Distributed Agents

Information and communication infrastructures underwent a rapid and extreme decentralization process over the past decade: From a world of statically and partially connected central servers rose an intricate web of millions of information sources loosely connecting one to another. Today, we expect to witness the extension of this revolution with the wide adoption of meta-data standards like RDF or OWL underpinning the creation of a semantic web. Again, we hope for global properties to emerge from a multiplicity of pair-wise, local interactions, resulting eventually in a self-stabilizing semantic infrastructure. This paper represents an effort to summarize the conditions under which this revolution would take place as well as an attempt to underline its main properties, limitations and possible applications.

In this paper we show how to go beyond the study of the topological properties of the Internet, by measuring its dynamical state using special active probing techniques and the methods of network tomography. We demonstrate this approach by measuring the key state parameters of Internet paths, the characteristics of queueing delay, in a part of the European Internet. In the paper we describe in detail the ETOMIC measurement platform that was used to conduct the experiments, and the applied method of queueing delay tomography. The main results of the paper are maps showing various spatial structure in the characteristics of queueing delay corresponding to the resolved part of the European Internet. These maps reveal that the average queueing delay of network segments spans more than two orders of magnitude, and that the distribution of this quantity is very well fitted by the log-normal distribution.

We present the Behaviosite paradigm, a new approach to affecting the behavior of distributed agents in a multiagent system, which is inspired by biological parasites with behavior manipulation properties. Behaviosites are special kinds of agents that "infect" a system composed of agents operating in that environment. The behaviosites facilitate behavioral changes in agents to achieve altered, potentially improved, performance of the overall system. Behaviosites need to be designed so that they are intimately familiar with the internal workings of the environment and of the agents operating within it, and behaviosites apply this knowledge for their manipulation, using various infection and manipulation strategies. To demonstrate and test this paradigm, we implemented a version of the El Farol problem, using behaviosites.

The MonALISA (Monitoring Agents in a Large Integrated Services Architecture) framework provides a set of distributed services for monitoring, control, management and global optimization for large scale distributed systems. It is based on an ensemble of autonomous, multi-threaded, agent-based subsystems which are registered as dynamic services. They can be automatically discovered and used by other services or clients. The distributed agents can collaborate and cooperate in performing a wide range of management, control and global optimization tasks using real time monitoring information.

This paper presents a reactive programming and triggering framework for the coordination of a large number of distributed agents with shared knowledge. At the heart of this framework is a highly structured shared store in the form of a constraint logic program (CLP), which is used as a knowledge base and being reacted to by agents through the use of "reactors". The biggest challenge arising from such a reactive programming framework using CLP is to develop a trigger mechanism that allows efficient "wakeup" of blocked reactors. This paper addresses the architecture of this open framework, and discusses a general methodology for doing triggering of logic conditions using views and abstractions.

We are developing a distributed system for the tracking of people and objects in complex scenes and environments using biologically based algorithms. An important component of such a system is its ability to track targets from multiple cameras at multiple viewpoints. As such, our system must be able to extract and analyze the features of targets in a manner that is sufficiently invariant of viewpoints, so that they can share information about targets, for purposes such as tracking. Since biological organisms are able to describe targets to one another from very different visual perspectives, by discovering the mechanisms by which they understand objects, it is hoped such abilities can be imparted on a system of distributed agents with many camera viewpoints. Our current methodology draws from work on saliency and center surround competition among visual components that allows for real time location of targets without the need for prior information about the targets visual features. For instance, gestalt principles of color opponencies, continuity and motion form a basis to locate targets in a logical manner. From this, targets can be located and tracked relatively reliably for short periods. Features can then be extracted from salient targets allowing for a signature to be stored which describes the basic visual features of a target. This signature can then be used to share target information with other cameras, at other viewpoints, or may be used to create the prior information needed for other types of trackers. Here we discuss such a system, which, without the need for prior target feature information, extracts salient features from a scene, binds them and uses the bound features as a set for understanding trackable objects.

Abstract. There are many computations and surveys which involve confidential data from many parties or organizations. As the concerned data is property of the organization or the party, preservation and security of this data is of prime importance for such type of computations. Although ...

This poster proposal presents operational experience of large-scale unicast network tomography, that samples a part of the European Internet. We describe in detail the ETOMIC measurement platform that was used to conduct the experiments, and its potential in future scaled-up measurements. Our main results are maps showing various spatial and temporal structure in the characteristics of the inferred queueing delay distributions. At the most loaded time of day we find that the distribution of average queueing delays among the different path segments follows closely a log-normal distribution.

Queuing delay tomography of the Internet is mostly a theoretical research topic, and measurements were mainly performed to prove the validity of a certain measurement methods. We propose a large scale Internet tomography survey to map the queueing delay in the European networks in great details, and the rest of the world to a lesser extent. The measurements will be based on the ETOMIC high accuracy packet capturing infrastructure and on DIMES vast distributed agent community. We present the rational behind the effort, the new technical tools developed to enable it, and some results from initial trials.

We propose in this paper a generic architecture to support, develop, manage, and interact with agentbased applications. In this context, we detail two main components of the architecture, namely: the AES (Agent Execution System) and the AEE (Agent Execution Environment). For both of them we describe their main concepts, data structures and interfaces. We also propose a set of characteristics that we believe will also be present in the next-generation AES, such as: independence between AES and AEE, ie, support to multi- ...

Low population areas and low-income people often face unreliable public transport characterized by long waiting times. Meanwhile, a private vehicle in the same area, on average, will remain inactive for long periods of time. Recent research in shared mobility systems addresses this situation. Specifically, Demand Responsive Transport, DRT, a transport alternative between a bus and taxi, and which has many specification variations, has pioneered provision of an alternative transport solution geared to reach under-served minorities. Solutions to DRT trip scheduling and costing mechanisms usually include constraints specific to the variation of DRT specifications. We aim to maximize use of a vehicle and we propose a design of a Community shared mobility service model that provides centralized management and ICT support in in delivering multiple services in a vehicle. Our design adds time constraints of vehicle schedule to the DRT problem. We 研究ノート

Recently, researchers have created many platforms and applications for mobile agents; however, current Agent-Oriented Software Engineering (AOSE) methodologies have yet not fully integrated the unique properties of these mobile agents. This paper attempts to bridge the gap between current AOSE methodologies and mobile agent systems by incorporating mobility into the established Multiagent Systems Engineering (MaSE) methodology. We accomplished this by adding a move command to the MaSE analysis models and then defined the required transformations to incorporate the required functionality into the design. Finally, we translated the design models into Java-based agents that operate within a mobile agent environment.

This paper describes a multiagent architecture for CAD/CAPP/CAM integration for the remote manufacturing of cylindrical parts using Internet, especially through the protocols associated with the World Wide Web. This methodology specifies the modules of the system, and how these modules interact with each other. The methodology is conceived starting from the modeling paradigm based on the synthesis of design features, in order to allow the integration of the design activities (CAD), process planning (CAPP) and manufacturing (CAM Planning and Execution). The procedure begins with the modeling of a part by features in a context of remote manufacturing using the Web as the communication means, in a client-server computer model. The system is conceived in a distributed environment of agents of interoperable pieces of software called Community of Manufacturing Agents, whose architecture is stratified in three levels: Design, Process Planning and Manufacturing. The flow of information among the three levels is accomplished through information encapsulation, being adopted the Knowledge Query and Manipulation Language (KQML) as the language pattern of messages among the agents.

In this paper, we consider several mathematical and algorithmic problems which arise naturally in the optimal deployment of modern network management systems. Specifically, we will consider the problem of minimizing the total communication costs within an architecture consisting of a distributed hierarchy of cooperating intelligent agents. We consider several communication cost models, and describe provable optimal schemes for distributing agents among machines in each of these models.

Space exploration will use humans and robots in partnership and leverage the capabilities of each where most appropriate and useful. For example, habitat modules, power systems, and human return vehicles can support human occupation on the Moon, would be landed autonomously, and assembled accurately by a group of communicating robots and sensors. A complex system of robots and sensors will complete their tasks only if supported by an effective communication network, and this white paper will discuss the properties of this communication architecture. Our main recommendations are that the communication infrastructure should consider end-to-end system-wide characteristics of sensing and actuation, constraints of power and computing resources, and interoperability. Moreover, the network platform must be evaluated early and continuously to ensure system efficiency and reliability. Such a communication network would enable flexible, sustainable, and affordable human-robotic space exploration. Human exploration of Moon's surface can be supported by a Wireless Surface Local Area Network (WSLAN) consisting of small solar-powered communication towers with simple switched patch antennas and access point transceivers, antennas, and access points on roving and fixed habitats. Each WSLAN subsystem routes voice, video, sensor streams, and data to and from space-suited humans, robots, sensors, rovers, and habitats. A WSLAN enables robots to coordinate their activities using video, sensor streams, control signals, and the exchange of software components. A companion subsystem has a microwave landmark capability whereby robots can determine precisely their position. Landmark capabilities are critical for example in a construction zone, which requires accurate landscaping and placement of structures. The WSLAN is the outpost of a space communication infrastructure that makes it possible to supervise autonomous robots from Earth-based, in-orbit, or CEV-based locations.

This paper presents the architecture of an onboard controller developed for the HERO autonomous helicopter, which is a low-cost unmanned aerial vehicle research platform. An embedded digital-signal-processor-based low-level controller is devoted to flight control, while a PC-based high-level controller is used for onboard perception tasks and interaction with other agents in a distributed system. The functional design, software architecture, and implementation of the low-level controller are analyzed in detail, focusing mainly on its runtime environment (JULIET) and its capability for flexible reconfiguration. The connectivity functions of the low-level controller with external possibly distributed agents are also addressed. Finally, the results of real autonomous flight experiments are presented, including the tracking of a smooth 3-D path described by over two hundred waypoints. Index Terms-Aerial robotics, digital signal processor (DSP), embedded system, unmanned aerial vehicle (UAV). I. INTRODUCTION U NMANNED AERIAL VEHICLES (UAVs) have become a fundamental area in the field of robotics research during the last 15 years; in this time span, they have evolved from prototypes to real working systems. Although military applications have been a fundamental drive for their success, there also exists a wide set of civil applications where they have shown their usefulness, such as forest fire detection and monitoring, disaster management, inspection of inaccessible structures, or traffic surveillance. Over the years, the focus of research has expanded from the development of hardware platforms and basic control approaches [1]-[8] to the design of fault-tolerant systems, which are able to deal with uncertain environments and react to unexpected events. There is also a trend to integrate UAVs in complex distributed systems to cooperate with other aerial or ground robots [9], [10] and wireless sensor networks,

We present the concept of mobile groups as a basic mechanism for the reliable coordination of mobile agents. Analogously to traditional group systems, mobile groups also provide message delivery guarantees and virtual synchrony. Furthermore, they make agent mobility not only visible for the group, but also consistently ordered with other group actions (such as crashes, joins, leaves, and other migrations). The mobile groups approach represents a novel mobility support mechanism, which can be used to handle reliability of mobile agents required at both, the application and system level (e.g., for coordinating distributed agents and for reliable agent migration, respectively). In this paper, we discuss the motivations for the mobile groups approach, formally define their properties, and present a membership protocol for such groups. We also discuss some implementation issues and related performance data, and present the advantages of mobile groups against mechanisms commonly employed for the coordination of mobile agents.

We introduce a novel view on how to deal with the problems of semantic interoperability in distributed systems. This view is based on the concept of emergent semantics, which sees both the representation of semantics and the discovery of the proper interpretation of symbols as the result of a self-organizing process performed by distributed agents exchanging symbols and having utilities dependent on the proper interpretation of the symbols. This is a complex systems perspective on the problem of dealing with semantics. We highlight some of the distinctive features of our vision and point out preliminary examples of its application.

Living organisms can establish an intricate network of mutually influencing relationships with their environment. In this study, a genetic algorithm is used in an attempt to model a complex adaptive system that addresses the El Farol Bar problem.
The problem was introduced by W. Brian Arthur to illustrate the dynamics of decision-making under conditions of bounded rationality. It describes a toy ecosystem with two different microhabitats and a threshold for access to resources. The agents of our computational model are the mathematical objects known as finite automata (here “flibs”, as labelled by A.K. Dewdney) that respond to binary inputs from the environment based on their internal state and the transition table encoded in their chromosomes. The number of states of the flibs, which regulates the diversity of possible chromosomes, parameterizes the degree of complexity of the system. Furthermore, by determining individual behaviours, chromosomes collectively shape the comfort level that the microhabitats currently provide. The environment retroactively exerts selection pressure on the flibs: the genetic algorithm mimics Darwinian evolution by evaluating the fitness function, random mutation and selective reproduction.
The dynamic equilibrium observed in previous El Farol bar implementations has been replicated under different threshold levels, producing more permissive or more restrictive environmental conditions.  Even populations of the simplest flibs with only one internal state can approach equilibrium. Increasing the number of states increases the efficiency of the algorithm in resource utilisation by promoting behavioural diversification through increasing genomic richness. Observation of the Gini index shows that resources are shared more equitably when the number of internal states increases: this is an emergent property that is not directly targeted by selection. Conversely, a narrower potential in genetic variability seems to lead a population to the formation of subgroups having unequal access to resources.
Further studies are needed to clarify whether the system spontaneously develops a tendency towards low or high genetic diversity.

This paper focuses on the development of a hybrid security model for Mobile Agents (MAs) system in a distributed network environment using RSA and DSA. The MA computing vulnerabilities include authentication, secure messaging, certification, trusted third parties, non-repudiation and resource control. These vulnerabilities have prevented widespread adoption of mobile agent technology. The proposed framework is able to counter these threats as agents must be protected at the same time. The model is an attempt at overcoming these issues for efficient messaging and host security. The methodology adopted is an Object-Oriented System Analysis and Design (OOSAD). The new system is implemented using JAVA programming technologies and execute on any environment with appropriate JAVA Virtual Machine.

The Cognitive Agent Architecture (Cougaar) is one of the most sophisticated distributed agent architectures developed today. As part of its research and evolution, Cougaar is being studied for application to large, logistics-based applications for the Department of Defense (DoD). Anticipiting future complex applications of Cougaar, we are investigating the Model Driven Architecture (MDA) approach to understand how effective it would be for increasing productivity in Cougarbased development efforts. Recognizing the sophistication of the Cougaar development environment and the limitations of transformation technologies for agents, we have systematically developed an approach that combines component assembly in the large and transformation in the small. This paper describes some of the key elements that went into the Cougaar Model Driven Architecture approach and the characteristics that drove the approach.

The Personal Satellite Assistant (PSA) is a softball-sized flying robot designed to operate autonomously onboard manned spacecraft in pressurized micro-gravity environments. We describe how the Brahms multi-agent modeling and simulation environment in conjunction with a KAoS agent teamwork approach can be used to support human-centered design for the PSA.

This paper proposes a distributed model predictive control strategy for a HVAC system that relies on its decomposition into subsystems based on the sensitivity analysis. An economic model predictive controller is implemented for every subsystem to optimize the operational cost of the building without compromising the thermal comfort of the occupants. Also, this work demonstrates the coordination strategy between the subsystem controllers using the sensitivity analysis approach. A discussion about the coordination strategy with the convergence property is provided. Finally, the proposed approach is illustrated using the benchmark building that considers the weather of Nancy, France.

Resumo. O ambiente SACI (Simple Agent Communication Infrastructure) (SACI, 2004) é uma ferramenta para a construção de Sistemas Multiagentes. O ambiente proporciona uma infraestrutura que facilita a implementação de agentes, permitindo que estes se conheçam e se comuniquem de forma transparente. Este trabalho teve como objetivo aperfeiçoar este ambiente através da implementação de uma nova funcionalidade. Trata-se do desenvolvimento uma biblioteca que permite ao ambiente reconhecer e monitorar os protocolos comunicação que regem os mecanismos de interação entre os agentes. Palavras-chave: Comunicação entre Agentes; protocolos de comunicação entre Agentes; SACI. 1 Introdução Um Sistema Multiagentes é uma técnica de Inteligência Artificial Distribuída em que é utilizado um conjunto de elementos, chamados agentes, que têm características como autonomia, pró-atividade, reatividade, deliberação, etc. (Wooldridge, M, 2002). A técnica de Sistemas Multiagentes tem sido utilizada nos mais variados domínios e tem se mostrada adequada para a resolução de certos problemas computacionais em que as técnicas tradicionais de computação não são aplicáveis. Desta forma, existe um grande interesse, tanto acadêmico como comercial, no aperfeiçoamento desta tecnologia. Diante disto, surgiu a necessidade de ferramentas que facilitem o desenvolvimento de tais sistemas. O ambiente SACI (Simple Agent Communication Infrastructure) (SACI, 2004) é uma destas ferramentas e fornece uma infra-estrutura para a construção de Sistemas Multiagentes. Como em um Sistema Multiagentes não existe um controle central nem um armazenamento global de informações, os agentes não têm capacidade ou informação suficiente para resolverem os problemas sozinhos. Neste contexto, faz-se necessário à existência de algum tipo de interação entre estes agentes para que o grupo como um todo consiga resolver o problema. Percebe-se então a necessidade de se definir protocolos que estabeleçam uma comunicação confiável entre estes agentes. Tais protocolos, também chamados protocolos de interação entre agentes, definem regras e estruturas que estabelecem um mecanismo para que a comunicação entre estes agentes seja objetiva e eficiente. O objetivo deste trabalho consistiu na inserção de mecanismos que possibilitam ao ambiente SACI reconhecer e implementar tais protocolos. A arquitetura interna do ambiente foi modificada acrescentando novos componentes que permitem ao SACI interpretar estes protocolos e gerenciar as interações entre os agentes. Nas próximas seções, são apresentados maiores detalhes sobre o trabalho. Na seção 2, são discutidos alguns detalhes referentes à padronização de linguagens de comunicação entre agentes.

DIMES is a distributed agent-based Internet mapping effort which at present has attracted thousands of volunteers running clients worldwide. Despite the "law of diminishing returns" seen when when adding agents in past mapping attempts, we find that significant amounts of new information can be found in the long tails of our measurement process. We use the k-pruning method to decompose the network into nodes with distinct roles. The analysis suggests a new picture of the AS-level Internet structure, which distinguishes a relatively large, redundantly connected core of nearly 100 ASes and two components that flow data in and out from this core. One component is fractally interconnected through peer links; the second makes direct connections to the core only. We plan to use this picture as a framework for measuring and extrapolating changes in the Internet's physical structure. Our analysis may also be relevant for estimating the function of nodes in the "scale free" graphs extracted from other naturally-occurring processes.

Monitoring systems in Grid infrastructures typically collect and aggregate data originating in distributed agents and store it in global, periodically refreshed repositories. However, in some scenarios access to real-time streams of monitoring information, rather than persistent data sets, would be beneficial. In this paper, we evaluate the Complex Event Processing (CEP) approach applied to real-time Grid monitoring and argue that CEP enables us to achieve two goals, otherwise difficult to combine: real-time access to monitoring data and advanced query capabilities. Monitoring for the purpose of dynamic allocation of Grid resources serves as a case study to demonstrate powerful real-time query capabilities provided by CEP. In addition, we show how to employ CEP for data reduction. For practical verification of our solution, a CEPbased Grid monitoring infrastructure, GEMINI2, has been developed. We have measured the overhead due to CEP-based monitoring and conclude that real-time Grid monitoring is possible without excessive intrusiveness for resources and network.

Monitoring systems in Grid infrastructures typically collect and aggregate data originating in distributed agents and store it in global, periodically refreshed repositories. However, in some scenarios access to real-time streams of monitoring information, rather than persistent data sets, would be beneficial. In this paper, we evaluate the Complex Event Processing (CEP) approach applied to real-time Grid monitoring and argue that CEP enables us to achieve two goals, otherwise difficult to combine: real-time access to monitoring data and advanced query capabilities. Monitoring for the purpose of dynamic allocation of Grid resources serves as a case study to demonstrate powerful real-time query capabilities provided by CEP. In addition, we show how to employ CEP for data reduction. For practical verification of our solution, a CEPbased Grid monitoring infrastructure, GEMINI2, has been developed. We have measured the overhead due to CEP-based monitoring and conclude that real-time Grid monitoring is possible without excessive intrusiveness for resources and network.

Process algebras like CSP and CCS inspired the original occam model of communication and process encapsulation. Later the π-calculus and various treatments handling mobility in CSP added support for mobility, as realised in practical programming systems such as occam-π, JCSP, CHP and Sufrin’s CSO, which allow a rather abstract notion of motion of processes and channel ends between parents or owners. Milner’s Space and Motion of Communicating Agents on the other hand describes the bigraph framework, which makes location more of a first-class citizen of the calculus and evolves through reaction rules which rewrite both place and link graphs of matching sections of a system state, allowing more dramatic dynamic reconfigurations of a system than simple process spawning or migration. I consider the tractability of the notation, and to what extent the additional flexibility reflects or elicits desirable programming paradigms.

In this paper, we consider several mathematical and algorithmic problems which arise naturally in the optimal deployment of modern network management systems. Specifically, we will consider the problem of minimizing the total communication costs within an architecture consisting of a distributed hierarchy of cooperating intelligent agents. We consider several communication cost models, and describe provable optimal schemes for distributing agents among machines in each of these models.

We study the development of distributed agent environments as distributed event-based systems specified in the Ambient Event Calculus (AEC). The AEC is a logic-based formalism that is developed here to support the representation of a distributed agent environment as a persistent composite structure evolving over time. Such a complex structure supports the interaction between agents, objects, and containers, entities that have their own external observable state and can be distributed over a network. Interactions between these entities are specified in terms of events that represent actions executed by agents on objects and other agents in the environment. When events happen they are stored in containers and are notified to agent sensors that subscribe to event descriptions and as a result perceive the interactions. The AEC formalism also allows changes caused by events to be delivered across distributed containers, according to the topology of the application environment. We illustrate the use of AEC and we show how to specify interactions within the GOLEM agent platform applied to a specific agent scenario.

Recent psychological theories of emotion have explicitly focused on modeling the multiplicity of levels of the human emotion system. This increased interest in creating multi-level theories of emotion matches similar efforts in the area of cognitive architectures. In this paper we report our initial work toward integrating a psychological multi-level model of emotions with Brahms, a multi-agent system that is used to model and simulate work practice.

Vehicle routing problems (VRP's) involve assigning a fleet of limited capacity service vehicles to service a set of customers. This paper describes an innovative, agent-based approach to solving a real-world vehicle-routing problem embedded in a highly dynamic, unpredictable domain. Most VRP research, and all commercial products for solving VRP's, make a static-world assumption, ignoring the dynamism in the real world. Our system is explicitly designed to address dynamism, and employs an in-time algorithm that quickly finds partial solutions to a problem, and improves these as time allows. Our fundamental innovation is a stochastic improvement mechanism that enables a distributed, agent-based system to achieve highquality solutions in the absence of a centralized dispatcher. This solution-improvement technology overcomes inherent weaknesses in the distributed problem-solving approach that make it difficult to find high-quality solutions to complex optimization problems. In previous work on similar problems, the MARS system of Fischer and Müller, et al., achieved an average route performance of roughly 124% of Solomon's algorithm for a VRP problem, which is known to achieve results that average roughly 107% of optimal. Our algorithm produces routes that average 106% those produced by an adaptation of Solomon's algorithm to a more general problem. The Pickup and Delivery Problem with Time Windows Vehicle routing problems (VRP's) involve assigning a fleet of limited capacity service vehicles to service a set of customers. These problems have been well studied in the Operations Research literature (e.g.,Laport 1992.) Insofar as the Travelling Salesman Problem is embedded in virtually every VRP of practical interest, this entire class of problems is intractable. The vast majority of vehicle routing research addresses a static problem in which all of the relevant data is known in advance. Typically, service vehicles are assumed to start at some initial location (the depot), all customer demands and constraints are known, as are the distance and travel times between each pair of customers, and between each customer and the depot. There are a number of commercial systems for solving

It is essential for robots working in close proximity to people to be both safe and trustworthy. We present a case study on formal verification for a high-level planner/scheduler for the Care-O-bot, an autonomous personal robotic assistant. We describe how a model of the Care-O-bot and its environment was developed using Brahms, a multiagent workflow language. Formal verification was then carried out by automatically translating this model to the input language of an existing model checker. Four sample properties based on system requirements were verified. We then refined the environment model three times to increase its accuracy and the persuasiveness of the formal verification results. The first refinement uses a user activity log based on real-life experiments, but is deterministic. The second refinement uses the activities from the user activity log nondeterministically. The third refinement uses "conjoined activities" based on an observation that many user activities can overlap. The four samples properties were verified for each refinement of the environment model. Finally, we discuss the approach of environment model refinement with respect to this case study.

Human-robot teams are likely to be used in a variety of situations wherever humans require the assistance of robotic systems. Obvious examples include healthcare and manufacturing, in which people need the assistance of machines to perform key tasks. It is essential for robots working in close proximity to people to be both safe and trustworthy. In this paper we examine formal verification of a high-level planner/scheduler for autonomous personal robotic assistants such as Care-O-bot R. We describe how a model of Care-O-bot and its environment was developed using Brahms, a multiagent workflow language. Formal verification was then carried out by translating this to the input language of an existing model checker. Finally we present some formal verification results and describe how these could be complemented by simulation-based testing and realworld end-user validation in order to increase the practical and perceived safety and trustworthiness of robotic assistants.

Distributed Services Architecture with support for mobile agents between services, offers significantly improved communication and computational flexibility. The uses of agents allow execution of complex operations that involve large amounts of data to be processed effectively using distributed resources. The prototype system Distributed Agents for Mobile and Dynamic Services (DIAMOnDS), allows a service to send agents on its behalf, to other services, to perform data manipulation and processing. Agents have been implemented as mobile services that are discovered using the Jini Lookup mechanism and used by other services for task management and communication. Agents provide proxies for interaction with other services as well as specific GUI to monitor and control the agent activity. Thus agents acting on behalf of one service cooperate with other services to carry out a job, providing inter-operation of loosely coupled services in a semi-autonomous way. Remote file system access functionality has been incorporated by the agent framework and allows services to dynamically share and browse the file system resources of hosts, running the services. Generic database access functionality has been implemented in the mobile agent framework that allows performing complex data mining and processing operations efficiently in distributed system. A basic data searching agent is also implemented that performs a query based search in a file system. The testing of the framework was carried out on WAN by moving Connectivity Test agents between AgentStations in CERN, Switzerland and NUST, Pakistan.

In this paper, we consider several mathematical and algorithmic problems which arise naturally in the optimal deployment of modern network management systems. Specifically, we will consider the problem of minimizing the total communication costs within an architecture consisting of a distributed hierarchy of cooperating intelligent agents. We consider several communication cost models, and describe provable optimal schemes for distributing agents among machines in each of these models.

In this paper we investigate measures over bounded lattices using algebraic geometry techniques. We prove that measures over a bounded lattice can be seen as a classical measure over a measurable space. Then, we define the measurability of a lattice and describe measures over finite lattices.

The promise of agent-based systems is leading towards the development of autonomous, heterogeneous agents, designed by a variety of research/industrial groups and distributed over a variety of platforms and environments. Teamwork among these heterogeneous agents is critical in realizing the full potential of these systems and scaling up to the demands of large-scale applications. Indeed, to succeed in highly uncertain, complex applications, the agent teams must be both robust and exible. Unfortunately, development of such agent teams is currently extremely di cult. This paper focuses on signi cantly accelerating the process of building such teams using a simpli ed, abstract framework called team-oriented programming (TOP). In TOP, a programmer speci es an agent organization hierarchy and the team tasks for the organization to perform, but abstracts away from the large number of coordination plans potentially necessary to ensure robust and exible team operation. We support TOP through a distributed, domain-independent teamwork layer that integrates core teamwork coordination and communication capabilities. We have recently used TOP to integrate a diverse team of heterogeneous distributed agents in performing a complex task. We outline the current state of our TOP implementation and the outstanding issues in developing such a framework.

Here we describe mechanisms for a half-dozen or so different types of learning to be implemented in “conscious” software agents, and speculate briefly on their implications for human learning and development. In particular, we're concerned with conceptual learning and behavioral learning in two “conscious” software agents, namely, CMattie and IDA. We offer computational mechanisms for such learning.

selectivity. The restriction of selectivity is mainly compelled by the specific characteristics of large-scale multi-agent systems. In such settings, it is reasonable to assume that the overhearing resources will be essentially limited, thus allowing the overhearing agent to overhear only a subset of inter-agent communications. Most previous investigations on overhearing ignore the limitation of selectivity, assuming that all relevant interagent communications can be overheard. In contrast, our work provides an empirical study of selective overhearing committed by both centralized and distributed teams of collaborative overhearing agents.

The role of humans in aviation and other domains continues to shift from manual control to automation monitoring. Studies have found that humans are often poorly suited for monitoring roles, and workload can easily spike in offnominal situations. Current workload measurement tools, like NASA TLX, use human operators to assess their own workload after using a prototype system. Such measures are used late in the design process and can result in expensive alterations when problems are discovered. Our goal in this work is to provide a quantitative workload measure for use early in the design process. We leverage research in human cognition to define metrics that can measure workload on belief-desire-intentions based multi-agent systems. These measures can alert designers to potential workload issues early in design. We demonstrate the utility of our approach by characterizing quantitative differences in the workload for a single pilot operations model compared to a traditional two pilot model.

Information systems must establish trust to cooperate effectively in open environments. We are developing an agent-based approach for establishing trust, where information systems are modeled as agents that provide and consume services. Agents can help each other find trustworthy parties by providing referrals to those that they trust. We propose a graph-based representation of services for modeling the trustworthiness of agents. This representation captures natural relationships among service domains and provides a simple means to accommodate the accrual of trust placed in a given party. When interpreted as a lattice, it enables less important services (e.g., low-value transactions) to be used as gates to more important services (e.g., high-value transactions). We first show that, where applicable, this approach yields superior efficiency (needs fewer messages) and effectiveness (finds more providers) than a vector representation that does not capture the relationships between services. Next, we study trade-offs between various factors that affect the performance of this approach.

The present work describes the design of a distributed agent system devoted to the Dial-a-Ride Problem. This routing and scheduling problem consists in finding a set of routes and schedules for each vehicle that satisfies a set of trip requests comming from users. The agent system distributes an improved insertion heuristic for the scheduling of passengers' trip requests over a fleet of vehicles. Agents make use of the contract-net protocol as base coordination mechanism for the planning and scheduling of passenger trips.

In this paper, we consider several mathematical and algorithmic problems which arise naturally in the optimal deployment of modern network management systems. Specifically, we will consider the problem of minimizing the total communication costs within an architecture consisting of a distributed hierarchy of cooperating intelligent agents. We consider several communication cost models, and describe provable optimal schemes for distributing agents among machines in each of these models.