1996 · 77 citations · 7 references
EngineeringComputer ScienceSystems EngineeringOrganizational DesignOrganizationsRoboticsAgent-based DesignMulti-agent SystemsMulti-robot CoordinationRobot TeamingAgent CoordinationAgent Oriented DesignDesign Methodology
The multi‑agent paradigm is widely applied to solve organizational problems, yet translating a global task specification into local agent behaviors remains a key challenge. This paper demonstrates the application of the Cassiopeia method to design and implement the organization of a soccer robot team. The authors justify the choice of this domain, describe the design process, and highlight its relevance and inherent difficulties, underscoring the need for a dedicated design methodology for distributed artificial intelligence. The Cassiopeia method is defined as a framework that articulates multi‑agent system design around the notion of organization.
The multi-agent paradigm is widely used to provide solutions to a variety of organizational problems related to the collective achievement of one or more tasks. All these problems share a common difficulty of design: how to proceed from a global specification of the collective task to the specification of the local behaviors, which are to be provided to the agents? We have defined the Cassiopeia method whose specificity is to articulate the design of a multi-agent system around the notion of organization. This paper reports the use of this method for designing and implementing the organization of a soccer robot team. We show why we chose this application and how we designed it, we discuss its interest and inherent difficulties, in order to clearly express the needs for a design methodology dedicated to DAI.
7
Hiroaki Kitano, Minoru Asada, Yasuo Kuniyoshi et al. · 1997 · 771 citations · Full text
Michael Erdmann, Tomás Lozano‐Pérez · Algorithmica · 1987 · 743 citations