Marcus J. Huber

University of Michigan | U-M · Division of Computer Science and Engineering

Most agent research seeks insights about a single technology, and problems are chosen to allow this focus. In contrast, many real-world applications do not lend themselves to a single technology, but require multiple tools. In an applied AI company, each tool often has its own advocate, whose specialized knowledge may lead her to overestimate her t...

To coordinate with other agents in its environment, an agent needs models of what the other agents are trying to do. When communication is impossible or expensive, this information must be acquired indirectly via plan recognition. Typical approaches to plan recognition start with a specification of the possible plans the other agents may be followi...

Differentiating between normal human activity and aberrant behavior via closed circuit television cameras is a difficult and fatiguing task. The vigilance required of human observers when engaged in such tasks must remain constant, yet attention falls off dramatically over time. In this paper we propose an architecture for capturing data and creati...

Many agents are fielded within environments requiring modeling traditional organizational structures such as military hierarchies and corporations, with their associated authority relationships and a strong form of responsibility associated with subordinate agents. Addressing the needs above, we have developed an integrated semantic framework for m...

As interactions between agents become more common, it will become very important to be able to characterize and perhaps even guarantee an agent's 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 intuition...

This paper describes a BDI-inspired abstraction layer that is implemented in Soar to increase the scalability and flexibility of knowledge engineering in Soar-based agent systems. We motivate the BDI abstraction with a specific example of a multi-agent system that is being developed to explore human command and control of heterogeneous robotic enti...

Many agents are fielded within environments requiring modeling traditional organizational structures such as military hierarchies and corporations, with their associated authority relationships and a strong form of responsibility associated with the subordinate agents. Furthermore, communications between agents placed in such environments benefit f...

This paper describes a system for implementing adjustable autonomy levels in simulated unmanned vehicles using an approach based upon the fields of deontics and Joint Intention Theory (JIT). It discusses Soar Technology's Intelligent Control Framework research project (ICF), the authors' use of deontics in the creation of adjustable autonomy for IC...

Trends in combat technology research point to an increasing role for uninhabited vehicles in modern warfare tactics. To support increased span of control over these vehicles human responsibilities need to be transformed from tedious, error-prone and cognition intensive operations into tasks that are more supervisory and manageable, even under inten...

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 processe...

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 ex...

Mediation services are becoming increasingly important in multiagent systems.

Less than half of the individuals of working age with visual impairments are employed and a significant barrier to employment is effective computer access. Screen reader applications offer some help but have limited context sensitivity and are of limited use in applications with dynamic “interfaces” like web pages. Sophisticated screen readers prov...

Conversation protocols are meant to achieve certain tasks or to bring about certain state of affairs in the world. Therefore, one may identify the landmarks or the state of affairs that must be brought about during the execution of a protocol in order to achieve its goal. Accordingly, the most important aspect of protocols is these landmarks rather...

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 n...

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...

Mediation services are becoming increasingly important in multiagent systems. An agent that can act on behalf of another agent is one important example of me- diation functionality commonly required. Within this paper, we define and ana- lyze PROXY and PROXY-WEAK communicative acts that formally specify se- mantics for interacting with middle agent...

By now, intelligent agents have been on the research agenda of the computer science community for roughly one decade. Still, control architectures for autonomous intelligent systems have been an important research issue for an even much longer time, going as far back as James Watt’s steam engine control mechanism based on mechanical feedback. More...

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, respe...

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...

Autonomous mobile robots need to integrate many different skills in order to perform complex tasks. In particular, they need to explore, sense, map and navigate in unknown or partially known environments. This paper describes a robot system that is designed to perform a find-and-deliver task in an office-building-like environment. The robot's initi...

Reducing the burden of interacting with complex systems has been a long standing goal of user interface design. In our approach to this problem, we have been developing user interfaces that allow users to interact with complex systems in a natural way and in high-level, task-related terms. These capabilities help users concentrate on making importa...

The University of Michigan Procedural Reasoning System (UM-PRS) is an object--oriented (C++) implementation of PRS as described in [1, 2, 3, 4, 6] and other papers. This document's primary purpose is to explain everything that is required to create an UM-PRS application. To a lesser extent, the document will familiarize readers to PRS's concepts. F...

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...

Plan recognition remains a largely unexplored paradigm for facilitating coordination. In this paper, we begin to explore domain, task, and agent charac- teristics which impact upon the utility of using plan recognition for coordinating multiple agents and, in particular, collections of agents organized into com- peting teams. Agents in our research...

Mobile robots operate in a wide variety of environments, and the tasks they are being designed to perform vary from the simplest pick-and-place factory jobs to space station construction, maintenance, and repair. Vision systems for mobile robots are used to help locate goal objects or locations, plan paths to the goal, avoid obstacles along the cho...

plan recognition as its primary means of acquiring the information necessary to coordinate the activities of agents. Preliminary research has demonstrated that the plan recognition system developed makes coordination of multiple agents possible. An important issue that arises when observation is the primary means of information acquisition is the i...

It is important, in situations where teams of agents can commit to joint, long-term objectives, that the agents be able to identify when the team objective is no longer important, or is futile. Prior work has typically assumed that agents should believe that joint commitments hold unless explicitly told otherwise by other agents. In this paper, we...

Plan recognition is the process of observing another agent's behavior(s) and inferring what, and possibly why, the agent is acting as it is. Plan recognition becomes a very important means of acquiring such information about other agents in situations and domains where explicit communication is either very costly, dangerous, or impossible. Performi...

The Procedural Reasoning System (PRS) is a general purpose reasoning system that is particularly suited for use in domains in which there are predetermined procedures for handling the situations that might arise. We have just completed an implementation of PRS written in C++, which we call the University of Michigan Procedural Reasoning System (UM-...

The University of Michigan Procedural Reasoning System (UM-PRS) is an object--oriented (C++) implementation of PRS as described in [1, 2, 3, 4, 6] and other papers. This document's primary purpose is to explain everything that is required to create an UM-PRS application. To a lesser extent, the document will familiarize readers to PRS's concepts. F...

To coordinate with other agents in its environment, an agent needs models of what the other agents are trying to do. When communication is impossible or expensive, this information must be acquired indirectly via plan recognition. Typical approaches to plan recognition start with a specification of the possible plans the other agents may be followi...

Researchers at the University of Michigan's AI Lab are working to produce mobile robot systems capa-ble of intelligently reacting to events and objects in a real world environment. A system has been built which integrates a reactive planner with perception and navigation capabilities in order to generate ble, responsive vehicle behavior. The system...

The issues that can arise in research associated with multiple, robotic agents are discussed. Two particular multi-robot projects are presented as examples. This paper was written in the hope that it might ease the transition from single to multiple robot research.

The Carmel project (computer-aided robotics for maintenance, emergency, and life support) which won the AAAI 1992 Robot Competition, is discussed. Carmel's design philosophy and architecture, obstacle avoidance, global path planning, vision sensing, landmark triangulation, and supervisory planning system are described. The Carmel project shows that...

This paper describes the design and implementation of an integrated system for combining obstacle avoidance, path planning, landmark detection and position triangulation. Such an integrated system allows the robot to move from place to place in an environment, avoiding obstacles and planning its way out of traps, while maintaining its position and...

This paper describes the design and implementation of an integrated system for combining obstacle avoidance, path planning, landmark detection and position triangulation. Such an integrated system allows the robot to move from place to place in an environment, avoiding obstacles and planning its way out of traps, while maintaining its position and...

In this paper we discuss the implementation and uses of the object recognition system used for CARMEL, the University of Michigan's winning entry in the AAAI-92 Autonomous Robot Competition. Following the rules of the competition, the robot was required to navigate within a large, unstructured environment performing exploration, and then a directed...

The University of Michigan's CARMEL and SRI International's FLAKEY were the first- and second-place 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 reaso...

Last summer, AAAI sponsored a mobile robot competition in conjunction with the AAAI-92 conference in San Jose, California. Ten robots from across the country competed in the competition, with CARMEL from the University of Michigan finishing first. CARMEL is a Cybermotion K2A mobile platform with a ring of 24 sonar sensors and a single black and whi...

Group communication is the core of societal interactions. Therefore, artificial agents should be able to communicate with groups as well as individuals. However, most contemporary agent communication languages, notably FIPA and KQML, have either no provision or no well-defined semantics for group communication. We give semantics for group communica...

Less than half of the individuals of working age with visual impairments are employed, and a significant barrier to em-ployment is effective computer access. Some of the screen reader applications offer some help but have limited con-text sensitivity and are of limited use in applications with dynamic "interfaces" like web pages. Sophisticated scre...

The Army's vision of the future for armored and mechanized military structure includes the use of mixed teams of human and robotic forces on a dynamic and rapidly changing battlefield. Successful implementation of this vision will require autonomous and semi-autonomous robotic forces and a command and control infrastructure that will allow human, r...

This paper discusses issues related to autonomy within BDI- based intelligent agent architectures and applications. One advantage that agent-based applications built upon formal theories such as BDI (Belief-Desire-Intention) theories have over standard application implementation paradigms is their explicit implementation of some of the mentalistic...

An agent operating in the real world must be able to coordinate its activities with those of other agents. Traditionally, work it/multiagent coordi-nation has assumed that agents can communicate about their intentions or that they are coordinated through the efforts of a third party. In many en-vironments, however, reliance on communication or on a...