Pavel Tichý

Distributed multi-agent architecture for automation systems

Abstract In the 21st century, industrial automation will be greatly benefited by the advances in electronics, information systems, and process technology. However, these technological advances are still separate islands of automation. We believe that multi-agent systems will help the future of automation by providing flexible and scalable ways to integrate the different parts. This paper reports preliminary results of an ongoing research project that demonstrates advanced automation in a highly distributed architecture that is made of a synergy of intelligent agents, control, and physical devices. This was built to achieve the goals of reduced manning and improved readiness and survivability in US Navy shipboard systems.

Industrial MAS for Planning and Control

In this paper, we summarize advantages of using agent technology in the design and operation of industrial control systems. To support appropriate responses to dynamically generated events, we divided the agent operation into two main levels: planning and real-time control. A higher-level software agent covers the planning phase and low-level distributed control agent provides the real-time control. The proposed architecture is targeted on applications where real-time response is essential but also more sophisticated higher-level control mechanisms are needed for efficient control. An implementation of such system is briefly described.

An Intelligent Agent Validation Architecture for Distributed Manufacturing Organizations

In this paper, we focus on validation of Multi-Agent System (MAS) behavior. We describe the simulation architecture and the system design methodology to accomplish the appropriate agent behavior for controlling a real-life automation system. The architecture is explained in the context of an industrial-sized water cooling system. Nevertheless, it is intended to operate in a wide spectrum of control domains. In general, after the design of the control system is accomplished, a set of validation procedures takes place. The current needs are to validate both the control and the agent levels as integrated parts. Hence there is a need to establish a general architecture and methodology for easing the commissioning process of the control solution.

A strategy to implement and validate industrial applications of holonic systems

Classical control systems are based on feedback techniques and models that generally cannot manage computational complexity, nonlinearity and uncertainty. Moreover, classical control cannot adapt well to the variability of the processes under control in a dynamic fashion. However, agent-based control eases combinatorial complexity by enabling a robust partitioning of knowledge and behaviors. It is a difficult challenge to create the infrastructure, development system and validation tools for agent systems. In this paper we discuss fundamental steps to achieve the foundation infrastructure for creating agents but also we address several guidelines to create the agents and the requirements to present this to non-agent specialists.

Chilled Water System Control, Simulation, and Visualization Using Java Multi-agent System

Abstract This paper describes a multi-agent application for control of a simulated shipboard chilled water system (CWS) implemented in Java language. The multi-agent approach provides robustness, flexibility, and scalability. Elements of the CWS are represented by three types of agents. Agents of type “Chiller” provide cold water, agents “Service” require cold water, and agents “Valve” connect segments of water piping system. Agents take actions to dynamically create cooling paths, which enables agents “Service” to stay close to their required temperature. Ability to isolate a leakage, when it occurs, to keep other parts of the system healthy, is another feature of the system. The graphical user interface of the application allows a user to watch running processes in the system as well as edit the system configuration.

Cost-Based Dynamic Reconfiguration System for Evolving Holarchies

Autonomous, highly distributed control architectures are composed of a significant number of holons/agents that can reason and act on behalf of represented processes or artifacts in a coordinated manner. Depending on the social organization capabilities of the agents, the autonomous system could evolve into complex agent organizations called temporal holarchies. Cost-based negotiation supports the holarchy formation. Dynamic hierarchical teamworks architecture of middle-agents is described to increase robustness of the architecture.

Multi-agent system design and integration via agent development environment

Design and integration of multi-agent systems are becoming more and more important since there is evidence of a shift of multi-agent application deployment from academic or lab testing areas to real world applications such as manufacturing, aerospace, logistics, and networking. Therefore, utilization of an environment that guides the development process starting from the design of templates that can be reused multiple times up to the code generation and debugging is essential. In this manuscript we present advancements in the Agent Development Environment that can be used not only for the development of agents designed for high-level decision-making, but also for distributed low-level control that is automatically integrated.

Methods to Observe the Clustering of Agents Within a Multi-Agent System

Cohesion and coupling are standard concepts in computer science. As in classic software, multi-agent systems (MASs) strive for high cohesion and low coupling. But MASs, by definition, do have some coupling, i.e., the agents seek out and discover other agents, and negotiate and cooperate with them, to accomplish some overall goal. Moreover, as different events occur, and different agents and different agent behaviors are called upon, the coupling values for the system change. This paper proposes a methodology for viewing these changing coupling relationships and clustering behaviors. In addition, the methodology is described as implemented as part of a standard agent monitoring tool.

Multi-agent technology for robust control of shipboard chilled water system

Abstract This paper reports preliminary results of an ongoing research project that demonstrates distributed architecture based on agents applied in the area of industrial automation. This architecture has been built to achieve the goals of improved survivability and readiness of US Navy shipboard systems. We show benefits of multiagent systems in the area where flexibility, survivability, and scalability are required. We present the architecture of multi-agent system, internal structure of agent, planning technique based on plan templates, fault-tolerant structure of middle-agents, and development environment for agents.

Multi-Agent Technology for Fault Tolerant and Flexible Control

One of the main characteristics of multi-agent systems (MAS) is fault tolerance. When an agent is unavailable for some reason, another agent with similar capabilities can theoretically compensate for this loss. Many key aspects of fault tolerance in MAS are described in this chapter including social knowledge, physical distribution, agent development, and validation. Therefore, the focus is not only on a fault tolerant agent platform with necessary services (e.g., fault tolerant social knowledge), but also on the design that can significantly reduce mistakes in agent programming and validation that can discover faults that manifest as failures during the testing phase.