However, many of the posts here refer to "un-alerted see and avoid", and nobody has seriously advocated "un-alerted" see and avoid, quite the contrary, as I have pointed out in writing, on many occasions.
Do you support Class E over Class D at Broome and Karratha?
- How do IFR and VFR receive ‘alerting’ in non-radar Class E?
- How do ATC [in a non-radar environment] confirm aircraft departing Class D [in to Class E] have the TXPDR selected ON for operation in Class E?
To briefly answer your ‘assertion/s’. Also from the O’Neill report:-
The NAS 2b safety case conducted prior to the implementation of these changes seems to have consisted primarily of subjecting those aspects of the 2b changes that were not consistent with the system in use in the United States, to examination by a panel of experts who discussed their likely safety acceptability.
Remember, this was a ‘design’ safety case, not the volume specific safety cases that were being asked for by industry that would have highlighted the deficiencies.
And, the nub of this discussion:-
There is no evidence that there was any attempt to subject proposed changes to formal quantitative analysis.
On the one hand, you support the introduction of NAS2b without ANY quantitative analysis, and on the other, you whine about AsA’s safety work to return system safety [previously utilised and demonstrated safety of Class C above D] as promptly as safely possible following the unfolding of the NAS2b Class E inadequacies? I find that to be the height of hypocrisy.
CASA provided input to the safety case
The ‘general design safety case’ prior to NAS2b
most notably in a minute of 12 September 2003. Extracts from that document stated:
• “CASA believes the US FAA NAS to be safer than the existing Australian airspace management system based on a simple comparative analysis of mid air collision statistics between the two countries”
• “The detailed evaluation of the presented design safety case evidences that the Stage 2b implementation will bring about an increase in risk beyond that which exists in Australia today.”
• “If the NASIG claim that it is necessary to the full implementation of NAS that Australia transition through this higher risk stage is valid, then clearly
the sooner NAS Stage Four is implemented and the potential safety benefits realised, the sooner this risk is removed”.
Shall we further the discussion and have a look at the merits [or otherwise] of the NAS stage 4 changes
Leadsled?
Back to the ‘rollback’
One issue that should be addressed is why more empirically based data modeling procedures were not examined in more detail. Thus an examination could have been made of the relationship between traffic flows and types of aircraft in various types of airspace and resultant incidents both in Australia and internationally. While it is by no means clear that this approach would definitely have been successful in producing highly accurate estimates, given the well-known difficulties associated with pure modeling approaches, a significant attempt to explore this option would seem to have been warranted.
It seems Professor O’Neill agrees with concerns raised, then, and most notably today!
You said:
anybody who believes a 1 in 1 to a 2 in 1 error rate for pilots executing ATC instructions is a reasonable error rate must have two willies, you can't get that stupid playing with one.
Who is saying that?
Just as much as ascribing an error rate of (by remarkable contrast) 1 in 1,000,000 to ATC persons.
Table 2 [page 28] of the O’Neill report
Error estimates for 2 VFR scenarios in E and C airspaces under Radar 45 Conditions
Generally the probabilities appear to show convincingly that VFR pilots are much less likely to make errors in Class C airspace.
A reasonable conclusion given the requirement for a ‘clearance’, and third party ‘error catch’ as part of those interactions in Class C.
In fact the table seems to show that VFR pilot errors are in some instances almost 85,000 times more likely in Class E airspace, a difference that is likely to have had very significant effects on the estimates of the relative safety of the two classes of airspace.
The problem with this conclusion is that the probabilities in the shaded cells, although entered into the fault tree, were not in fact estimated by the panel but directly assigned by AA staff. In other words, in these cases AA merely decided that these probabilities were remote and arbitrarily assigned them a value of 1 in 1,000,000.
Are you agreeing that the expert panel estimate of 85,000 times greater chance of VFR error in Class E was reflective of reality? Or are you questioning the expert [pilot and ATC] panellist responses?
Perhaps the data capture of that particular question was displayed in a potentially confusing way which resulted in an anomalous scoring by expert panellist?
You also say:
To further suggest that such figures do not materially alter the outcome of the analysis is (at best) revealing the writer has no in-depth knowledge of the various qualitative and quantitative methods of risk assessment.
From the O'Neill report:
AA expressed confidence that although the absolute values of the risk estimates may not be reliable, the relative results (i.e. between airspace types) were.
Clearly, this is reasonable, and correct. Were it [and more generally the AsA case for rollback] not, it would not have occured given the 'political' pressure cooker of the time!
That same political pressure cooker might well be on the heat today
