Reactive Execution in a Command, Planning and Control Environment.
Abstract
This research proposed a design of a reactive execution agent (REA) which accepts and executes task directives in a dynamic environment. This design drew upon an amalgam of ideas emerging from situated rational agency research. The result was an agent which was able to accept tasks directives (or reject them), reason about directives to determine how to achieve them (maintaining commitments to achieving other directives), respond to execution failures, and communicate with a superior agent when further deliberation is required beyond the abilities of the REA. The primary contributions of the research have been (1) a characterised model of rational behaviour, (2) an inter-agent communication language (IACL) for adapting execution time behaviour, and (3) the use of causal structure to predict potential execution failures. The research has shown how reasoning about a commitment to tasks, failure related to tasks, and tasks with temporal; extent interact with the basic set of capabilities to provide a robust model for competent and rational situated behaviour. (AN)
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 18, 1995
- Accession Number
- ADA299381
Entities
People
- Glen A. Reece
Organizations
- University of Edinburgh