degrees of positive or negative desires allow the agent to set di_erent levels of preference or rejection respectively. Intention degrees also give a preference measure but, in this case, modelling the cost/benefit trade off of achieving an agent's goal. Then, agents having different kinds of behaviour can be modelled on the basis of the representation and interaction of their graded attitudes. The formalization of the g-BDI agent model is based on Multi-context systems and in order to represent and reason about the beliefs, desires and intentions, we followed a many-valued modal approach. Also, a sound and complete axiomatics for representing each graded attitude is proposed. Besides, in order to cope with the operational semantics aspects of the g-BDI agent model, we first defined a Multi-context calculus for Multi-context systems execution and then, using this calculus we give this agent model computational meaning.Furthermore, a software engineering process to develop graded BDI agents in a multiagent scenario is presented. The aim of the proposed methodology is to guide the design of a multiagent system starting from a real world problem. Through the development of a Tourism recommender system, where one of its principal agents is modelled as a g-BDI agent, we show that the model is useful to design and implement concrete agents.Finally, using the case study we have made some experiments concerning the exibility and performance of the g-BDI agent model, demonstrating that this agent model is useful to develop agents showing varied and rich behaviours. We also show that the results obtained by these particular recommender agents using graded attitudes improve those achieved by agents using non-graded attitudes.

Agent-oriented software engineering is faced with challenges that impact on the adoption of agent technology by the wider software engineering community. This is generally due to lack of adequate comprehension of the concepts of agent technology. This thesis is based on the premise that the comprehension of the concepts of and the adoption of agent technology can be improved. Two approaches are explored: the first approach is the analysis and structuring of the interactions in multiagent systems; the second approach is sharing of experiences of what works and what does not in agent-oriented software engineering using software patterns. While analysis of interactions in multiagent systems improves the understanding of the behaviour of multiagent systems, sharing multiagent system development experience improves the understanding of the concepts of agent technology as well as the challenges that face the engineering of multiagent systems. It is therefore believed that interaction analysis and experience sharing can enhance the comprehension of agent technology and hence, the adoption of the technology by the wider community of software practitioners. This thesis addresses the challenges facing agent-oriented software engineering by presenting a dedicated approach for developing agent interaction protocols to guide the interactions in a multiagent system; and a comprehensive framework for classifying, analyzing and describing agent-oriented patterns for the purpose of sharing multiagent systems development experiences.

The thesis solves a challenging problem in military logistics for tasks such as transportation scheduling and combinatorial auctions. A conceptual model has been developed that captures the organisational business processes involved and an effective implementation suitable for computer software agents. The protocol facilitates planning and task allocation among organisations in decentralised, dynamic and open environments.

A tecnologia de agentes surge como uma solução que fornece flexibilidade e robustez para lidar com domínios dinâmicos e complexos. Tal flexibilidade pode ser alcançada através da adoção de abordagens já existentes baseadas em agentes, como a arquitetura BDI, que provê agentes com características mentais de crenças, desejos e intenções. Essa arquitetura é altamente personalizável, deixando lacunas a serem preenchidas de acordo com aplicações específicas. Uma dessas lacunas é o algoritmo de seleção de planos, responsável por selecionar um plano para ser executado pelo agente buscando atingir um objetivo, e tendo grande influência no desempenho geral do agente. Grande parte das abordagens existentes requerem considerável esforço para personalização e ajuste a fim de serem utilizadas em aplicações específicas. Nessa dissertação, propomos uma abordagem para seleção de planos apta a aprender quais planos possivelmente terão os melhores resultados, baseando-se no contexto atual e nas preferências do agente. Nossa abordagem é composta por um meta-modelo, que deve ser instanciado a fim de especificar metadados de planos, e uma técnica que usa tais metadados para aprender e predizer resultados da execução destes planos. Avaliamos nossa abordagem experimentalmente e os resultados indicam que ela é efetiva. Adicionalmente, fornecemos uma ferramenta para apoiar o processo de desenvolvimento de agentes de software baseados em nosso trabalho. Essa ferramenta permite que desenvolvedores modelem e gerem código-fonte para agentes BDI com capacidades de aprendizado. Um estudo com usuários foi realizado para avaliar os benefícios de um método de desenvolvimento baseado em agentes BDI auxiliado por ferramenta. Evidências sugerem que nossa ferramenta pode auxiliar desenvolvedores que não sejam especialistas ou que não estejam familiarizados com a tecnologia de agentes. / Agent technology arises as a solution that provides flexibility and robustness to deal with dynamic and complex domains. Such flexibility can be achieved by the adoption of existing agent-based approaches, such as the BDI architecture, which provides agents with the mental attitudes of beliefs, desires and intentions. This architecture is highly customisable, leaving gaps to be fulfilled in particular applications. One of these gaps is the plan selection algorithm that is responsible for selecting a plan to be executed by an agent to achieve a goal, having an important influence on the overall agent performance. Most existing approaches require considerable effort for customisation and adjustment to be used in particular applications. In this dissertation, we propose a plan selection approach that is able to learn plans that provide possibly best outcomes, based on current context and agent’s preferences. Our approach is composed of a meta-model, which must be instantiated to specify plan metadata, and a technique that uses such metadata to learn and predict plan outcomes. We evaluated our approach experimentally, and results indicate it is effective. Additionally, we provide a tool to support the development process of software agents based on our work. This tool allows developers to model and generate source code for BDI agents with learning capabilities. A user study was performed to assess the improvements of a tool-supported BDI-agent-based development method, and evidences suggest that our tool can help developers that are not experts or are unfamiliar with the agent technology.