Catálogo de publicaciones - libros

When constructing multiagent systems, the designer may approach the system as a collection of individuals or may view the entire system as a whole. In addition to these approaches, it may be beneficial to consider the interactions between the individuals and the whole. Borrowing ideas from the notion of social construction and building on previous work in synthetic social construction, this paper presents a framework wherein autonomous agents engage in a dialectic relationship with the society of agents around them. In this framework, agents recognize patterns of social activity in their societies, group such patterns into institutions, and form computational representations of those institutions. The paper presents a design framework describing this method of institutionalization, some implementation suggestions, and a discussion of possible applications.

Although, the software agent paradigm has been widely accepted, there is still a problem with situating (implementing) concrete agents in real environments. Agent is an abstract notion so that there is no straightforward mapping of real entities like robots, spacecrafts or any other autonomous systems to agents. An interesting discussion on this subject was done within the framework of LOGOS and ACT Agent Architecture in the context of ground and space systems, see [10,11]. In the paper we discuss the problem of situating agents in open systems consisting of cognitive heterogeneous robots that are supposed to perform jointly complex tasks. The starting point of the discussion is the archtecture of the M-agent.

Research done during the last three years has studied the emersion properties of Complex Adaptive Systems (CAS). The deployment of Artificial Intelligence (AI) techniques applied to remote Unmanned Aerial Vehicles has led the author to investigate applications of CAS within the field of Autonomous Multi-Agent Systems. The core objective of current research efforts is focused on the simplicity of Intelligent Agents (IA) and the modeling of these agents within complex systems. This research effort looks at the communication, interaction, and adaptability of multi-agents as applied to complex systems control. The embodiment concept applied to robotics has application possibilities within multi-agent frameworks. A new framework for agent awareness within a virtual 3D world concept is possible where the vehicle is composed of collaborative agents. This approach is considered for application to the complex tetrahedron structure system from NASA Goddard Space Flight Center (GSFC) developed under the Autonomous Nano Technology Swarm (ANTS) program.

A classic autonomous robot is an autonomous agent for open, unpredictable environments. Such an agent is inherently autonomous but not independent. Independence implies unpredictability, which is incompatible with agency. The current robot models – behavior based and artificial intelligence – have not been effective at implementing the classic autonomous robot model due to limitations in their definitions. The artificial intelligence model cannot deal with unpredictable environments, and neither model directly includes the concept of agency. Animal training as a model for robotics has the potential to avoid these problems. Animal training has several advantages. It has an inherent model of agency. Goals and behavior are formally separated into human goals and animal behavior. The animal is autonomous, requiring conversation between human and animal, but it is not an independent entity. A robot designed using this model is an articulate machine, programmed as an agent for the user.

As e-Commerce systems evolve, and B2B models flourish, more and more complex payment transactions are performed electronically. A consistent topic periodically addressed by e-Commerce systems is how to reduce the costs associated with these electronic transactions, such that it is of minimal detriment to the growth in the volume of transactions that are conducted online. This question is of particular interest in the area of micro-payments, where the cost of the transaction is a significant percentage of the total cost of the goods exchanged between the merchant and the buyer. This paper proposes a novel architecture and associated artifacts for online payment processing providers and their client applications (online order capture and processing systems, payment capture and processing systems, and other online-based merchant software). New data structures, message exchange patterns and optimization algorithms are developed to support this new paradigm. These artifacts are molded following the concepts defined by the Autonomic Computing and Intelligent Agents disciplines.

Inspired by the autonomic aspects of the human central nervous system, the vision of “” arrived with a fully-formed wish list of characteristics that such systems should exhibit, essentially those self-referential aspects required for effective self-management. Here, the authors contend that the biologically-inspired managerial cybernetics of Beer’s Viable System Model (VSM) provides significant conceptual guidance for the development of a general architecture for the operation and management of such complex, evolving, adaptive systems. Consequently, the VSM has been used as the basis of a theoretically-supported reference model that provides the "blueprint" for an extensible intelligent agent architecture. Of course, normal use of the VSM relies heavily on human agency to realize the adaptive capabilities required by the model. Therefore, artificially replicating such activities represents a significant challenge, however the authors show that some progress can be made using algorithmic hot swapping and in particular Holland’s Genetic Algorithms (GA’s) to generate, in specific circumstances, a repertoire of tailored responses to environmental change. The authors then speculate on the use of the associated Learning Classifier Systems (LCS) approach to allow the system to develop an adaptive environmental model of appropriate, optimized responses.

In this paper we argue that self-modifying code can become a better strategy for realizing long-lived autonomous software systems than static code, regardless how well it was validated and tested. We base our discussion on three facets – self-repairing software, adaptive software and networked systems – for which we point out ongoing and related work before presenting a roadmap towards a controlled framework for self-modifying code.

This paper introduces a reference architecture for developing agent-based systems that preserves core concepts of agenthood while minimizing cumbersome features found in other agent architectures. This parsimony has been found useful in addressing complex problems.

We address tactical behaviors of unmanned ground vehicles (UGVs) and focus on their coordination within hierarchical units. This research thus addresses multiagent systems but more particularly structures where autonomous parts coordinate to form wholes that in turn are parts in more inclusive wholes. To capture these notions, we introduce the concept of a holarchy (a hierarchy of holons, which are both parts and wholes) and use an enhancement of the Statechart notation called Parts/whole Statecharts. We relate aspects of this notation to the technical notion of common knowledge, which is a necessary condition for coordination, and we sketch techniques for building holarchic models using Parts/whole Statecharts. An implementation of one such model in Matlab’s Stateflow tool is sketched.

The systemic theory associates open systems and complexity closely. This article presents a particular articulation between these two concepts using the Systemion Model. Two types of opening are defined, characterized and illustrated using examples.