No full-text available
To read the full-text of this research,
you can request a copy directly from the author.
... Since, all of the agents are on the same team and have the same utility function, R4 interprets this proposal as an indication that R1 is unaware of R4's other commitments: it therefore shares that information with R1 and makes a counter-proposal. R1 is not sure whether R3 can help (in fact, given the shared preferences, it is unsure whether R4 knows whether it can help) this explains exchange (5). Robot R4 interprets this as a request for information and therefore shares the information in (6) (we can assume that, for example, R3's antenna just went d o wn and R1 cannot communicate with it). ...
... 3 Agents are assumed to have access to a common library of recipes 6] that describe how tasks are decomposed. In our implementation, recipes are represented using the Procedural Reasoning System (PRS) 4,5]. Figure 3 is an example of a PRS recipe for searching an area using a particular sensor. ...
... If agent i proposes some action which represents a division of some resource or task with agent j, then the proposal should either: (1) con rm the typical distribution of the task or resource, or (2) give reasons for a departure from the typical distribution. 5 For example, if i proposes to divide the task of patrolling a particular area in a less than even way, i should communicate a reason for doing so: for example, because of other commitments. The prevention clause corresponds to the reason for the agent's proposal. ...
Structured negotiation is proposed as a new method through which collaborating agents can seek consensus on the apportionment of tasks and resources. The approach draws on research in collaborative planning and human dialog understanding: agent interactions are organized in a manner that reflects the structure of a shared plan. Negotiations are incremental and interleaved with the shared planning process while communications supporting negotiations are made efficient by drawing on knowledge of a prevailing context. Agent proposals to team members are annotated with causal information that compactly expresses relationships between new proposals and the current context. Normative guidelines for proposal generation further restrict communications of ancillary information to only those fragments that represent departures from the norm. Finally, a set of interpretation rules allows agents to infer information not explicitly communicated.
... 22: The goal stacks at the time of P1.47 Steps P1.48{P1.75 and Steps P2.43{P2.75 ...
... 23: The goal stacks at the time of P1.75 ...
... Indeed, considerable research and effort has gone into developing multiagent planning systems to allow agents to anticipate and avoid unintended negative interactions, or conflicts [2] [9] [10] [13]. ...
... Others have looked not at the single-agent plan merging problem but at ways of coordinating the plans of multiple agents in a multiagent domain [6] [11] [17]. Georgeff's coordination mechanism for agents acting in a shared environment imposes synchronization constraints between agents' actions, to guarantee that their combined execution does not cause conflicts [10]. An key element of Georgeff's strategy was to identify, through exhaustive pairwise comparisons , the subset of actions that could conflict, and grouping these together into critical regions to reduce the combinatorics of synchronization. ...
It is critical for agents in a multiagent environment to avoid interfering with each other when carrying out their tasks. However, to avoid execution inefficiencies, they also should capitalize on cooperative opportunities. In state oriented domains [14], identifying overlapping effects between agents' plans enables some agents to leave some tasks to others, thereby reducing the cost of execution and improving the overall efficiency of the multiagent system. This is what we term synergy. In this paper, we define criteria for finding a certain type of synergy involving agents with overlapping goals. We also develop algorithms for discovering this synergy between planning agents that exploit hierarchical plan representations. Our results show that our approach not only can reduce the costs of finding synergies compared to non-hierarchical strategies, but can also find synergies that might otherwise be missed.
... 22: The goal stacks at the time of P1.47 Steps P1.48{P1.75 and Steps P2.43{P2.75 ...
... 23: The goal stacks at the time of P1.75 ...
Typescript. Thesis (Ph. D.)--Duke University, 1994. Vita. Includes bibliographical references (leaves 355-380).
... These e orts in DAI and others that have followed dealt with negotiations in the case of cooperative systems which are designed to achieve a common general task, or in which the agents belong to the same organization or unit (see for example George 1983] which describes a method for synthesizing multi-agent plans from simple single-agent plans, Sathi et al. 1986] which deals with project management, Durfee 1988; Durfee and Lesser 1989] which deals with the vehicle monitoring domain, and Sathi and Fox 1989] which deals with resource reallocations). Con icts among the agents in these environments may arise while each tries to achieve its own sub-tasks (for example, they may need to share the same resources), but their overall task is the same. ...
... Research in Distributed Arti cial Intelligence (DAI) is concerned with how automated agents can be designed to interact e ectively (for a survey of DAI see Bond and Gasser 1988;Gasser 1991]). The subject of negotiation, that we have discussed in this paper, is one of the subjects that has occupied the e orts of the DAI community (e.g., Smith and Davis 1983;George 1983;Sycara 1987;Malone et al. 1988; Conry et al. 1991;Durfee 1988;Rosenschein and Genesereth 1985;Sathi and Fox 1989;Zlotkin and Rosenschein 1990;Kreifelts and Martial 1990;Durfee and Montgomery 1990]. ...
Negotiations are very important in a multi-agent environment, particularly, in an environment where there are conflicts between the agents, and cooperation would be beneficial. We have developed a general structure for a Negotiating Automated Agent that consists of five modules: a Prime Minister, a Ministry of Defense, a Foreign Office, a Headquarters and Intelligence. These modules are implemented using a dynamic set of local-agents belonging to the different modules. We used this structure to develop a Diplomacy player, Diplomat. Playing Diplomacy involves a certain amount of technical skills as in other board games, but the capacity to negotiate, explain, convince, promise, keep promises or break them, is an essential ingredient in good play. Diplomat was evaluated and consistently played better than human players. Key Words: Automated Negotiations, Multi-Agent Environment, Game Playing, Localagents, Diplomacy. Subject Category: Cognitive Science, Knowledge Representati...
... Another important issue that distinguishes between various DAI research e orts is whether the goals themselves need to be adjusted, that is, whether there may be any fundamental con icts among di erent agents' goals. Thus, for example, George 's early work on multiagent planning assumed that there was no basic con ict among agent goals, and that coordination was all that was necessary to guarantee success (George , 1983(George , , 1984Stuart, 1985). Similarly, planning in the context of Lesser, Corkill, Durfee, and Decker's research (Decker & Lesser, 1992 often involves coordination of activities (e.g., sensor network computations) among agents who have no inherent con ict with one another (though surface con ict may exist). ...
... The synthesis, synchronization, or adjustment process for multiple agent plans thus constitute some of the (varied) foci of DAI planning research. Synchronization through con ict avoidance (George , 1983(George , , 1984Stuart, 1985), distribution of a single-agent planner among multiple agents (Corkill, 1979), the use of a centralized multiagent planner (Rosenschein, 1982), and the use of consensus mechanisms for aggregating subplans produced by multiple agents (Ephrati & Rosenschein, 1993b), have all been explored, as well as related issues (Cohen & Perrault, 1979;Morgenstern, 1987;von Martial, 1992avon Martial, , 1992bKreifelts & von Martial, 1991;Kamel & Syed, 1989;Grosz & Sidner, 1990;Kinny, Ljungberg, Rao, Sonenberg, Tidhar, & Werner, 1992;Ferber & Drogoul, 1992;Kosoresow, 1993). ...
This paper lays part of the groundwork for a domain theory of negotiation, that is, a way of classifying interactions so that it is clear, given a domain, which negotiation mechanisms and strategies are appropriate. We define State Oriented Domains, a general category of interaction. Necessary and sufficient conditions for cooperation are outlined. We use the notion of worth in an altered definition of utility, thus enabling agreements in a wider class of joint-goal reachable situations. An approach is offered for conflict resolution, and it is shown that even in a conflict situation, partial cooperative steps can be taken by interacting agents (that is, agents in fundamental conflict might still agree to cooperate up to a certain point). A Unified Negotiation Protocol (UNP) is developed that can be used in all types of encounters. It is shown that in certain borderline cooperative situations, a partial cooperative agreement (i.e., one that does not achieve all agents' goals) might be preferred by all agents, even though there exists a rational agreement that would achieve all their goals. Finally, we analyze cases where agents have incomplete information on the goals and worth of other agents. First we consider the case where agents' goals are private information, and we analyze what goal declaration strategies the agents might adopt to increase their utility. Then, we consider the situation where the agents' goals (and therefore stand-alone costs) are common knowledge, but the worth they attach to their goals is private information. We introduce two mechanisms, one 'strict', the other 'tolerant', and analyze their affects on the stability and efficiency of negotiation outcomes.
... This paper introduces a new concept of local and autonomous operational planning of electric vehicles as the means to meet system-wide objectives of Smart Grids. The motivation here is that if adjustments in power demand can be pre-computed and scheduled, operational uncertainties are minimized and more effective regulatory actions can be applied under several operational scenarios, e.g., failures of power generators, price peaks, weather events influencing the availability of renewable energy resources, etc. Planning is a well-established approach in literature [36][37][38][39] and in several related real-world application domains [11,12,40,41]. ...
The penetration of electric vehicles becomes a catalyst for the sustainability of Smart Cities. However, unregulated battery charging remains a challenge causing high energy costs, power peaks or even blackouts. This paper studies this challenge from a socio-technical perspective: social dynamics such as the participation in demand-response programs, the discomfort experienced by alternative suggested vehicle usage times and even the fairness in terms of how equally discomfort is experienced among the population are highly intertwined with Smart Grid reliability. To address challenges of such a socio-technical nature, this paper introduces a fully decentralized and participatory learning mechanism for privacy-preserving coordinated charging control of electric vehicles that regulates three Smart Grid socio-technical aspects: (i) reliability, (ii) discomfort and (iii) fairness. In contrast to related work, a novel autonomous software agent exclusively uses local knowledge to generate energy demand plans for its vehicle that encode different battery charging regimes. Agents interact to learn and make collective decisions of which plan to execute so that power peaks and energy cost are reduced system-wide. Evaluation with real-world data confirms the improvement of drivers’ comfort and fairness using the proposed planning method, while this improvement is assessed in terms of reliability and cost reduction under a varying number of participating vehicles. These findings have a significant relevance and impact for power utilities and system operator on designing more reliable and socially responsible Smart Grids with high penetration of electric vehicles.
... Early work tended to focus on either communication (Georgeff (1983); Grosz and Kraus (1996)) or on the plan coordination problem (Durfee and Lesser, 1991), the latter involving methods for merging existing plans into global plans and dealing with insincere agents Rosenschein, 1993b, 1994). ...
Classical planning problems consist of an environment in a predefined state; a set of
deterministic actions that, under certain conditions, change the state of the environment;
and a set of goal conditions. A solution to a classical planning problem is a sequence of
actions that leads from the initial state to a state satisfying the problem’s goal conditions.
There are many methods for finding solutions to classical planning problems, and a
popular technique is to exploit structures that commonly occur. One such structure,
apparent in many planning domains, is a breakdown of the problem into multiple agents.
However, methods for finding and exploiting multiagent structures are not prevalent in
the literature and are currently not competitive.
This thesis sets out to rectify this problem. Its first main contribution, is to introduce
a domain independent algorithm for extracting multiagent structure from classical
planning problems. The algorithm relies on identifying a generalisable property of
agents in planning; namely, that agents are entities with an internal state, a part of the
planning problem that, under a certain distribution of actions, only they can modify.
Once this is appropriately formalised, the decomposition algorithm is introduced and
is shown to produce identifiably multiagent decompositions over all of the classical
planning domains used in the International Planning Competitions, even finding more
detailed decompositions than are used by humans in certain cases.
Solving multiagent planning problems can be challenging because a solution may
require complex inter-agent coordination. The second main contribution of the thesis
is a heuristic planning algorithm that effectively exploits the structure of decomposed
domains. The algorithm transforms the coordination problem into a process of subgoal
generation that can be solved efficiently under a well-known relaxation in planning. The
generated subgoals guide the search so that it is always performed by one single-agent
subproblem at a time. The algorithm is evaluated and shown to greatly outperform
current state-of-the-art planners over decomposable domains.
The thesis also includes discussion of the possible extensions of this work, to include
the multiagent concepts of self-interested agents and concurrent actions. Results from
the multiagent planning literature are improved upon and a new solution concept is
presented that accounts for the ‘farsightedness’ inherent in planning. A method is
then presented that can find stable solutions for a certain class of multiagent planning
problems. A new method is introduced for modelling concurrent actions that allows
them to be written without requiring knowledge of each other agent in the domain, and
it is shown how such problems can be solved by a translation to single-agent planning.
... If we are going back to our study case, there will be a dynamic related to the Energy, one to Finance and the last one to Carbon Emission Control. In the context of a spatial [6], [7] and highly communicating [8] MAS, we also can define two more dynamics for its position and its communication. ...
The multi-agent systems are successfully used in modeling of dynamic complex systems, and simulations of these models reinforce the knowledge of experts and even allow them to explore new horizons or to cross boundaries. This is the reason why the models being tackled are increasingly varied, and as one goes along with experimentations, these models are completed, intercrossed. Consequently they become increasingly complex. In our previous work [1], we proposed a first modeling approach to support this complexity increase: the Dynamic- Oriented Modeling (DOM). The application of this approach can effectively support the increase of the model. This increase applies to both agents and environments. This DOM approach tackles the problem of the latter by splitting in multiple parts. But if DOM led to organize properly the multiple environments that come into play, little support is provided to organize and manage the increasing complexity of the agents themselves... Inevitably, when we reach a quite advanced stage of evolution of the model, the agents eventually reach a critical mass (either in formalization or code) that makes them more and more hard to comprehend. In this paper, we illustrate this phenomenon and show that it quickly takes the upper hand against the benefits of DOM, as it can eventually block the potential development, or even reuse, of the model. Then we explain that a solution to this ”side effect” could structure the architecture of agents, a structure capable of maintaining readability and flexibility of the formalization of the agent throughout the growth process of the global model.We study a well known pattern in software engineering: the MVC pattern, which can be reused here to meet this objective. We will present in details how this pattern could be instantiated in the field of MAS architecture, and how, ultimately, it can be an effective new way to formalize agents in a method called Multi- Behaviors Modelization.
... In the following, four prominent approaches are described in more detail: Fischer's MAGSY, the GRATE* architecture by Jennings, Steiner's MECCA system, and the COSY architecture developed by Sundermeyer and Burmeister. For further reading, see, e.g., George (1983), Finin and Fritzson (1994), Rosenschein and Zlotkin (1994), McCabe and Clark (1995), Barbuceanu and Fox (1996), Chaib-Draa (1996) and May®eld et al. (1996). ...
The objective of this paper is twofold. In its first part, we survey the state of the art in research on agent
architectures. The architecture of an agent describes its modules and capabilities, and how these operate together. We structure the field by investigating three important research threads, i.e. architectures for reactive agents, deliberative agents and interacting agents. Then we describe various hybrid approaches that reconcile these three threads, aiming at a combination of different features like reactivity, deliberation and the ability to interact with other agents. These approaches are contrasted with architectural issues of recent agent-based work, including software agents, softbots, believable agents, as well as commercial agent-based systems. The second part of the paper addresses software engineers and system designers who are interested in applying agent technology to their problem domains. The objective of this part is to assist these readers in deciding which agent architecture to choose for a specific application. We characterise the most important domains to which the different approaches described in the first part have been applied, propose an application-related taxonomy of agents, and give a set of guidelines to select the right agent (architecture) for a given application.
... One of the major goals of distributed articial intelligence is to obtain, through interactions between agents, eective and useful emergent decisions or computations fromm relatively simple actions by each agent in a given environment (e.g. [5, 13, 14, 15, 26]). Research toward this goal faces the dicult challenge of designing each agent's actions and interactions with others in an ever changing environment to obtain a desired macroscopic computation or action. ...
One of the major challenges of distributed artificial intelligence is to obtain useful and effective emergent behaviors of agents in the system based on the local decision-making of each agent. The effectiveness of the system as a whole is as much dependent on the form of interactions between agents as on the capabilities or strategies of each one. The focus and main theme of this paper is to put forth the idea of distributed interactions through an elementary medium. In this conceptual picture, the interaction between agents is distributed over the collective behavior of a basic unit or particle, which we call computon. The key feature of distributed interaction is the distribution of contents of information among objects or computational agents. This model allows each agent to make a decision on its behavior based on simple all-local transactions for a possibly effective emergent collective behavior. In order to evaluate and examine the feasibility and possibility of distributed interactions, We consider two examples of distributed interaction models with computons. The first example is a conceptual discussion of a Quantized Computational Field Model. In this model, computon is introduced as a fundamental particle of the computational field. Interaction of objects and computational field is envisioned as an interaction between objects via exchanges of computons. To gain more quantitative insight into distributed interactions, we constructed a model using computons to address the problem of load balancing. A dynamic load balancing model applied to a ring of processors was investigated using simulations. When compared with a load balancing model without computons, the load was found to be distributed better over a model ring of processors. Through these examples, we infer and discuss general advantages and problems of distributed interactions among distributed agents or computational resources.
... However, in dynamic systems, allocation of tasks in real-time may be necessary. Task allocation can be done using a single centralized agent (e.g., [22, 37]), or by a collection of agents (e.g., [18, 35, 43]). In this paper we consider the problem of distributed dynamic task allocation by a collection of agents. ...
This paper considers the problem of distributed dynamic task allocation by a set of cooperative agents. The paper describes a rather specific situation. However, its methods have wide application and, thus, it can be useful to solve general problems of computer science. One of its main ideas is to combine optimization questions with the symmetries of initial objects. There are different types of tasks that are dynamically arriving to a system. Each of the agents can satisfy only a subset of the tasks. The main goal of the agents is to maximize the overall performance of the system and to fulfill the tasks as soon as possible. The agents are modeled using a stochastic closed queueing network. The problem is divided into two subproblems: to determine a distributed policy of optimal task distribution and to find the optimal effort levels of the agents subject to certain constraints. For the first subproblem, a distributed polynomial allocation algorithm for determining an instantaneous probabilistic optimal policy for task allocation is presented. The policy is independent of the state of the system and thus does not require information exchange among the agents during the performance of the tasks. For the second subproblem, an analytical solution to find the optimal effort levels for the agents is given.
... Negotiation has been a subject of central interest in DAI, proposed specially as a means for autonomous, self-interested agents to communicate and compromise in Multiagent Systems (Conry et al.the main focus of DAI research from a logical perspective has been that of \planning for multiple agents", which considers issues inherent in centrally directed multi-agent execution (George , 1983;George , 1984;George , 1986;Rao et al., 1992;Rosenschein, 1982). We propose a logical framework to represent an incentive negotiation procotocol about task distribution in bilateral Multi-Agent Systems. ...
... AI research in planning for multiple agents can be classi ed into two main approaches. The rst is centralized multiagent planning, in which one of the agents, chosen either by the other agents Cammarata et al., 1983] or by the human designer George , 1983], collects the separate plans of the agents and analyzes them to nd potential con icts. It resolves con icts by suggesting alterations to plans or by inserting additional synchronization steps. ...
this paper, I describe a line of research that has led to a view of coordination as a distributed search problem, where the search space provides a common representation for organizations, plans, and schedules. The evolving approach is thus a candidate foundation for a truly interdisciplinary study of coordination from both computational and social perspectives
... The command driven / master agent relationship employs traditional centralized planning or distributed centers of planning control. These approaches are described by Corkill [1979] and Georgeff [1983]. Planning under the influence of a master agent has been examined specifically by Ephrati and Rosenschein [1992]. ...
The level of autonomy at which individual agents function is of critical importance to the overall operation of multi-agent systems. The term level of autonomy refers to the type of interactions between an agent and other agents in its system. In well-defined contexts, agents can be designed for a single level of autonomy by predicting the type of problems that will be faced. However, in dynamic systems, the appropriate level of autonomy may depend on the situation. Therefore, substantial performance benefits for agent-based systems can be realized by agents that are capable of dynamically adapting their level of autonomy during system operation. This paper develops a representation for agent autonomy level and discusses how dynamic adaptive autonomy can be used to create flexible multi-agent systems applicable to manufacturing environments. Accepted to Artificial Intelligence and Manufacturing: A Research Planning Workshop Albuquerque, NM Contact person: Leslie Interrante...
... The study of concurrent action in relation to planning (Georgee 1984) has improved our understanding of how agents can reason about their interactions in order to avoid connicts during concurrent plan execution . Connicts can be avoided by reducing or eliminating interactions by localizing plan eeects to particular agents (Lansky 1990), and by merging the individual plans of agents by introducing synchronization actions (Georgee 1983). In fact, planning and merging can be interleaved, such that agents can propose next-step extensions to their current plans and reconcile connicts before considering extensions for subsequent steps. ...
Interacting agents that interleave planning, plan coordination, and plan execution for hierarchical plans (e.g. HTNs or procedures for PRS) should reason about abstract plans and their concurrent execution before they are fully refined. Poor decisions made at abstract levels can lead to costly backtracking or even failure. We claim that better decisions require information at abstract levels that summarizes the preconditions and effects that must or may apply when a plan is refined. Here we formally characterize concurrent hierarchical plans and a method for deriving summary information for them, and we illustrate how summary conditions can be used to coordinate the concurrent interactions of plans at different levels of abstraction. The properties of summary conditions and rules determining what interactions can or might hold among asynchronously executing plans are proven to support the construction of sound and complete coordination mechanisms for concurrent hier...
... Negotiation has been a subject of central interest in DAI, proposed specially as a means for autonomous, self-interested agents to communicate and compromise in Multiagent Systems (Conry et al.the main focus of DAI research from a logical perspective has been that of \planning for multiple agents", which considers issues inherent in centrally directed multi-agent execution (George , 1983;George , 1984;George , 1986;Rao et al., 1992;Rosenschein, 1982). We propose a logical framework to represent an incentive negotiation procotocol about task distribution in bilateral Multi-Agent Systems. ...
Negotiation has been a subject of central interest in DAI, proposed specially as a means for autonomous, self-interested agents to communicate and compromise in Multiagent Systems (Conry et al., 1988; Kraus, 1996; Kraus et al., 1995; Rosenschein and Zlotkin, 1994; Smith, 1980). However, the main focus of DAI research from a logical perspective has been that of "planning for multiple agents", which considers issues inherent in centrally directed multi-agent execution (Georgeff, 1983; Georgeff, 1984; Georgeff, 1986; Rao et al., 1992; Rosenschein, 1982). We propose a logical framework to represent an incentive negotiation procotocol about task distribution in bilateral Multi-Agent Systems. We first extend Bell's logic (Bell, 1995) to represent autonomous agents by adding utilities to his many-sorted first-order modal logic, and then introduce a strategic model of negotiation that takes into consideration goal relationships. In so doing we represent not just a communication mecha...
An intelligent agent is an entity that performs its task in a given environment by exploiting the knowledge acquired from its interaction with the environment during problem-solving process. Over the past two decades, multi-agent systems have emerged as a new methodology to address the issue of organizing a large-scale system by assembling and coordinating individual agents to achieve a goal jointly. Remarkable features of multi-agent systems resulting in their immense real-world applications include low implementation cost, adaptability with dynamicity of environment, enhanced flexibility, great robustness, and ease of maintenance. A multi-agent system is primarily characterized by goal-oriented coordination among its agents, both in cooperative and in competitive circumstances. This chapter introduces the basic concepts of cooperative and competitive multi-agent coordination. It begins with formal definitions of agency and elaborately discusses the perceptual and learning capability of an agent based on its architecture. Gradually, the chapter unveils the emergence of multi-agent coordination due to handshaking of distributed artificial intelligence and machine intelligence. The chapter then highlights the significance of planning and learning in multi-agent coordination to solve real-world problems. The chapter next demonstrates the scope of evolutionary optimization algorithms to maximize coordination efficiency in multi-agent robotics by optimal utilization of system resources. The chapter ends with a discussion on enhancing the performance of traditional evolutionary optimization algorithms to handle measurement noise in real-world multi-robot coordination problems.
Multiagent systems serve as an enabling technique for modern manufacturing and logistics applications; e.g., in “Industry 4.0”. However, these settings require agents to coordinate their actions in a complex environment. This gives rise to the following questions:
How can autonomous agents coordinate their distributed activities
a) with incomplete knowledge and limited perception of their
environment?
b) in mixed settings with cooperating as well as competing entities?
c) in the presence of concurrently acting other agents?
To answer these questions, Jan Ole Berndt presents a novel approach to multiagent coordination in Distributed Artificial Intelligence. Inspired by Niklas Luhmann’s theory of social systems, he investigates social order generation in concurrent negotiation processes. This book provides an interdisciplinary foundation of coordination and connects sociological concepts with reinforcement learning techniques. It includes a comprehensive formalization of Self-Organizing Multiagent Negotiations as well as a formal analysis of their algorithmic properties. In addition, that method is evaluated in a simulation and its practical application to a real-world logistics coordination problem is demonstrated.
Multiagent systems serve as an enabling technique for modern manufacturing and logistics applications; e.g., in “Industry 4.0”. However, these settings require agents to coordinate their actions in a complex environment. This gives rise to the following questions:
How can autonomous agents coordinate their distributed activities
a) with incomplete knowledge and limited perception of their
environment?
b) in mixed settings with cooperating as well as competing entities?
c) in the presence of concurrently acting other agents?
To answer these questions, Jan Ole Berndt presents a novel approach to multiagent coordination in Distributed Artificial Intelligence. Inspired by Niklas Luhmann’s theory of social systems, he investigates social order generation in concurrent negotiation processes. This book provides an interdisciplinary foundation of coordination and connects sociological concepts with reinforcement learning techniques. It includes a comprehensive formalization of Self-Organizing Multiagent Negotiations as well as a formal analysis of their algorithmic properties. In addition, that method is evaluated in a simulation and its practical application to a real-world logistics coordination problem is demonstrated.
One of the main challenges in human–swarm interactions is the construction of suitable abstractions that make an entire robot team amenable to human control. For such abstractions to be useful, they need to scale gracefully as the number of robots increases. In this work, we consider the use of time-varying density functions to externally influence a robot swarm. Density functions abstract away the size of the robot team and describe instead the concentration of agents over the domain of interest. This allows a human operator to design densities so as to manipulate the robot swarm as a whole, instead of at the individual robot level. We discuss coverage of time-varying density functions as a mechanism to translate densities into robotic movement, and provide a series of control laws that guarantee optimal coverage by the robot team. Distributed approximations allow the solutions to scale with the size of the robot team. This renders coverage a viable choice of method for influencing a robot swarm. Finally, we provide a framework for the design of density functions that shape the swarm to achieve specified geometric configurations within the domain of interest. We show through robotic implementation in two different platforms the viability of human–swarm interactions
with the proposed schemes.
Partial global planning is used to provide a framework for coordinating multiple AI systems that are cooperating in a distributed sensor network. By combining a variety of coordination techniques into a single, unifying framework, partial global planning enables separate AI systems to reason about their roles and responsibilities as part of group problem solving, and to modify their planned processing and communication actions to act as a more coherent team. Partial global planning is uniquely suited for coordinating systems that are working in continuous, dynamic, and unpredictable domains because it interleaves coordination with action and allows systems to make effective decisions despite incomplete and possibly obsolete information about network activity. The authors implement and evaluate partial global planning in a simulated vehicle monitoring application and identifying promising extensions to the framework
Planning systems have been an active research topic within Artificial Intelligence for over two decades. There have been a number of techniques developed during that period which still form an essential part of many of today's planners. This paper introduces the techniques, attempts to classify some of the important research themes in Al planning and describes their historical development. There has been a recent surge of interest in planning systems at both the research and applications levels. This paper should act as a brief introduction to the large volume of literature on the subject. An extensive bibliography is provided and cross-referenced throughout the text.
Optimisation is a requirement of many commercial as well as theoretical problems; the optimal use of available resources, maximising savings, minimising costs, extending useful lives, increasing time available for other tasks, using fewer resources. In many of these problems, there are reasons, geographic, political or arbitrary, for splitting the problem into smaller parts and solving them separately, but the need to optimise the combination is still present. In this report I chart the progress in two problem arenas that ooer the potential for the use of a distributed method of optimisation. The rst, Teaching Space Utilisation, seeks a union between local and global optima of similar problem solvers operating in a centralisable domain. The second, British Airways' Aircraft Retimer, shares this in part of its problem, but has additional aspects that means it cannot be centralisable, there are non-friendly, potentially hostile peers, as well as the need to integrate heterogeneous problem solvers with only partial global information into a globally optimal solution.
Coordination can be required whenever multiple agents plan to achieve their individual goals independently, but might mutually benefit by coordinating their plans to avoid working at cross purposes or duplicating effort. Although variations of such problems have been studied in the literature, there is as yet no agreement over a general characterization of them. In this paper, we formally define a common coordination problem subclass, which we call the Multiagent Plan Coordination Problem, that is rich enough to represent a wide variety of multiagent coordination problems. We then describe a general framework that extends the partial-order, causal-link plan representation to the multiagent case, and that treats coordination as a form of iterative repair of plan flaws between agents. We show that this algorithmic formulation can scale to the multiagent case better than can a straightforward application of the existing plan coordination techniques, highlighting fundamental differences between our algorithmic framework and these earlier approaches. We then examine whether and how the Multiagent Plan Coordination Problem can be cast as a Distributed Constraint Optimization Problem (DCOP). We do so using ADOPT, a state-of-the-art system that can solve DCOPs in an asynchronous, parallel manner using local communication between individual computational agents. We conclude with a discussion of possible extensions of our work.
Due to recent market challenges organizational researchers have developed a variety of strategies how organizations can continuously survive in highly dynamic, sometimes even hostile environments. One of the most important strategies aims to enhance the flexibility of enterprises through widespread decentralization, while another well-known approach advocates customer orientation through systematic business process (re-) engineering. This paper addresses organizational flexibility and business process orientation from the perspective of information systems. It starts from a requirements analysis which investigates the challenges of contemporary organizational strategies and then proceeds towards an approach that supports the flexible modeling of business processes by linking decentralized organizational procedures. For this purpose a set of process modeling and process interaction operators is defined. These operators also allow to automatically create and customize configurations of computerized business processes. This progress in cooperative information processing technology contributes significantly to the recently emerged concept of the computerized enterprise. The concepts are presented in the context of a banking application, namely the Credit Advisory Subsystem of our banking application MAMBA.
LNCS n°1038, http://dx.doi.org/10.1007/BFb0031848 One of the major interests of Multi-Agent Systems (MAS), which are able to handle distributed planning, is coordination. This coordination requires both an adequate plan representation and efficient interacting methods between agents. Interactions are based on information exchange (e.g. data, partial or global plan) and allow agents to update their own plans by including the exchanged information. Coordination generally produces two effects: it cancels negative interactions (e.g. resource sharing) and it takes advantage of helpful ones (e.g. handling redundant actions). A coordination model should satisfy the following requirements: domain independence, broad covering of interacting situations, operational coordination semantics and natural expression for the designer. This paper presents an adequate framework for the representation and handling of plans in MAS. It then shows how an approach based on a plan representation by means of a partial order model enables the definition of a coordination algorithm for the possible enrichment of plans. oui
Research in multiagent systems (MAS), or distributed artificial intelligence (DAI), dates back to the late 1970s. Initial work in the area focused on distributed interpretation of sensor data, organizational structuring and generic negotiation protocols. But several recent developments have helped reshape the focus of the field. Like the rest of Al, the field has matured from being largely exploratory in nature to focusing on formal theories of negotiation, distributed reasoning, multiagent learning and communication languages. The field is also maturing to the point of developing its first few fielded applications. The recent widespread interest in the Internet, the World Wide Web, and intelligent agent applications have further fueled the need for techniques and mechanisms by which agents representing users can interact effectively with other agents in open, dynamic environments. The development of several new international workshops and conferences has helped focus research in the area. The field is poised at a critical juncture, with stimulating problems and challenges promising some very exciting developments in the next few years.
Resumen La investigación en el diseño de sistemas de toma de decisión colectiva busca mejorar las tasas y la robustez de clasificación. Los clasificadores que tienen algún grado de diversidad en su conocimiento pueden presentar diferentes patrones de errores. Esta observación sugiere que un sistema de clasificadores, dependiendo de la inteligencia de su combinación, puede mejorar la precisión de un único clasificador. La eficacia del sistema radica en la capacidad del método de combinación de evitar el traslado de errores individuales en la decisión colectiva. Esta tesis está dedicada al estudio del diseño eficiente de sistemas de clasificación conjunta. El método propuesto considera un conjunto de número variable de clasificadores heterogéneos, generalistas, y no adaptados a una regla de combinación específica. La propuesta de diseño permite una selección de subconjuntos cooperativos haciendo uso de la caracterización del comportamiento individual y colectivo de los clasificadores. Los aspectos del comportamiento considerados en la selección están condicionados en distintos grados al método de combinación a ser usado, pero en ningún caso se utilizará una descripción exhaustiva. Abstract The research on the design of collective decision making systems looks for the classification rates and robustness improvement. The diversity on the classifiers knowledge may provide different error patterns. This observation suggests that multiple classifiers can enhance the accuracy of a single one, depending on the combination cleverness. The key point for the multiclassifier efficacy is the ability of decision combination method to avoid the transference of individual mistakes to the collective result. This thesis is devoted to the study of efficient design of classifier systems, starting from a dynamic set of heterogeneous classifiers, which are generalist and non adapted to a specific combination rule. The proposed designing method allows a systematic selection of cooperative classifiers subsets by utilizing the characterization of individual and collective behavior of classifiers. The considered behavior characterization can be more or less combination rule adaptive, but in no case the exhaustive classifiers description will be considered.
This thesis describes an expert system to aid in the management of operations in complex qualitative domains characterized by multiple parallel activities with time-critical relationships. An extension to "standard" temporal logic required for reasoning about inferred allocation of resources and a detailed representation of temporally dependent facts, including persistence, is presented. The non-linear planning paradigm commonly used in planning programs is extended into the temporal domain to facilitate scheduling as well as ordering of plan steps. This enhancement requires new structures and analytical methods for the detection and resolution of serendipitous interactions and conflicts between proposed schedules. A computer implementation of these concepts is discussed in detail. The expert system is organized into three modules: the time map manager or temporal database manager which stores, organizes, and retrieves time dependent knowledge; the temporal system analyzer which uses this knowledge to forecast and analyze domain dynamics; and the planner/scheduler which formulates and schedules activities in order to satisfy goals generated by the temporal system analyzer. Finally, Tower Chief, an application of the system to scheduling runway configuration changes and maintenance at large airports, is described.
The authors present an overview of cooperative distributed problem solving (CDPS), an emerging research area that combines aspects of AI (artificial intelligence) and distributed processing. CDPS can be used to study how a loosely coupled network of sophisticated problem-solving nodes can solve a complex problem which consists of a set of interdependent subproblems. Subproblems arise because of spatial, temporal, and functional distribution of data, knowledge, and processing capabilities. Application areas include distributed interpretation, distributed planning and control, cooperating expert systems, and computer-supported human cooperation. The authors survey the important approaches and empirical investigations that have been developed. The approaches covered include negotiation, functionally accurate cooperation, organizational structuring, multiagent planning, sophisticated local control, and theoretical frameworks
An artificial social system is a set of restrictions on agents' behaviors in a multiagent environment. Its role is to allow agents to coexist in a shared environment and pursue their respective goals in the presence of other agents. This paper argues that artificial social systems exist in practically every multi-agent system, and play a major role in the performance and effectiveness of the agents. We propose artificial social systems as an explicit and formal object of study, and investigate several basic issues that arise in their design.
this paper is to examine the crucial area of co-ordination in multi-agent systems. It does not attempt to provide a comprehensive overview of the co-ordination literature; rather, it highlights the necessity for co-ordination in agent systems and overviews briefly various co-ordination techniques. It critiques these techniques and presents some conclusions and challenges drawn from this literature. 1. Introduction The scope of this paper is limited mainly to software multi-agent systems (MAS). The breakdown of the paper is as follows. Section 2 argues more cogently why coordination is vital, not only in DAI and agent-based systems, but also to open distributed systems. Section 3 overviews briefly various co-ordination techniques. Negotiation, overviewed in section 4, is one of the techniques used to achieve co-ordination in agent-based systems. It is overviewed separately because it has an extensive literature in its own right. Section 5 presents some lessons learned and some heuristics on co-ordination, and section 6 concludes the paper.
ADONIS is an adaptive optics user friendly instrument for the ESO 3.6-m telescope. The modal control technique is presented as a tool for optimizing the control of ADONIS. The perturbed wavefront is spread over a base of modes treated by the system in terms of their behavior with respect to the observing conditions (turbulence characteristics, anisoplanatic angle and star magnitude). This optimization and the multiple parameters influencing the system performance are difficult to handle by the astronomer. Moreover, the wide range of situations the astronomical AO system has to deal with requires a flexible configuration which has to be optimized for each case. The ADONIS instrument also has to be easy to use by an astronomer who doesn't have a deep knowledge of AO techniques. Finally, it must make efficient use of the telescope time. Hence, a new human interface and an intelligent control system have been designed. 1 INTRODUCTION The ADONIS instrument 2,14 is based on the adaptive o...
The objective of this paper is to examine the crucial area of coordination in multi-agent systems. It does not attempt to provide a comprehensive overview of the coordination literature; rather, it highlights the necessity for coordination in agent systems and overviews briefly various coordination techniques. It critiques these techniques and presents some conclusions and challenges which we draw from this literature. 1 Introduction Coordination is a central issue in software agent systems in particular and in Distributed Artificial Intelligence (DAI) in general. However, it has also been studied by researchers in diverse disciplines in the social sciences including organisation theory, political science, social psychology, anthropology, law and sociology. For example, organisation theorists have investigated the coordination of systems of human beings, from small groups to large formal organisations [13,38]. Economists have studied coordination in markets of separate profit-maximisi...
. This paper introduces ECM, a new coordination model, and one of his language binding, STL++. STL++ is aimed at making parallel and distributed programs' construction easier by providing a set of powerful tools that clearly describe the coordination part of an application. These tools, embedded in an object-oriented computation language, enable distributed components to be dynamically reconfigured. STL++ acts also as a coordination language especially well suited for distributed implementations of multi-agent systems. It is aimed at giving basic constructs for the implementation of generic multi-agent platforms. The underlying architecture on which STL++ develops, is based on Pt-pvm, a software platform providing rich message passing facilities for lightweight processes (threads), which is itself built on top of PVM. Keywords: Coordination, Distributed Systems, Threads, Concurrency, Communication Models, Agents, Multi-Agent Systems, PVM, Pt-pvm. 1 Introduction Coordination constit...
One of the major interests of Multi-Agent Systems (MAS), which are able to handle distributed planning, is coordination. This coordination requires both an adequate plan representation and efficient interacting methods between agents. Interactions are based on information exchange (e.g. data, partial or global plan) and allow agents to update their own plans by including the exchanged information. Coordination generally produces two effects: it cancels negative interactions (e.g. resource sharing) and it takes advantage of helpful ones (e.g. handling redundant actions). A coordination model should satisfy the following requirements: domain independence, broad covering of interacting situations, operational coordination semantics and natural expression for the designer. This paper presents an adequate framework for the representation and handling of plans in MAS. It then shows how an approach based on a plan representation by means of a partial order model enables the definition of a coordination algorithm for the possible enrichment of plans. Keywords: Multi-Agent Systems, Distributed Planning, Coordination, Plan Interactions. 1
There has been increased research interest in systems composed of multiple autonomous mobile robots exhibiting collective behavior. Groups of mobile robots are constructed, with an aim to studying such issues as group architecture, resource conflict, origin of cooperation, learning, and geometric problems. As yet, few applications of collective robotics have been reported, and supporting theory is still in its formative stages. In this paper, we give a critical survey of existing works and discuss open problems in this field, emphasizing the various theoretical issues that arise in the study of cooperative robotics. We describe the intellectual heritages that have guided early research, as well as possible additions to the set of existing motivations. 1 Preliminaries There has been much recent activity toward achieving systems of multiple mobile robots engaged in collective behavior. Such systems are of interest for several reasons: (1) tasks may be inherently too complex for a singl...
ResearchGate has not been able to resolve any references for this publication.