Question regarding Integrity Analysis

Forum for discussing and asking questions about the MISRA "Guidelines for safety analysis of vehicle based programmable systems"

Moderator: david ward

Post Reply
david ward
Moderator
Moderator
Posts: 489
Joined: Wed May 05, 2004 9:05 am
Location: MIRA

Question regarding Integrity Analysis

Post by david ward » Tue Mar 04, 2008 12:42 pm

The following question has been received by MISRA and is posted on behalf of the questioner:
Both the MISRA Integrity Report and the Safety Analysis Guidelines refer to the following probabilities:

Reasonably Possible
Unlikely
Remote
Very Remote
Extremely Improbable

However there is never an explicit definition of the probablities associated with these categories and the categories do not directly map onto the probabilities defined in Table A.2 of the integrity report. Are there explicit definitions for these categories or can I assume the following?

Probable = Reasonably Possible
Occasional = Unlikely
Remote = Remote
Improbable = Very Remote
Implausible = Extremely Improbable
Dr David Ward
MISRA Project Manager

MISRA Reply
Posts: 52
Joined: Mon Dec 06, 2004 12:24 pm

MISRA reply to this question

Post by MISRA Reply » Tue Mar 04, 2008 12:46 pm

It should be noted that each of the MISRA Reports (including the \"Integrity\" report to which you refer) relate to the original work that was done during the early stages of the original MISRA project, and that sometimes they describe possible multiple approaches, one of which was then chosen by the MISRA Consortium to become the official recommended approach. This is what happened in the \"Integrity\" report (note that Section 5.2 says \"This section describes the three approaches that have been considered for the MISRA Guidelines\").

Thus Appendix A and Appendix B of the 1995 \"Integrity\" report describe two separate and distinct methods and hence, for example, the contents of Table A.2 and Table B.1 are not compatible.

The MISRA consortium selected the method described in Appendix B, though this has been enhanced a number of times since then. The probabilities mentioned in Table B.1 (and used in Table 1 of the Development Guidelines for Vehicle Based Software) have not been given any numeric values and this is deliberate. It was always intended that if a company needed to have numeric values then it was its responsibility to provide them.

This decision was reviewed during the preparation of the new Guidelines for Safety Analysis of Vehicle Based Programmable Systems and, whilst it was decided to keep the words used in Table 1 of the original Guidelines in Section 4.5 of the new Guidelines, possible numeric values are now also offered in Appendix E. However, not all companies will wish to use them.

Post Reply

Return to “MISRA SA discussions”