Towards executable specification : combining i* and AgentSpeak(L).

Agent-oriented conceptual modelling (AoCM) approaches in Requirements Engineering (RE) have received considerable attention recently. Semi-formal modeling frameworks such as i* assist analysts in requirements elicitation and reasoning of early-phase RE. AgentSpeak(L) is a widely accepted agent programming language. The Strategic Rationale (SR) model of the i* framework naturally lends itself to AgentSpeak(L) programs. Furthermore, the Strategic Dependency (SD) component of the i* framework prescribes the interaction between the agents in a multi-agent environment. This paper proposes a formal methodology for transforming a SR model to an AgentS- peak(L) agent. The constructed AgentSpeak(L) agents will then form the essential components of a multi-agent system, MAS.

AgentSpeak+: AgentSpeak with Probabilistic Planning

AgentSpeak is a logic-based programming language, based on the Belief-Desire-Intention (BDI) paradigm, suitable for building complex agent-based systems. To limit the computational complexity, agents in AgentSpeak rely on a plan library to reduce the planning problem to the much simpler problem of plan selection. However, such a plan library is often inadequate when an agent is situated in an uncertain environment. In this paper, we propose the AgentSpeak+ framework, which extends AgentSpeak with a mechanism for probabilistic planning. The beliefs of an AgentSpeak+ agent are represented using epistemic states to allow an agent to reason about its uncertain observations and the uncertain effects of its actions. Each epistemic state consists of a POMDP, used to encode the agent’s knowledge of the environment, and its associated probability distribution (or belief state). In addition, the POMDP is used to select the optimal actions for achieving a given goal, even when facing uncertainty.

Probabilistic Planning in AgentSpeak using the POMDP framework.

AgentSpeak is a logic-based programming language, based on the Belief-Desire-Intention (BDI) paradigm, suitable for building complex agent-based systems. To limit the computational complexity, agents in AgentSpeak rely on a plan library to reduce the planning problem to the much simpler problem of plan selection. However, such a plan library is often inadequate when an agent is situated in an uncertain environment. In this paper, we propose the AgentSpeak+ framework, which extends AgentSpeak with a mechanism for probabilistic planning. The beliefs of an AgentSpeak+ agent are represented using epistemic states to allow an agent to reason about its uncertain observations and the uncertain effects of its actions. Each epistemic state consists of a POMDP, used to encode the agent’s knowledge of the environment, and its associated probability distribution (or belief state). In addition, the POMDP is used to select the optimal actions for achieving a given goal, even when facing uncertainty.

Leveraging new plans in AgentSpeak(PL)

In order to facilitate the development of agent-based software, several agent programming languages and architectures, have been created. Plans in these architectures are often self-contained procedures with an associated triggering event and a context condition, while any further information about the consequences of executing a plan is absent. However, agents designed using such an approach have limited flexibility at runtime, and rely on the designer’s ability to foresee all relevant situations an agent might have to handle. In order to overcome this limitation, we have created AgentSpeak(PL), an interpreter capable of performing state-space planning to generate new high-level plans. As the planning module creates new plans, the plan library is expanded, improving performance over time. However, for new plans to be useful in the long run, it is critical that the context condition associated with new plans is carefully generated. In this paper we describe a plan reuse technique aimed at improving an agent’s runtime performance by deriving optimal context conditions for new plans, allowing an agent to reuse generated plans as much as possible.

Programmazione ad Agenti e Pianificazione: AgentSpeak(PL) come caso di studio

In questa tesi si discutono inizialmente i concetti chiave di agente e sistema multi-agente e si descrivono in ogni dettaglio il linguaggio di programmazione AgentSpeak(L) e la piattaforma Jason, fornendo le basi per poter programmare con il paradigma AOP. Lo scopo centrale di questa tesi è quello di estendere il modello di pianificazione dell’interprete di AgentSpeak(L), considerato come caso specifico, con un approccio che può essere integrato in qualsiasi linguaggio di programmazione ad agenti. Si espone un’evoluzione di AgentSpeak(L) in AgentSpeak(PL), ossia la creazione ed esecuzione di piani automatici in caso di fallimento attraverso l'uso di un algoritmo di planning state-space. L'approccio integrativo modifica il Ciclo di Reasoning di Jason proponendo in fase di pianificazione automatica un riuso di piani già esistenti, atto a favorire la riduzione di tempi e costi nel long-term in un sistema multi-agente. Nel primo capitolo si discute della nozione di agente e delle sue caratteristiche principali mentre nel secondo capitolo come avviene la vera e propria programmazione con AgentSpeak(L). Avendo approfondito questi argomenti base, il terzo capitolo è incentrato sull’interprete Jason e il quarto su una migliore estensione dell'interprete, in grado di superare i limiti migliorando le performance nel tempo. Si delineano infine alcune considerazioni e ringraziamenti nel quinto e ultimo capitolo. Viene proposta con scrittura di carattere divulgativo e non ambiguo.

Can(Plan)+: Extending the Operational Semantics of the BDI Architecture to deal with Uncertain Information

The BDI architecture, where agents are modelled based on their beliefs, desires and intentions, provides a practical approach to develop large scale systems. However, it is not well suited to model complex Supervisory Control And Data Acquisition (SCADA) systems pervaded by uncertainty. In this paper we address this issue by extending the operational semantics of Can(Plan) into Can(Plan)+. We start by modelling the beliefs of an agent as a set of epistemic states where each state, possibly using a different representation, models part of the agent's beliefs. These epistemic states are stratified to make them commensurable and to reason about the uncertain beliefs of the agent. The syntax and semantics of a BDI agent are extended accordingly and we identify fragments with computationally efficient semantics. Finally, we examine how primitive actions are affected by uncertainty and we define an appropriate form of lookahead planning.

Context-dependent Combination of Sensor Information in Dempster-Shafer Theory for BDI

There has been much interest in the belief–desire–intention (BDI) agent-based model for developing scalable intelligent systems, e.g. using the AgentSpeak framework. However, reasoning from sensor information in these large-scale systems remains a significant challenge. For example, agents may be faced with information from heterogeneous sources which is uncertain and incomplete, while the sources themselves may be unreliable or conflicting. In order to derive meaningful conclusions, it is important that such information be correctly modelled and combined. In this paper, we choose to model uncertain sensor information in Dempster–Shafer (DS) theory. Unfortunately, as in other uncertainty theories, simple combination strategies in DS theory are often too restrictive (losing valuable information) or too permissive (resulting in ignorance). For this reason, we investigate how a context-dependent strategy originally defined for possibility theory can be adapted to DS theory. In particular, we use the notion of largely partially maximal consistent subsets (LPMCSes) to characterise the context for when to use Dempster’s original rule of combination and for when to resort to an alternative. To guide this process, we identify existing measures of similarity and conflict for finding LPMCSes along with quality of information heuristics to ensure that LPMCSes are formed around high-quality information. We then propose an intelligent sensor model for integrating this information into the AgentSpeak framework which is responsible for applying evidence propagation to construct compatible information, for performing context-dependent combination and for deriving beliefs for revising an agent’s belief base. Finally, we present a power grid scenario inspired by a real-world case study to demonstrate our work.

From SMART to agent systems development

n order for agent-oriented software engineering to prove effective it must use principled notions of agents and enabling specification and reasoning, while still considering routes to practical implementation. This paper deals with the issue of individual agent specification and construction, departing from the conceptual basis provided by the smart agent framework. smart offers a descriptive specification of an agent architecture but omits consideration of issues relating to construction and control. In response, we introduce two new views to complement smart: a behavioural specification and a structural specification which, together, determine the components that make up an agent, and how they operate. In this way, we move from abstract agent system specification to practical implementation. These three aspects are combined to create an agent construction model, actsmart, which is then used to define the AgentSpeak(L) architecture in order to illustrate the application of actsmart.

Interaction among agents that plan

The development of practical agent languages has progressed significantly over recent years, but this has largely been independent of distinct developments in aspects of multiagent cooperation and planning. For example, while the popular AgentSpeak(L) has had various extensions and improvements proposed, it still essentially a single-agent language. In response, in this paper, we describe a simple, yet effective, technique for multiagent planning that enables an agent to take advantage of cooperating agents in a society. In particular, we build on a technique that enables new plans to be added to a plan library through the invocation of an external planning component, and extend it to include the construction of plans involving the chaining of subplans of others. Our mechanism makes use of plan patterns that insulate the planning process from the resulting distributed aspects of plan execution through local proxy plans that encode information about the preconditions and effects of the external plans provided by agents willing to cooperate. In this way, we allow an agent to discover new ways of achieving its goals through local planning and the delegation of tasks for execution by others, allowing it to overcome individual limitations.

Extending Agent Languages for Autonomy

BDI agent languages provide a useful abstraction for complex systems comprised of interactive autonomous entities, but they have been used mostly in the context of single agents with a static plan library of behaviours invoked reactively. These languages provide a theoretically sound basis for agent design but are very limited in providing direct support for autonomy and societal cooperation needed for large scale systems. Some techniques for autonomy and cooperation have been explored in the past in ad hoc implementations, but not incorporated in any agent language. In order to address these shortcomings we extend the well known AgentSpeak(L) BDI agent language to include behaviour generation through planning, declarative goals and motivated goal adoption. We also develop a language-specific multiagent cooperation scheme and, to address potential problems arising from autonomy in a multiagent system, we extend our agents with a mechanism for norm processing leveraging existing theoretical work. These extensions allow for greater autonomy in the resulting systems, enabling them to synthesise new behaviours at runtime and to cooperate in non-scripted patterns.

Norm-based behaviour modification in BDI agents

While there has been much work on developing frameworks and models of norms and normative systems, consideration of the impact of norms on the practical reasoning of agents has attracted less attention. The problem is that traditional agent architectures and their associated languages provide no mechanism to adapt an agent at runtime to norms constraining their behaviour. This is important because if BDI-type agents are to operate in open environments, they need to adapt to changes in the norms that regulate such environments. In response, in this paper we provide a technique to extend BDI agent languages, by enabling them to enact behaviour modification at runtime in response to newly accepted norms. Our solution consists of creating new plans to comply with obligations and suppressing the execution of existing plans that violate prohibitions. We demonstrate the viability of our approach through an implementation of our solution in the AgentSpeak(L) language.

Linguaggi di programmazione ad agenti: Jason come caso di studio.

L'informatica è al giorno d'oggi al centro di un incredibile sviluppo e tumulto innovativo; è la scienza che coniuga ogni sviluppo tecnologico, il quale inevitabilmente, d'ora in poi, verrà in qualche modo condizionato dalle innovazioni di tale disciplina. Fra gli innumerevoli nuovi trend che si sono affacciati prepotentemente negli ultimi anni sul panorama informatico, il paradigma per la programmazione ad agenti è uno dei più interessanti, in accordo con i recenti e prossimi sviluppi della tecnologia in generale. Questa tesi tratterà tale argomento partendo da un punto di vista generico ed introduttivo volutamente esaustivo, per poi concentrarsi su una specifica tecnologia implementativa, ovvero il linguaggio Jason, mostrandola infine nella pratica con la presentazione di un'applicazione nel capitolo tre.

Programmazione di sistemi multi-agente: la piattaforma Jason come caso di studio

L’obiettivo principale di questo elaborato è di mostrare in un primo momento i concetti fondamentali che stanno alla base del paradigma ad agenti. Una volta introdotti, essi verranno collocati in un determinato ambiente di programmazione attraverso una piattaforma specifica chiamata Jason. Come sarà facile capire dalla lettura di questa trattazione, un sistema ad agenti è costituito dagli agenti stessi e dall’ambiente in cui sono situati. L’ambiente risulta quindi un altro tassello fondamentale ed è stato introdotto allo scopo un nuovo paradigma per la programmazione di ambienti chiamato Agent & Artifact. Nello specifico, verrà ampiamente descritto il framework di riferimento di tale paradigma: CArtAgO. Dopo aver illustrato i concetti e gli strumenti per poter agilmente programmare e progettare sistemi ad agenti, verrà infine mostrato un esempio di applicazione di tale tecnologia attraverso un case study. Il progetto del sistema in questione riguarda un reale caso aziendale e integra la tecnologia RFID con quella ad agenti per fornire la soluzione ad un problema noto come quello del controllo periodico delle scorte.