This is the guidance document that helps interpret the Part 1 standard provided separately on this site.
Annex D is of some interest as it provides a table that essentially ranks the Part 1 requirements based on the safety integrity level needed for the application. This could help medical device manufacturers in planning and justifying different levels of rigor and documentation for different potential software impact/hazards levels.
Annex D also states:
“Much of the main text of the standard contains requirements that are considered to be ‘good standard practice’ for all SRS of all safety integrity levels and hence, for these requirements, the table shows no variation with integrity level. In other cases an activity may be required for all safety integrity levels, but the extent of the activity or the amount of evidence may be reduced with reducing integrity levels. The table is intended to provide developers with flexibility for the development of software of lower integrity levels and permits them to justify the development approach taken.”
This statement is consistent with FDA’s statements that rigor of software validation can vary based on potential impact/hazards. This statement can help convince others that complete consistency in rigor and detail across all software applications is not a requirement for safety and scaling to potential impact is acceptable. Of course, the details of what is sufficient for each application needs some justification and in the case of the UK Defence standard an approach to this justification is through use of the Annex D table.
Annex E presents the concept of “Process Risk Analysis” together with FMEA and FTA methods. The approach is to identify the parts of the software development process that could affect safety and to determine approaches to each part of the process to prevent such safety problems.
Annex E Section E.3.5.2 also discusses safety cases and safety arguments which can be a useful tool in focusing on critical failures and justifying/demonstrating coverage and effectiveness of mitigations/methods of control.
SoftwareCPR keywords: white box, traceability, United Kingdom, military.