Bryan Horling

A survey of multi-agent organizational paradigms

Many researchers have demonstrated that the organizational design employed by an agent system can have a significant, quantitative effect on its performance characteristics. A range of organizational strategies have emerged from this line of research, each with different strengths and weaknesses. In this article we present a survey of the major organizational paradigms used in multi-agent systems. These include hierarchies, holarchies, coalitions, teams, congregations, societies, federations, markets, and matrix organizations. We will provide a description of each, discuss their advantages and disadvantages, and provide examples of how they may be instantiated and maintained. This summary will facilitate the comparative evaluation of organizational styles, allowing designers to first recognize the spectrum of possibilities, and then guiding the selection of an appropriate organizational design for a particular domain and environment.

Using self-diagnosis to adapt organizational structures

The specific organization used by a multi-agent system is crucial for its effectiveness and efficiency. In dynamic environments, or when the objectives of the system shift, the organization must therefore be able to change as well. In this paper we propose using a general diagnosis engine to drive this process of adaptation, using the \tems\ modeling language as the primary representation of organizational information. Results from experiments employing such a system in the Producer-Consumer-Transporter domain are also presented.

Distributed sensor network for real time tracking

In this paper we describe our solution to a real-time distributed resource allocation application involving distributed situation assessment. The hardware configuration consists of a set of reconfigurable sensors at fixed locations, each having local processing and low-bandwidth communication capabilities with other sensor nodes. The objective is to track objects moving in the environment in real-time as best as possible, given uncertainty and constraints on sensor loads, communication, power consumption, action characteristics, and clock synchronization. Once the target is detected, the sensors must communicate and cooperate so that, within a given window of time, the data needed to triangulate the position of the target can be collected. Our solution to this problem decomposes the environment into a number of sectors, where individual sensor nodes in a sector are specialize dynamically to address different parts of the goal. We describe our solution to this problem in detail, including the high-level architecture and a number of the more interesting implementation challenges. Results and future direction are also covered.

BIG: an agent for resource-bounded information gathering and decision making

Abstract The World Wide Web has become an invaluable information resource but the explosion of available information has made Web search a time consuming and complex process. The large number of information sources and their different levels of accessibility, reliability and associated costs present a complex information gathering control problem. This paper describes the rationale, architecture, and implementation of a next generation information gathering system—a system that integrates several areas of Artificial Intelligence research under a single umbrella. Our solution to the information explosion is an information gathering agent, BIG, that plans to gather information to support a decision process, reasons about the resource trade-offs of different possible gathering approaches, extracts information from both unstructured and structured documents, and uses the extracted information to refine its search and processing activities.

Cooperative negotiation for soft real-time distributed resource allocation

In this paper we present a cooperative negotiation protocol that solves a distributed resource allocation problem while conforming to soft real-time constraints in a dynamic environment. Two central principles are used in this protocol that allow it to operate in constantly changing conditions. First, we frame the allocation problem as an optimization problem, similar to a Partial Constraint Satisfaction Problem (PCSP), and use relaxation techniques to derive conflict (constraint violation) free solutions. Second, by using overlapping mediated negotiations to conduct the search, we are able to prune large parts of the search space by using a form of arc-consistency. This allows the protocol to both quickly identify situations where the problem is over-constrained and to identify the appropriate fix to the over-constrained problem. From the global perspective, the protocol has a hill climbing behavior and because it was designed to work in dynamic environments, is an approximate one. We describe the domain which inspired the creation of this protocol, as well as discuss experimental results.

Diagnosis as an integral part of multi-agent adaptability

Agents working under real world conditions may face an environment capable of changing rapidly from one moment to the next, either through perceived faults, unexpected interactions or adversarial intrusions. The members of a multi-agent system can gracefully and efficiently handle such situations by adapting, either by evolving internal structures and behavior or repairing or isolating those external influences believed to be malfunctioning. The first step in achieving adaptability is diagnosis-being able to accurately detect and determine the cause of a fault based on its symptoms. In this paper we examine how domain independent diagnosis plays a role in multi-agent systems, including the information required to support and produce diagnoses. Particular attention is paid to coordination based diagnosis directed by a causal model. Several examples are described in the context of an Intelligent Home environment, and the issue of diagnostic sensitivity versus efficiency is addressed.

The UMASS intelligent home project

Intelligent environments are an interesting development and research application problem for multi-agent systems. The functional and spatial distribution of tasks naturally lends itself to a multi-agent model and the existence of shared resources creates interactions over which the agents must coordinate. In the UMASS Intelligent Home project we have designed and implemented a set of distributed autonomous home control agents and deployed them in a simulated home environment. Our focus is primarily on resource coordination, though this project has multiple goals and areas of exploration ranging from the intellectual evaluation of the application as a general MAS testbed to the practical evaluation of our agent building and simulation tools.

Implementing soft real-time agent control

Real-time control has become increasingly important as technologies are moved from the lab into real world situations. The complexity associated with these systems increases as control and autonomy are distributed, due to such issues as precedence constraints, shared resources, and the lack of a complete and consistent world view. In this paper we describe a real-time environment requiring distributed control, and how we modified our existing multi-agent technologies to meet this need. Two types of enhancements are covered: those which enable planning to meet real-time constraints, such as our task representation, meta-level costing, alternative plan selection, and partial-order scheduling, and those which facilitate on-line real-time control, including scheduling flexibility, caching, and windowed commitments.

The Soft Real-Time Agent Control Architecture

Real-time control has become increasingly important as technologies are moved from the lab into real world situations. The complexity associated with these systems increases as control and autonomy are distributed, due to such issues as temporal and ordering constraints, shared resources, and the lack of a complete and consistent world view. In this paper we describe a soft real-time architecture designed to address these requirements, motivated by challenges encountered in a real-time distributed sensor allocation environment. The system features the ability to generate schedules respecting temporal, structural and resource constraints, to merge new goals with existing ones, and to detect and handle unexpected results from activities. We will cover a suite of technologies being employed, including quantitative task representation, alternative plan selection, partial-order scheduling, schedule consolidation and execution and conflict resolution in an uncertain environment. Technologies which facilitate on-line real-time control, including meta-level accounting, schedule caching and variable time granularities are also discussed.

Farm: A Scalable Environment for Multi-agent Development and Evaluation

In this paper we introduce Farm, a distributed simulation environment for simulating large-scale multi-agent systems. Farm uses a component-based architecture, allowing the simulation to be easily modified and augmented, as well as distributed to spread the computational load and improve running time. Technical details of Farms architecture are described, along with discussion of the rationale behind this design. Performance graphs are provided, along with a brief discussion of the environments currently being modeled with Farm.