Linking Spatially Explicit Parallel Continuous and
Boleslaw K. Szymanski and Gilbert Chen (Rensselaer Polytechnic Institute)
This paper advocates the use of mobile agents for linking simulations running on different computers. A Mobile Component approach is proposed to enhance reusability of existing simulations and to improve efficiency of component based simulations of complex systems. A basic unit of the mobile component simulation is a simulation server with a communication interface to mobile agents. Each mobile agent links and coordinates component's execution. We used this approach to implement a combined Lyme disease simulation. It consists of a partial differential equation based continuous simulation and parallel discrete event simulation with explicit space representation. The performance of this implementation is presented to demonstrate the feasibility of the Mobile Component approach. In addition, a process-port model of simulation is discussed. Its implementation allows efficient linkage of simulation servers, if they are programmed in a simulation language supporting the process-port model. We finally show that the performance of the Mobile Component approach could be significantly improved by using compiler techniques to eliminate overhead of communication among simulation servers.
An Agent-Based Framework for Linking Distributed
Linda F. Wilson, Daniel Burroughs, Jeanne Sucharitaves, and Anush Kumar (Dartmouth College)
Simulations often operate on static datasets and data sources, but many simulations would benefit from access to dynamic data. This paper describes our work developing a software agent-based framework for dynamically linking distributed simulations and other remote data resources. The framework allows independently-designed simulations to communicate seemlessly with no a priori knowledge of the details of other simulations and data sources. In this paper, we discuss our architecture and current implementation developed using the D'Agents mobile agent system. To demonstrate the feasibility of our system, we present a prototype for a hypothetical search and rescue mission.
'Plug and Test' - Software Agents in Virtual
Adelinde M. Uhrmacher (University Rostock) and Bernd G. Kullick (University Ulm)
James - a Java Based agent modeling environment for simulation has been developed to support the compositional construction of test beds for multi-agent systems and their execution in distributed environments. The modeling formalism of James imposes only few constraints on the modeling of agents and facilitates a ``plug and test'' with pieces of agent code which has been demonstrated in earlier work. However, even entire agents can be run in James as they are run in their run-time environment. The integration of agents as a whole is based on model templates which serve as the agents' interface and representative during the simulation run. The effort which is put into defining model templates for selected agent systems obviates the need for the single agent programmer to get acquainted with the underlying modeling and simulation formalism. Instead the agent programmer can compose the experimental frame and test the programmed agents as they are. The approach is illustrated with agents of the mobile agent system Mole.
Domain-General Simulation and Planning with Physical
Marc S. Atkin, David L. Westbrook, and Paul R. Cohen (University of Massachusetts at Amherst)
Physical schemas are representations of simple physically grounded relationships and interactions such as ``move,'' ``push,'' and ``contain.'' We believe they are the conceptual primitives an agent employs to understand its environment. Physical schemas can be used at varying levels of abstraction across a variety of domains. We have designed a domain-general agent simulation and control testbed based on physical schemas. If a domain can be described in physical terms as agents moving and applying force, it can be simulated in this testbed. Furthermore, we show that physical schemas can be viewed as the basis for abstract plans and a domain-general planner, GRASP. Our simulation and planning system is currently being evaluated in a continuous, dynamic, and adversarial domain based on the game of Capture the Flag. The paper concludes with an example of how GRASP was applied to the problem of Course of Action generation and evaluation.
Using Agent-Based Modeling to Capture Airpower
Richard K. Bullock (Air Force Studies and Analyses Agency) and Gregory A. McIntyre and Raymond R. Hill (Air Force Institute of Technology)
Airpower's strength lies in quickly striking the enemy directly where they are vulnerable while being unhampered by geography and surface forces. Airpower theory suggests the effects of these strikes propagate throughout an opponent's military system yielding catastrophic output or strategic effects. Despite this theory being a cornerstone of US Air Force doctrine, current Air Force models do not seem to capture airpower's inherent strength. Since these models are used to support budgetary decision making, the US may not be funding the airpower capability it needs. This effort focuses on developing an approach to capture strategic effects in models. The approach establishes a basis for the effects in military theory as well as the field of Complex Adaptive Systems. Using these concepts as a foundation, a simulation model called the Hierarchical Interactive Theater Model (HITM) is constructed and exercised. HITM output depicts a cascading deterioration in force effectiveness and eventual total collapse resulting from destruction of vital targets. This outcome is consistent with the expected results of strikes against centers of gravity defined in Air Force doctrine suggesting agent-based modeling is an effective way to simulate strategic effects at the operational level of war.
A Simulation Test-Bed to Evaluate Multi-Agent Control
of Manufacturing Systems
Robert W. Brennan and William O (University of Calgary)
Current research in the area of manufacturing planning and control has moved away from traditional centralized solutions towards distributed architectures that range from hierarchical to heterarchical. Between these two extremes of the control architecture spectrum lies the holonic manufacturing systems paradigm, where partial dynamic hierarchies of agents cooperate to meet global system objectives in the face of disturbances. This paper describes a simulation test bed for the evaluation of a distributed multi-agent control architecture for holonic manufacturing systems that integrates discrete-event simulation software into its design to allow the control architecture to be evaluated with a variety of emulated manufacturing systems.
Agent-Directed Simulation - Challenges to Meet Defense
and Civilian Requirements
Tuncer I. Íren (Information Technologies Research Institute), S. K. Numrich (U.S. Naval Research Laboratory), Adelinde M. Uhrmacher (University of Rostock), Linda F. Wilson (Dartmouth College) and Erol Gelenbe (University of Central Florida)
The aim of this panel session is to point out the importance of agent-directed simulation, as a scientific concept and technological possibility, to enhance the potential of simulation in both civilian and defense applications. The members of the panel (organized by Dr. Íren) are: Dr. Erol Gelenbe, Dr. S. K. Numrich, Dr. Adelinde Uhrmacher, and Dr. Linda Wilson. The position statements of the panel members are given separately. Íren bases his arguments on the NATO Modelling and Simulation Master Plan. He points out the need to proactively advance simulation science and technology to satisfy the requirements of the sophisticated defense applications. He stresses that, among other methodological advance possibilities, the three categories of agent-directed simulation have to be properly developed and/or tailored for defense applications. Gelenbe's interests include goal-directed knowledge processing abilities of agents in hostile environments. Numrich stresses on the need for command and search agents in defense applications. Uhrmacher states challenges for the users and the simulationists on the need of agents for modelling and agents for testing. Wilson covers four key challenges to agent-directed simulation that are: security, standards in communication, computer resources, and system management and monitoring.
Multi-Agent Enabled Modeling and Simulation Towards
Collaborative Inventory Management in Supply Chain
Yonghui Fu and Rajesh Piplani (Nanyang Technological University), Robert de Souza (Viewlocity) and Jingru Wu (National University of Singapore)
This paper is framed to address the preliminary approaches towards process-oriented collaborative inventory management in supply chain taking advantage of multi-agent technology in terms of modeling and simulation. Initially, a SCM Support Model is proposed as a foundation to combine the supply chain processes with the multi-agent system. In succession, a simple PC assembling case is investigated and simulated mainly to validate the SCM Support Model. As a result, the combination has the potential to make possible a real strategic competitive advantage for the entire supply chain and will enable new forms of business, namely, collaborative inventory management. Accordingly, a theoretical framework of collaborative inventory management is highlighted to refine and extend the SCM Support Model with the purpose to synchronize decisions as well as actions.
Agent-Based Simulation of Dynamic Online
Hideyuki Mizuta (IBM Research) and Ken Steiglitz (Princeton University)
The need to understand dynamic behavior in auctions is increasing with the popularization of online auctions. Applications include designing auction mechanisms, bidding strategies, and server systems. We describe simulations of a typical online auction, where the duration is fixed, and the second-highest price is continuously posted and determines the winner's payment. We modeled agents of exactly two types, idealizations and simplifications of those observed in practice: early bidders, who can bid any time during the auction period, and snipers, who wait till the last moments to bid. This allows us to study the interactions of the two types of bidders during the course of auctions, and the effects of the two strategies on the probability of winning, the final price, and the formation of price consensus in iterated auctions. Results show that 1) early bidders can win with a lower price on average than snipers, but much less often; 2) the late bidding strategy of snipers is effective; and 3) in iterated auctions, adjustment feedback of motivational parameters can lead to effective price consensus with small fluctuations.
Use of Discrete Event Simulation to Validate an Agent
Based Scheduling Engine
Shubhabrata Biswas and Sara Merchawi (DELMIA Corporation)
This paper discusses the use of simulation in a new context. Most often QUEST is viewed as a standalone simulation tool to analyze and understand shop floor behavior. It has rarely been used in conjunction with other proprietary software. This paper attempts to demonstrate how QUEST is being used in conjunction with an Agent Based Scheduling Software (henceforth referred to as Scheduler) developed using JAVA. The Scheduler is a multi-threaded application using software agents. It can run in a distributed manner over a LAN. This paper deals with the Agent Based Architecture of the Scheduler as well as the important role of Simulation in validating the Scheduler.