The AADL Error Library: An Operationalized Taxonomy of System Errors

Abstract

The problem of how to best classify system errors has been a topic of research for years. In this paper, we present an established taxonomy that draws on a broad range of previous work in this area: the Architecture Analysis and Design Languages (AADL) EMV2 Error Library. The error library is now part of an international standard and has been used in a range of systems and domains. In this work, we describe its features, including that: a) it is deeply integrated in a rich, semi-formal system modeling language (AADL); b) the errors it includes have formalized semantics; and c) it is designed to be easily extensible by system developers to become domain or system-specific. We describe the original inspirations and prior work that informed the library's design, document the error families that comprise the taxonomy, and discuss the library's usage in an architecturally-integrated system assurance process.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 06, 2019
Accession Number
AD1068363

Entities

People

  • Peter Feiler
  • Sam Procter

Organizations

  • Carnegie Mellon University

Tags

Communities of Interest

  • Cyber

DTIC Thesaurus Topics

  • Chemical Industry
  • Classification
  • Engineering
  • Errors
  • Failure Mode And Effect Analysis
  • Hard Copy
  • Hazards
  • Hierarchies
  • Language
  • Massachusetts
  • Models
  • Safety
  • Safety Analysis
  • Software Development
  • Specifications
  • Standards
  • Taxonomy

Fields of Study

  • Computer science

Readers

  • Software Engineering.
  • Systems Analysis and Design