THE OPTIMAL BURN-IN TESTING OF REPAIRABLE EQUIPMENT.

Abstract

The infant mortality effect observed in the statistical treatment of reliability denotes a decreasing with time of the conditional probability of failure of a device which exhibits it. This widely present effect may be utilized to improve the reliability by means of burn-in testing which seeks to discriminate between high and low quality units by accumulating operating experience upon all units. Burn-in testing is applicable to both unrepairable and repairable devices. The latter case is more difficult since the state of a failed and repaired unit usually depends upon the entire past history of the unit and upon the nature of the repair process. A general burn-in test problem for repairable devices is formulated based upon the explicit modeling of the repair process. The case of unrepairable devices is treated as a special case. Considering a particular conjugate form of the repair rule the burn-in test optimization is formulated as a sequential decision problem and the solution is discussed in terms of dynamic programming. An adaptive reformulation which allows a parameter of the failure process to be considered unknown is also given. Specific models of decreasing failure rate processes, based upon the population heterogeneity cause of decreasing failure rate, are given. (Author)

Document Details

Document Type
Technical Report
Publication Date
Oct 01, 1966
Accession Number
AD0642105

Entities

People

  • John Michael Cozzolino

Organizations

  • Massachusetts Institute of Technology

Tags

Communities of Interest

  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Computer Programming
  • Dynamic Programming
  • Heterogeneity
  • Mathematical Programming
  • Mathematics
  • Optimization
  • Probability
  • Reliability

Fields of Study

  • Engineering

Readers

  • Electrical Engineering
  • Operations Research
  • Software Engineering