The concept of an agent has become important in both Artificial Intelligence (AI) and mainstream computer science. Our aim in this paper is to point the reader at what we perceive to be the most important theoretical and practical issues associated with the design and construction of intelligent agents. For convenience, we divide these issues into three areas (though as the reader will see, the divisions are at times somewhat arbitrary). Agent theory is concerned with the question of what an agent is, and the use of mathematical formalisms for representing and reasoning about the properties of agents. Agent architectures can be thought of as software engineering models of agents; researchers in this area are primarily concerned with the problem of designing software or hardware systems that will satisfy the prop-erties specified by agent theorists. Finally, agent languages are software systems for programming and experimenting with agents; these languages may embody principles proposed by theorists. The paper is not intended to serve as a tutorial introduction to all the issues mentioned; we hope instead simply to identify the most important issues, and point to work that elaborates on them. The article includes a short review of current and potential applications of agent technology.

Cooperative multi-agent systems are ones in which several agents attempt, through their interaction, to jointly solve tasks or to maximize utility. Due to the interactions among the agents, multi-agent problem complexity can rise rapidly with the number of agents or their behavioral sophistication. The challenge this presents to the task of programming solutions to multi-agent systems problems has spawned increasing interest in machine learning techniques to automate the search and optimization process. We provide a broad survey of the cooperative multi-agent learning literature. Previous surveys of this area have largely focused on issues common to specific subareas (for example, reinforcement learning or robotics). In this survey we attempt to draw from multi-agent learning work in a spectrum of areas, including reinforcement learning, evolutionary computation, game theory, complex systems, agent modeling, and robotics. We find that this broad view leads to a division of the work into two categories, each with its own special issues: applying a single learner to discover joint solutions to multi-agent problems (team learning), or using multiple simultaneous learners, often one per agent (concurrent learning). Additionally, we discuss direct and indirect communication in connection with learning, plus open issues in task decomposition, scalability, and adaptive dynamics. We conclude with a presentation of multi-agent learning problem domains, and a list of multi-agent learning resources. 1

ABSTRACT: “Smart ” sensors with embedded microprocessors and wireless communication links have the potential to change fundamentally the way civil infrastructure systems are moni-tored, controlled, and maintained. Indeed, a 2002 National Research Council report noted that the use of networked systems of embedded computers and sensors throughout society could well dwarf all previous milestones in the information revolution. However, a framework does not yet exist that can allow the distributed computing paradigm offered by smart sensors to be employed for structural health monitoring and control systems; current algorithms assume that all data is centrally collected and processed. Such an approach does not scale to systems with densely instru-mented arrays of sensors that will be required for the next generation of structural health moni-toring and control systems. This paper provides a brief introduction to smart sensing technology and identifies some of the opportunities and associated challenges. 1

In this article we will describe our research efforts in coping with a trade-off that can be often found in the control and optimization of todays business processes. Though centralized control may achieve nearto -optimum results in optimizing the system behavior, there are usually social, technical and security restrictions on applying centralized control. Distributed control on the other hand may cope with these restrictions but also entails sub-optimality and communicational overhead. Our concept of composable agents tries to allow a dynamic and fluent transition between globalization and localization in business process control by adapting to the current real-world system structure. We are currently evaluating this concept in the framework of patient flow control at Charit'e Berlin. Introduction Research in Distributed Artificial Intelligence (DAI, (Bond & Gasser 1988)) has been traditionally divided into Distributed Problem Solving (DPS) and Multi Agent Systems (MAS). However, r...

In this paper we describe an architecture of a multi-agent system for the assistance and supervision of the execution of medical protocols in hospital environments. A medical protocol specifies sequences of actions that could be performed in front of a particular pathology. We model medical services in hospital environments as specialized domain agents and interpret a medical protocol as a negotiation process between agents. A medical service can be involved in multiple medical protocols, and so specialized domain agents are independent of negotiation processes and autonomous system agents perform monitoring tasks. Agent interactions are performed as electronic communications through the Internet. As the Internet is not a secure communication channel, we define a robust communication interface that provides privacy, integrity and authentication during the process of exchanging information between agents.

Neural network based agents were trained on incomplete sets of classified instances that experienced doctors could suggest. Trained neural networks correctly classified the presented examples. A user does not give additional instructions to train the neural network based agents. The algorithm of selforganization is capable for define minimal number of features, as well as the number of neurons and layers in the trained neural networks. Trained neural networks were adequately represented in symbolic form that is easy to understanding. One simple form is similar to a syndrome-complex that the doctors use typically. Other is the set of logical formulas that can be easily tabulated and interpreted to be useful for diagnostic goals. The decision rules provide the evaluations of their confidence in which interested a doctor. The conducted clinical verification has shown that most decisions that symbolic rules that were discovered in databases have coincided with the doctor's conclusions.

Software Agents are the new revolution in software usability. Agents are described with anthropomorphic features such as intelligence, autonomy, and such. The school of researchers from the fields of Artificial Intelligence (AI), Distributed AI (DAI), and other more pragmatic researchers are engaged in defining these "software entities" and creating this new field of study and development along the way. In the mean time, while researchers fight for formal definitions, this new technology is bringing powerful and useful tools to the end user, in particular managing huge amount of data, such as that produced by the Internet, and taking decisions on what the user may be interested in. 1. Introduction `Intelligent Agent' and the more generic term `Agent' are terms that have been used interchangeably, since some kind of "intelligence" has been consistently described as one of the attributes of a software agent [Franklin, 96]. One of the definitions of software agents describes them as "cha...

In recent years, much research effort has been devoted in the development of systems in the context of intelligent agents. Several architectures have been proposed concerning the way agents are situated in a multi-agent system (MAS), how they act, negotiate and form teams in order to accomplish a simple or complex task. In this paper, we propose a generic MAS architecture that can be used in teaching a wide range of Medical subjects. This architecture is a result of our experience in building several autonomous educational applications. This proposed architecture summarizes the common application features and promotes modularity, scalability and reusability of educational content and instructional procedures. Keywords: Computer-Based Medical Education, Educational Software, Intelligent Agents, instructional Agents, Multi-Agent Architectures ######## ########, ##### ########## ########## #### ######### ### ######## ######### ######## #### ########## ### ###### ############# (intelligent agents). ####### ############## ##### ######## #### #### ####, # ##### ## ########### ### ##### # ### ##### ## ############# #### ######### ######### ############# (multi-agent system), ########, ################ ### ############# ##### #### #### #### ## ############ #### # ####### ####. #### ####### #### ########### ## ###### ############# #### ######### ######### ############# ### ##### ## ############# #### ############ ########## #### ###### ####### ######## ######. #### # ############# ##### ######### ### ######## ## ### ####### ### ### ######### ######## ######## ############# ########. # ########### ############# ####### ### ########## ### ############ #### ######### ## ######## ######, ### ######### ### ### ########...

Although several agent-based systems now exist on the network, most are essentially centralized on a single agent. However, the distributed, large-scale, dynamic nature of many problem spaces such as Medical Education calls for open, flexible and scalable solutions. This work proposes and implements a multi-agent based system that teaches a subject of Orthopedics. The proposed solution supports easy upgrade of the system during runtime and easy adaptation of the current system to teach other medical subjects. This paper deals with the chosen agent organization architecture and the inter-agent communication mechanism, which are the keys for the development of a modular, open, and adaptive solution. Keywords: Inscrutability, Organization of agent societies, multi-agent teams, multi-agent communication, CORBA. 1.

One of the main characteristics of a subway line is its large transport capacity (e.g., about 60000 travelers per hour in the Parisian subway) combined with a regular transport supply. The regularity is particularly important at rush time - peak hours - when an incident can provoke important delays. Experience shows that the consequences of an incident are highly dependent on the context in which the incident occurs (e.g., peak hours or not). The decisions taken by the operators are heavily relied on the incident context, and operators often make different decisions for the same incident in different contexts. The project SART (French acronym for Support system for traffic control) aims at developing an intelligent decision support system able of helping the operator in making decisions to solve an incident occurring on a line. This system relies on the notion of context. Context includes information and knowledge on the situation that do not intervene directly in the incident solving, but constrain the way in which the operator will choose a strategy at each step of the incident solving. The paper describes the SART project and highlights how Artificial Intelligence (AI) techniques can contribute to knowledge acquisition and knowledge representation associated with its context of use. Particularly we discuss the notion of context and show how we use this notion to solve a real-world problem.