[SystemSafety] New IET report

[Peter Bernard Ladkin]

The IET Transport Policy Panel 
https://www.theiet.org/impact-society/thought-leadership/expert-panels/transport-policy-panel/  has 
published a new report on automation and safety in air, rail, water and road transport.

https://www.theiet.org/media/8502/advancing-safety-in-transport-through-automation.pdf

Among its 6 recommendations are
(3) establish a dedicated road accident investigation organisation;
(4) develop a new standard for functional safety of programmable safety-related systems; and
(5) review cybersecurity standards for adequacy.

There is one international standard for functional safety of programmable safety-related systems 
(except for medical systems), IEC 61508. There are two IEC 61508 Maintenance Teams, one for general 
and HW and the second for SW, have some 90 members each, working on the next edition. The SW Team, 
for example, has been meeting regularly since December 2014 (the HW/general Team "only" since 
November 2017). Current projection for a CD is Autumn-Winter 2021, and then there are two years to 
get to a final version, so it's going to be well over 8 years and maybe 9.

The IET Policy Panel says we need a new one (Recommendation 4). We can endorse that in theory (there 
is much in IEC 61508 not to like, and it won't disappear). But in practice I see almost no chance 
that a volunteer effort will come up with an alternative. The IEC has also just put out a draft of 
5479, the AI-and-functional-safety tech report. Quite a few people have put considerable effort into 
that.

I think we can wholheartedly endorse Recommendation 5 and also guess what the result might be. The 
IET/NSCS Cybersecurity-for-Safety Guidelines are out. I am going to be comparing them with IEC 63069 
(which I criticised in SSS'20) for SSS'22. The IET/NSCS guidelines are way more appropriate, if only 
because the IET committee went to some lengths to incorporate advice from people with a technical 
background in practical cybersecurity (the NSCS). IEC 63069 is nominally only for ICS, but there is 
an effort to make it into a "horizontal" guideline, indeed into a standard rather than a technical 
report.

I also don't see civil aviation joining in with any such effort. They have their own rules, 
regulations and de facto standards-setting bodies (SAE, RTCA, EUROCAE). Also rail goes mainly 
through the European Agency for Railways, ERA, which works with CENELEC on the EN 5012x series for 
functional safety of programmable rail systems. As far as I know, the rail people are fully involved 
with that series, in the sense of all the safety-knowledgable people I know in rail automation. I 
don't see that any alternative effort would find such people to draft it.

The Panel are right to point out that rail, air and maritime have their own government-supported 
accident investigation processes; that road doesn't have such an entity in the UK, but it might well 
need one with increasing autonomy. The US of course has the NHTSA, which is and has been looking 
hard at the accidents possibly involving the Tesla Autopilot. The UK has the Office of Rail and 
Road, but it has not to date undertaken accident investigations in the way the NHTSA/NTSB has in the 
US.

PBL

Prof. Peter Bernard Ladkin, Bielefeld, Germany
ClaireTheWhiteRabbit RIP
Tel+msg +49 (0)521 880 7319  www.rvs-bi.de





-------------- next part --------------
A non-text attachment was scrubbed...
Name: OpenPGP_signature
Type: application/pgp-signature
Size: 840 bytes
Desc: OpenPGP digital signature
URL: <https://lists.techfak.uni-bielefeld.de/pipermail/systemsafety/attachments/20210805/45d3c529/attachment.sig>