[SystemSafety] NYTimes: The Next Accident Awaits

[Peter Bernard Ladkin]

> On 3 Feb 2014, at 21:52, Andrew Rae <andrew.rae at york.ac.uk> wrote:
> 
> Peter,
> The simple answer is that there is no particular reason to believe that they don't. The targets are set based on how safe aircraft were and the techniques tell them that new aircraft will be about as safe.

We therefore conclude, please, that confirmation bias is not a problem.

> There's no published evidence that shows that civilian aerospace safety analysis is capable of distinguishing safe from unsafe design. 

There's no published evidence that shows that any engineering discipline is capable of distinguishing safe from unsafe design. Indeed, IEC 61508 as well as plenty of books such as Hal Lewis's exemplary Technological Risk suggest that determining what is acceptably safe, and what unacceptable, is essentially a social, or sociological, task. 

Indeed, this is implicitly acknowledged by the German MGS and French GAMAB criteria for acceptability of new designs. Which is that a new design will be shown to be

> about as safe

or better, than what is currently used.

PBL

Prof. Peter Bernard Ladkin, University of Bielefeld and Causalis Limited