ASFIT: AUTOSAR-based software fault injection test for vehicles

Jihyun Park, Byoungju Choi

Research output: Contribution to journalArticlepeer-review

2 Scopus citations

Abstract

With recent increases in the amount of software installed in vehicles, the probability of automotive software faults that lead to accidents has also increased. Because automotive software faults can lead to serious accidents or even mortalities, vehicle software design and testing must consider safety a top priority. ISO 26262 recommends fault injection testing as a measure to verify the functional safety of vehicles. However, the standard does not clearly specify when and where faults should be injected, and the tools to support fault injection testing for automotive software are also insufficient. In the present study, we define faults that may occur in Automotive Open System Architecture (AUTOSAR)-based automotive software and propose a fault injection method to be applied during the software development process. The proposed method can inject different types of faults that may occur in AUTOSAR-based automotive software, such as access, asymmetric, and timing errors, while minimizing performance degradation due to fault injection, and without using any separate hardware devices. The superior performance of the proposed method is demonstrated through empirical studies applied to fault injection testing of a range of vehicle electronic control unit software.

Original languageEnglish
Article number0850
JournalElectronics (Switzerland)
Volume9
Issue number5
DOIs
StatePublished - May 2020

Keywords

  • AUTOSAR
  • Fault injection automation
  • Software fault injection test

Fingerprint

Dive into the research topics of 'ASFIT: AUTOSAR-based software fault injection test for vehicles'. Together they form a unique fingerprint.

Cite this