Skip to main content

Über dieses Buch

Since the mid 1980s, software agents and multi-agent systems have grown into a very active area of research and also commercial development activity. One of the limiting factors in industry take-up of agent-technology, however, is the lack of adequate software engineering support. The Agent-Oriented Software Engineering Workshop, AOSE, focuses on the synergies and cross fertilization between software engineering and agent research.

This volume presents both thoroughly revised selected papers from the AOSE 2010 workshop held at AAMAS 2010 in Toronto, Canada in May 2010 as well as invited articles by leading researchers in the field. The papers cover a broad range of topics related to software engineering and agent-based systems, with particular attention to the integration of concepts and techniques from multi-agent systems with conventional engineering approaches on the one hand, and to the integration of agent-oriented software engineering and methodologies with conventional engineering processes on the other hand.



Engineering Methods

Principles for Value-Sensitive Agent-Oriented Software Engineering

As software plays an increasingly important role in people’s lives, the impact it has on their values frequently becomes apparent. Many software design methods address “soft issues”, but very few address values explicitly. We present six principles that design methods should meet in order to properly deal with values. One area in which adherence to stakeholder values is important, is Agent-Oriented Software Engineering (AOSE). The Tropos AOSE method, with its concept of soft-goal, comes close to meeting our principles, but does not address values explicitly. Value-Sensitive Design is a methodology that does explicitly address value issues, but it offers little guidance in operationalizing them. We discuss a case study in which we attempt to capture values in Tropos’ soft-goals after eliciting them using Value-Sensitive Design. Subsequently, we discuss to what extent Tropos adheres to our principles. Finally, we propose the introduction of values as a first-class entity in Tropos in order to meet our aims of dealing with values.
Christian Detweiler, Koen Hindriks, Catholijn Jonker

Analyzing Contract Robustness through a Model of Commitments

We address one of the challenges in developing solutions based on multiagent systems for the problems of cross-organizational business processes and commerce generally. Specifically, we study how to gather and analyze requirements embodied within business contracts using the abstractions from multiagent systems.
Commerce is driven by business contracts. Each party to a business contract must be assured that the contract is robust, in the sense that it fulfills its goals and avoids undesirable outcomes. However, real-life business contracts tend to be complex and unamenable both to manual scrutiny and domain-independent scientific methods, making it difficult to provide automated support for determining or improving their robustness. As a result, establishing a contract is nontrivial and adds significantly to the transaction costs of conducting business. If the adoption of multiagent systems approaches in supporting business interactions is to be viable, we need to develop appropriate techniques to enable tools to reason about contracts in relation to their robustness.
To this end, we propose a powerful approach to assessing the robustness of contracts, and make two main contributions. First, we demonstrate a novel conceptual model for contracts that is based on commitments. Second, we offer a methodology for (i) creating commitment-based models of contracts from textual descriptions, and (ii) evaluating the contract models for robustness. We validate these contributions via a study of real-world contracts.
Amit K. Chopra, Nir Oren, Sanjay Modgil, Nirmit Desai, Simon Miles, Michael Luck, Munindar P. Singh

A Case for New Directions in Agent-Oriented Software Engineering

The state-of-the-art of Agent-oriented Software Engineering (AOSE) is insufficiently reflected in the state-of-practice in developing complex distributed systems. This paper discusses software engineering (SE) areas that have not been widely addressed in the context of AOSE, leading to a lack of mechanisms that support the development of Multiagent Systems (MASs) based on traditional SE principles, such as modularity, reusability and maintainability. This discussion is based on an exploratory study of the development of a family of buyer agents following the belief-desire-intention model and using a Software Product Line architecture. Based on the discussion presented in this paper, we hope to encourage the AOSE community to address particular SE issues on the development of MAS that have not yet been (widely) considered.
Ingrid Nunes, Donald Cowan, Elder Cirilo, Carlos J. P. de Lucena

Requirements Engineering and Testing

Engaging Stakeholders with Agent-Oriented Requirements Modelling

One advantage of using the agent paradigm for software engineering is that the concepts used for high-level modelling, such as roles, goals, organisations, and interactions, are accessible to many different stakeholders. Existing research demonstrates that including the stakeholders in the modelling of systems for as long as possible improves the quality of the development and final system because inconsistencies and incorrect behaviour are more likely to be detected early in the development process. In this paper, we propose three changes to the typical requirements engineering process found in AOSE methodologies, with the aim of including stakeholders over the requirements engineering process, effectively using stakeholders as modellers. These changes are: withholding design commitment, delaying the definition of the system boundary, and delaying the stakeholder “sign-off” of the requirements specification. We discuss our application of these changes to a project with an industry partner, and present anecdotal evidence to suggest that these changes can be effective in maintaining stakeholder involvement.
Tim Miller, Sonja Pedell, Leon Sterling, Bin Lu

Towards Requirement Analysis Pattern for Learning Agents

Learning is a capability that can be incorporated into software agents to handle the complexity of dynamic and unexpected situations, exploiting available artificial intelligence (AI) techniques. Despite design techniques for learning agents have been discussed in agent oriented software engineering literature, how to identify and analyze the requirements for learning agents is still poorly addressed. In this paper, we introduce a pattern for requirement analysis of learning agents. This analysis pattern contains a group of related, generic meta-classes of learning and their relations in a domain neutral manner which can be described as elements of conceptual modeling of learning requirement of agents. The applicability of the pattern has been investigated through the development of a book trading case study.
Shiva Vafadar, Ahmad Abdollahzadeh Barfourosh

Test Coverage Criteria for Agent Interaction Testing

By the very definition of complex systems, complex behaviour emerges from the interactions between the individual parts. This emergent behaviour may be difficult or impossible to predict by analysing the parts. As a result, systematic and thorough testing of the interactions of complex systems, including multi-agent systems, is an important part of the verification and validation process. This paper defines two sets of test coverage criteria for multi-agent interaction testing. The first uses only the protocol specification, while the second considers also the plans that generate and receive the messages in the protocol. We describe how an existing debugging agent can be used as a test oracle for assessing correctness of a test, and how the Petri Net representation of the debugging agent can be annotated to support test coverage measurements. This work both specifies, and shows how to measure, the degree of thoroughness of a set of test cases. It also provides a basis for the future specification of test case input, designed to provide good coverage.
Tim Miller, Lin Padgham, John Thangarajah

Model-Driven Approaches

Using ASEME Methodology for Model-Driven Agent Systems Development

This paper shows how an AOSE methodology, the Agent Systems Engineering Methodology (ASEME), uses state of the art technologies from the Model-Driven Engineering (MDE) domain. We present the Agent Modeling Language (AMOLA) metamodels and the model transformation tools that we developed and discuss our choices. Moreover, the way that non-functional requirements are used throughout the software development lifecycle is discussed and presented with two real-world case studies. Finally, we compare ASEME with a set of existing AOSE methodologies.
Nikolaos Spanoudakis, Pavlos Moraitis

Towards the Automatic Derivation of Malaca Agents Using MDE

The automatic transformation of software agent designs into implementations for different agent platforms is currently a key issue in the MAS development process. Recently several approaches have been proposed using model driven development concepts to specify generic agent metamodels and/or define a set of transformation rules from the design phase for different agent implementation platforms. However, all these approaches propose different sets of transformation rules for each target agent platform, thereby making the integration of new agent platforms more difficult. In this paper we propose transforming PIM4Agents, a generic agent metamodel used at the design phase, into Malaca, an agent specific platform-neutral metamodel for agents. With only one set of transformations it is possible to specify platform-neutral agents and to generate a partial implementation in Malaca, which can be executed on top of different FIPA compliant platforms.
Inmaculada Ayala, Mercedes Amor, Lidia Fuentes

$\mathcal{F}_{or}\mathcal{MAAD}$ : Towards a Model Driven Approach for Agent Based Application Design

Current trends in multi-agent systems development show a move towards adopting the Model Driven Architecture (MDA) approach to improve the development process and the quality of the agent-based software. Our work has two main contributions. First, it presents a reformulation of the \(\mathcal{F}_{or}\mathcal{MAAD}\) methodology in terms of the MDA paradigm by using the AML language. Second, it proposes a translation of each model to a formal language, \(\mathcal{T}_{emporal}\mathcal{Z}\) that integrates linear temporal logic to the Z notation, in order to guarantee a formal verification of the models. Furthermore, we make extensions to the StarUML tool to support the proposed models and use the transition rules. Our work is illustrated by developing an agent-based solution for the air traffic control problem
Zeineb Graja, Amira Regayeg, Ahmed Hadj Kacem

Software Architecture and Middleware

An Architectural Perspective on Multiagent Societies

This paper attempts to provide an architectural foundation to multiagent societies through a systematic application of the notion of software connector. It shows that multiagent societies can be explained as a Component & Connector architectural style, made up of high-level connectors defined in terms of common normative, communicative and organizational abstractions. This is expected to yield a better alignment of agent technology with mainstream software engineering practice and conventional architectural styles. Moreover, we show that connectors are a powerful metaphor for the design of organizational and communicative abstractions. Last, the paper challenges a common architectural assumption, namely the application-independence of software connectors.
Juan Manuel Serrano, Sergio Saugar

A Methodology for Developing an Agent Systems Reference Architecture

The slow adoption of agent-oriented methodologies as a paradigm for developing industry systems is due in part to their lack of integration and general-purpose use. There exists a need to define common patterns, relationships between components, and structural qualities that a reference architecture for agent-based systems would solve. However, there is little, if any, consensus on how to create a reference architecture for agent-based systems. This paper presents a methodology for developing a reference architecture that documents agent-based systems from different system viewpoints. Rather than the traditional approach of studying existing systems, the documentation methodology relies on forensic software analysis of agent frameworks (i.e., APIs and libraries for constructing agent systems). We demonstrate the methodology by describing the process used to create the Agent System Reference Architecture.
Duc N. Nguyen, Kyle Usbeck, William M. Mongan, Christopher T. Cannon, Robert N. Lass, Jeff Salvage, William C. Regli, Israel Mayk, Todd Urness

A Middleware Model in Alloy for Supply Chain-Wide Agent Interactions

To support the complex coordination activities involved in supply chain management, more and more companies have autonomous software agents acting on their behalf. Due to confidentiality concerns, such as hiding sensitive information from competitors, agents typically only have a local view on the supply chain. In many situations, however, companies would like to expand the view of their agents to share valuable information such as transportation tracking and service delays. Non of the participating companies, however, has enough knowledge or authority to realize such interactions in a controlled manner.
In this paper, we present an organization middleware that offers a collaboration platform and enables agents to interact across the boundary of local interactions. Policies and laws enable companies to define the scope of interactions of their agents and the restrictions on their exposed information. Using Alloy, we formally define the relation between the interactions offered by the middleware, the exposed information and the provided policies and laws. This allows us to guarantee a number properties which are of particular interest to companies using the middleware.
Robrecht Haesevoets, Danny Weyns, Mario Henrique Cruz Torres, Alexander Helleboogh, Tom Holvoet, Wouter Joosen

A Delegation-Based Architecture for Collaborative Robotics

Collaborative robotic systems have much to gain by leveraging results from the area of multi-agent systems and in particular agent-oriented software engineering. Agent-oriented software engineering has much to gain by using collaborative robotic systems as a testbed. In this article, we propose and specify a formally grounded generic collaborative system shell for robotic systems and human operated ground control systems. Collaboration is formalized in terms of the concept of delegation and delegation is instantiated as a speech act. Task Specification Trees are introduced as both a formal and pragmatic characterization of tasks and tasks are recursively delegated through a delegation process implemented in the collaborative system shell. The delegation speech act is formally grounded in the implementation using Task Specification Trees, task allocation via auctions and distributed constraint problem solving. The system is implemented as a prototype on Unmanned Aerial Vehicle systems and a case study targeting emergency service applications is presented.
Patrick Doherty, Fredrik Heintz, David Landén


Weitere Informationen

Premium Partner

BranchenIndex Online

Die B2B-Firmensuche für Industrie und Wirtschaft: Kostenfrei in Firmenprofilen nach Lieferanten, Herstellern, Dienstleistern und Händlern recherchieren.



Best Practices für die Mitarbeiter-Partizipation in der Produktentwicklung

Unternehmen haben das Innovationspotenzial der eigenen Mitarbeiter auch außerhalb der F&E-Abteilung erkannt. Viele Initiativen zur Partizipation scheitern in der Praxis jedoch häufig. Lesen Sie hier  - basierend auf einer qualitativ-explorativen Expertenstudie - mehr über die wesentlichen Problemfelder der mitarbeiterzentrierten Produktentwicklung und profitieren Sie von konkreten Handlungsempfehlungen aus der Praxis.
Jetzt gratis downloaden!