Title: Hazard analysis of Unified Modelling Language sequence and state charts using software fault tree analysis

Authors: Pankaj Vyas; R.K. Mittal

Addresses: Department of Computer Science and Information System, Birla Institute of Technology and Science, Pilani-333031, Rajasthan, India ' BITS Pilani, Dubai Campus, P.O. Box 345055, D.I.A.C, Dubai, UAE

Abstract: Software fault tree analysis (SFTA) is a well known software safety analysis technique for identifying those software related erroneous events which can force the system into a hazardous state. Over the years, the effective strengths of SFTA approach have been demonstrated by applying it practically at every phase of software development such as coding, design and requirement analysis. The focus of SFTA application research efforts at the software design phase is the automatic or semi-automatic construction of a software fault tree for the selected hazardous state of the system directly from the given Unified Modelling Language (UML) models such as sequence and state charts. But programmed application of SFTA has been successfully carried out so far only in the coding phase whereas at other phases especially at design and requirement analysis its application remains a manual process so far. This paper presents a semi-automated approach to construct a software fault tree from the UML sequence and state charts drawn for a particular functional scenario. The effective strength(s) of the proposed approach have been demonstrated by applying it on two use-case scenarios selected from an elevator controller application.

Keywords: software safety analysis; hazard analysis; software FTA; fault tree analysis; SFTA; Unified Modelling Language; UML; sequence diagrams; state diagrams; state charts; elevator control; lift control.

DOI: 10.1504/IJCCBS.2013.056512

International Journal of Critical Computer-Based Systems, 2013 Vol.4 No.2, pp.173 - 197

Received: 15 Dec 2012
Accepted: 22 Jul 2013

Published online: 29 Apr 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article