Managing Capability in Software Dependability Testing Through Generic Test Rigs

doi:10.1007/978-3-032-01241-8_8

Return

Managing Capability in Software Dependability Testing Through Generic Test Rigs

2026-01-01

PRE

Silke Kuball *

Lee Demetrius Walker

Martin Rendell

DOI：10.1007/978-3-032-01241-8_8

Original

Original request for help

Save

Abstract

En 中文

Operational-profile-based testing or statistical software testing (ST) has been a tool in determining quantitative indicators for on -demand failure probabilities of software-based devices for many decades. In the UK nuclear industry, the technique has found widespread application and been shown to be effective in supporting safety justification. In a time where a changeover of generations, changes in working styles and a pervasive skills shortage all work on and affect safety culture, new ways are needed to build up and maintain competency in this and other complex assurance techniques. In the case of ST, specific process steps in designing and generating valid ST test-cases show a high degree of repeatability between projects - especially for statistical testing of commercial-offthe-shelf computer-based devices. Encapsulating some of the expert input within this process into configurable components that can be reused, easily combined and configured mimics this repeatability. This has two advantages: firstly, a significant reduction of cost and effort in performing ST, secondly a significant improvement in how competency in the technique can be developed and maintained. A prototype for a generic and reuseable statistical test rig (SDITS - Smart Device Integrity Test Station) has been developed and is introduced in this paper. Initial results are presented and areas for future work and collaboration highlighted.

Keywords:

Software Dependability

Statistical Software Testing Competency

Test Rigs