Management of Temporal Constraints for Factory Scheduling.

Abstract

This paper, presents constraint propagation techniques used in the OPIS scheduling system to update schedule descriptions and detect introduced inconsistencies. This approach is summarized as follows: 1) a hierarchical model is used to represent resources and operations to be performed. Schedules are developed and maintained at different levels of precision which are explicitly associated with resources and operations; constraint propagation is correspondingly performed at different levels, 2) Various scheduling constraints are attached to resources and operations, and combined to derive time bound constraints. A description of the original constraints that collectively impose a bound is explicitly recorded, 3) Time bound constraints are maintained by an object-oriented propagation process: through messages, resources and operations communicate constraints and cooperate to compute time bounds, and 4) When time bounds are inconsistent, their origins provide the information required to construct an appropriate description of the conflicting situation. This description provides OPIS with information needed to make reactive decisions.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jun 01, 1987
Accession Number
ADA184337

Entities

People

  • Claude Le Pape
  • Stephen F. Smith

Organizations

  • Carnegie Mellon University

Tags

Communities of Interest

  • Autonomy
  • C4I
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Artificial Intelligence
  • Cognitive Science
  • Consistency
  • Expert Systems
  • Hierarchies
  • Inequalities
  • Information Systems
  • Intelligent Systems
  • Intervals
  • Job Shop Scheduling
  • Knowledge Based Systems
  • Manufacturing
  • Natural Languages
  • Precision
  • Production
  • Scheduling (Production)
  • Time Intervals

Fields of Study

  • Computer science

Readers

  • Joint Military Operations and Doctrine.
  • Software Engineering.
  • Statistical inference.