Cosmin Carabelea

Coordinating Agents in Organizations Using Social Commitments

One of the main challenges faced by the multi-agent community is to ensure the coordination of autonomous agents in open heterogeneous multi-agent systems. In order to coordinate their behaviour, the agents should be able to interact with each other. Social commitments have been used in recent years as an answer to the challenges of enabling heterogeneous agents to communicate and interact successfully. However, coordinating agents only by means of interaction models is difficult in open multi-agent systems, where possibly malevolent agents can enter at any time and violate the interaction rules. Agent organizations, institutions and normative systems have been used to control the way agents interact and behave. In this paper we try to bring together the two models of coordinating agents: commitment-based interaction and organizations. To this aim we describe how one can use social commitments to represent the expected behaviour of an agent playing a role in an organization. We thus make a first step towards a unified model of coordination in multi-agent systems: a definition of the expected behaviour of an agent using social commitments in both organizational and non-organizational contexts.

Using social power to enable agents to reason about being part of a group

One of the main challenges in multi-agent systems is the coordination of autonomous agents. In order to achieve this coordination, the agents are considered to be part of what we call a group (e.g., organization, institution, team, normative society, etc.). Our goal is to enable an agent to reason about the implications of being part of a group: what does it gain or lose, what are the constraints imposed on its behaviour. The theory of social power has been proposed as a paradigm to describe the agents behaviour. In this paper we use this theory, we formalize it and we extend it to include group-related aspects. We then show how, using this theory, an agent is able to reason about the constraints imposed on its behaviour by the group, for example to decide whether it should enter or not a group.

Benefits and Requirements of Using Multi-agent Systems on Smart Devices

Due to the emergence of Internet and embedded computing, humans will be more and more faced to an increasing intrusion of computing in their day-to-day life by what it is called now smart devices. Agent characteristics like pro-activeness, autonomy and sociability and the inherent distributed nature of multi-agent systems make them a promising tool to use in the smart devices applications. This paper tries to illustrate the benefits of using multi-agent systems on smart communicating objects and discusses the need of using multi-agent platforms. We then present an overview of multi-agent platforms created for use on small devices (i.e. devices with limited computing power) and for each reviewed multi-agent platform we try to identify what are its main characteristics and architectural concepts.

Genetic prediction of a multi-agent environment evolution

This paper deals with application of genetic algorithms to action planning and environment evolution prediction in a multi-agent system (MAS). The considered system is represented by a community of competitive agents exploiting natural renewable resources. From the large number of such systems that can be identified in real life, we have considered the case of fishing companies exploiting a group of resources represented by fish banks. In order to validate the MAS predictions, a company action plans simulation model was developed. The simulation results can be further used by the MAS for action re-planning.

Engineering a protocol server using strategy-agents

This paper addresses the idea of using a server of negotiation protocols to facilitate negotiations in an open multi-agent system. The agents that do not know the negotiation protocols of the system can query the server for a description of the protocol and use it in negotiations. After pointing out some of the major problems encountered by the existing approaches, we propose a special type of agent (called strategy-agent) that knows how to negotiate using a fixed protocol and negotiation strategy. Such an agent will then negotiate for the benefit of other agents that do not know the required protocol.

Genetic Models for the Rational Exploitation of Resources

The paper presents a multi-agent system aimed to solve the problem of rational exploitation of natural renewable resources by self-interested agents, a problem known also as the Tragedy of Commons. The system is geared towards a particular instance of this problem, namely the FishBanks game. The agents use genetic algorithms to build plans for fishing in several fishing banks and to buy or sell ships, in order to maximize their profit. We investigate the use of several genetic models based on cooperative co-adapted species to model the multi-agent world from the point of view of a particular agent and to predict society and environment evolution. The system is evaluated under different assumptions regarding the agent profiles and environment features, and the corresponding experimental results are discussed. An auction model for buying and selling ships is also proposed.