Marcus J. Huber

JAM: a BDI-theoretic mobile agent architecture

JAM is a hybrid intelligent agent architecture that draws upon the theories and ideas of the Procedural Reasoning System (PRS), Structured Circuit Semantics (SCS), and Act plan interlhtgua. Furthermore, JAM draws upon the implementation pragmatics of the University of Michigan’s and SRI Internatlonal’s implementation of PRS (UMPRS and PRS-CL, respectively). JAM provides rich and extensive plan and procedural representations, metalevel and utility-based reasoning over multiple simultaneous goals, and goal-driven and event-driven behavior that are an amalgam of all of the sources listed above. The JAM agent architecture also provides an agentGo primitive function utilizing Java’s object serialization mechanism to provide widely-supported mobility capabilities.

Representing and executing protocols as joint actions

Families of conversation protocols can be expressed formally as partially ordered landmarks where the landmarks represent the state of affairs that must be brought about during the goal-directed execution of a protocol. Then, concrete protocols represented as joint action expressions can be derived from the partially ordered landmarks and executed directly by joint intention interpreters, thus nearly eliminating the need to implement a separate protocol handling system. This approach also supports (1) flexibility in the actions used to achieve landmarks, (2) shortcutting protocol execution, and (3) application of the joint intention theory to provide automatic exception handling along with a correctness criterion for protocols.

CARMEL Versus FLAKEY A Comparison of Two Winners

■ The University of Michigan’s CARMEL and SRI International’s FLAKEY were the first- and secondplace finishers, respectively, at the 1992 Robot Competition sponsored by the American Association for Artificial Intelligence. The two teams used vastly different approaches in the design of their robots. Many of these differences were for technical reasons, although time constraints, financial resources, and long-term research objectives also played a part. This article gives a technical comparison of CARMEL and FLAKEY, focusing on design issues that were not directly reflected in the scoring criteria.

Integrated mobile-robot design-Winning the AAAI 1992 robot competition

The Carmel project (computer-aided robotics for maintenance, emergency, and life support) which won the AAAI 1992 Robot Competition, is discussed. Carmels design philosophy and architecture, obstacle avoidance, global path planning, vision sensing, landmark triangulation, and supervisory planning system are described. The Carmel project shows that mobile robots can perform carefully chosen tasks reliably and efficiently, although this requires extensive integration of components and a solid engineering effort. >

Multiple roles, multiple teams, dynamic environment: autonomous Netrek agents

We describe the architecture and performance of autonomous agents that play the complex, multi-faceted internet game called Netrek. To perform competently within Netrek, an agent must be capable of 1) reacting in real-time to arcade-like tasks and environment changes within the local surroundings, 2) reasoning about strategy and long-term tasks at a slower pace and on a global scale, and 3) coordinating both with teammates and against members of the opposing team. Agents capable of operating within the Netrek environment, therefore, require a reactive, exible, multi-level framework. We describe in detail our current agent implementation, which is based upon the UMPRS (University of Michigan Procedural Reasoning System) architecture. As multi-agent coordination plays a signi cant role within the Netrek domain, we describe recent research in extending the current communication-based intra-team coordination scheme with a plan recognition scheme that enhances both intraand inter-team coordination.

TAIPE: tactical assistants for interaction planning and execution

Downsizing the number of operators controlling complex systems can increase the decision-making demands on remaining operators, particularly in crisis situations. An answer to this problem is to offload decision-making tasks from people to computational processes, and to use these processes to focus and expedite human decision making. In this paper, we describe a system comprised of multiple computational agents that has demonstrated an ability to help operators prioritize their tasks better, process their tasks faster, and enlist the aid of other operators more transparently. In developing this system, we have of course encountered challenges, particularly in devising content languages that adequately convey the right information (to be interpreted correctly) across the heterogeneous agents. We here summarize our work that addresses this challenge, and illustrate how our system improves performance for operators in naval situations.

Agent Autonomy: Social Integrity and Social Independence

As interactions between agents become more common, it will become very important to be able to characterize and perhaps even guarantee an agents level of autonomy. We will both want agents to perform tasks on their own while at the same time both remaining controllable by ourselves and secure from control and manipulation by others. Most intuitions of autonomy seem to involve the notion that it is related to dependence/independence. Our model of autonomy captures the notion that, in one sense, autonomy represents security from corruption and manipulation by external influences (i.e., its social integrity). Another aspect of our model captures that autonomy also represents an agents ability to perform its tasks without dependence upon others, (i.e., its social independence). This paper therefore presents a multidimensional conceptualization of autonomy and introduces a pragmatic interpretation of our scheme that is applicable to the characterization of the autonomy level of any software entity but which is especially amenable to agent-based systems

Toward a suite of performatives based upon joint intention theory

Agent communication languages defined using joint intention theory have enjoyed a long research history. A number of performatives have been defined and refined in this literature with particular emphasis on the basic performatives of REQUEST and INFORM, which subsequently have many subtle versions. Even these less common performatives have been extended and refined multiple times. In many cases the underlying definitions upon which the various performatives are based have been modified as well. While working toward implementing a multi-agent system with communications based upon joint intention semantics, it quickly became apparent that it was going to be difficult to identify a single set of performatives with correct and compatible definitions. We also realized that a set of performatives with enough breadth to cover the needs of real fielded multi-agent systems has not yet been defined. We intend this paper to provide in a single place a broadly applicable set of compatible performatives defined using joint intention semantics. Many of the performatives previously defined in the literature have been brought to the same semantic basis while we have also defined a number of new performatives to increase the breadth of performatives available to agent developers.

Direct execution of team specifications in STAPLE

Formal theories of teamwork are typically treated as software design specifications of team behavior. We take a different approach to programming teamwork by directly executing logical specifications of joint commitment and joint intention. This approach leads to a domain-independent framework for programming teamwork where one can modify (or add new) behavior for a team of agents just by modifying (or adding) logical sentences. One may also be able to predict the behavior of an agent team offline using its team intention specification and verify it by running the actual system.

The Search for Coordination: Knowledge-Guided Abstraction and Search in a Hierarchical Behavior Space

Coordination is a search process, where individuals must find appropriate activities that allow them to achieve individual and collective goals. In this paper, we motivate and summarize the elements of coordination search, and use these elements to highlight how traditionally distinct coordination techniques can be viewed as similar search processes but at different levels of abstraction. In particular, the temporal extents and relationships among activities leads to search as organizational design, as multiagent planning, or as distributed resource scheduling. We update the current status of some of our search techniques for coordination, emphasizing issues such as the effects of abstraction in controlling distributed search and how knowledge can be used in hypothesizing alternative activities. The practical implications of coordination search are briefly illustrated in the context of multirobot control and distributed meeting scheduling.