Interactive Execution Monitoring of Agent Teams

  title={Interactive Execution Monitoring of Agent Teams},
  author={David E. Wilkins and Thomas J. Lee and Pauline M. Berry},
  journal={J. Artif. Intell. Res.},
There is an increasing need for automated support for humans monitoring the activity of distributed teams of cooperating agents, both human and machine. We characterize the domain-independent challenges posed by this problem, and describe how properties of domains influence the challenges and their solutions. We will concentrate on dynamic, data-rich domains where humans are ultimately responsible for team behavior. Thus, the automated aid should interactively support effective and timely… 

Figures from this paper

Execution Monitoring in Agent Teams
A monitoring framework for integrating many domain-specific and task-specific monitoring techniques and then using the concept of value of an alert to avoid operator overload is described, which is implemented in teams of autonomous ground and air vehicles.
Execution Monitoring and Replanning with Incremental and Collaborative Scheduling
We describe the Flight Manager Assistant (FMA), a prototype system, designed to support real-time management of airlift operations at the USAF Air Mobility Command (AMC). In current practice, AMC
A Temporal Logic-Based Planning and Execution Monitoring System
A general planning and execution monitoring system where formulas in an expressive temporal logic specify the desired behavior of a system and its environment and a unified domain description for planning and monitoring provides a solid shared declarative semantics permitting the monitoring of both global and operator-specific conditions.
Handling Coordination Failures in Large-Scale Multi-Agent Systems
A brief survey of the evolution of two key approaches to handling coordination failures in large-scale teams are provided: Restricting the number of agents that must be monitored, and using model-based rather than fault-based detection methods.
Model-Free Execution Monitoring in Behavior-Based Robotics
It is shown that pattern recognition techniques can be applied to realize model-free execution monitoring by classifying observed behavioral patterns into normal or faulty execution.
Detecting disagreements in large-scale multi-agent teams
  • G. Kaminka
  • Computer Science
    Autonomous Agents and Multi-Agent Systems
  • 2008
New bounds on the number of agents that must be monitored in a team to guarantee disagreement detection are presented, and YOYO, a highly scalable disagreement-detection algorithm which guarantees sound detection, is presented.
Task inference and distributed task management in the Centibots robotic system
A task inference algorithm is described that identifies potential team commitments that collectively balance constraints such as reachability, sensor coverage, and communication access and a dispatch algorithm for task distribution and management that assigns resources depending on either task density or replacement requirements stemming from failures or power shortages is described.
Monitoring Agents using Declarative Planning
In this paper we consider the following problem: Given a particular description of a multi-agent system (MAS), is it implemented properly? We assume that we are given (possibly incomplete)
Determining the value of information for collaborative multi-agent planning
A decision-theoretic algorithm for determining the value of information the system might acquire, query-reduction methods that decrease the number of queries the algorithm makes to the scheduler, and methods for ordering the queries to enable faster decision-making are provided.
Monitoring large-scale multi-agent systems using overhearing
This research attempts to provide a comprehensive theoretical model for overhearing and then, based on this model, to systematically cover various aspects related to overhearing.


Monitoring deployed agent teams
This paper presents a non-intrusive approach based on plan-recognition, in which the monitored agents' state is inferred from observations of their routine actions, and illustrates that monitoring based on observed routine communications enables significant monitoring accuracy, while not being intrusive.
TAIPE: tactical assistants for interaction planning and execution
This paper describes 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.
Rationale-Based Monitoring for Planning in Dynamic Environments
A framework for planning in dynamic environments is described, which introduces rationale-based monitors, which represent the features of the world state that are included in the plan rationale, i.e., the reasons for the planning decisions so far made.
Planning and reacting in uncertain and dynamic environments
The Cypress system is a domain-independent framework for defining persistent agents with this full range of behaviour, used for several demanding applications, including military operations, real-time tracking, and fault diagnosis.
Towards Focused Plan Monitoring: A Technique and an Application to Mobile Robots
  • C. McCarthy, M. Pollack
  • Computer Science
    Proceedings 1999 IEEE International Symposium on Computational Intelligence in Robotics and Automation. CIRA'99 (Cat. No.99EX375)
  • 1999
The original approach to rationale-based monitoring, originally introduced by Veloso, Pollack, and Cox (1998), is reviewed, how it can be adapted for a causal-link planner is described, and experimental results demonstrate that it can lead to improved plans without consuming excessive overhead.
Human directability of agents
A framework for the directability of agents, in which a human supervisor can define policies to influence agent activities at execution time, and is implemented within a PRS environment and applied to a multiagent intelligence-gathering domain.
I'm OK, you're OK, we're OK: experiments in distributed and centralized socially attentive monitoring
This work explores SAM, a novel complementary framework for social monitoring that utilizes knowledge of social relationships among agents in monitoring them, and presents a set of techniques for practical, efficient implementations with rigorously proven performance guarantees, and systematic empirical validation.
A Call for Knowledge-Based Planning
The terms knowledge-based and primitive-action planning are defined and an analogy from the current focus of the planning community on disjunctive planners to the experiences of the machine learning community over the past decade is drawn.
Remote Agent: To Boldly Go Where No AI System Has Gone Before
Determining the Loci of Anomalies Using Minimal Causal Models
The focus of this paper is a new technique for attention focusing that is used to detect both anomalous system parameters and "broken" causal dependencies in the system being monitored.