PPRuNe Forums - View Single Post - How do we avoid fatalities?
View Single Post
Old 11th Jul 2004, 10:48
  #9 (permalink)  
bookworm
 
Join Date: Aug 2000
Location: UK
Posts: 3,648
Likes: 0
Received 1 Like on 1 Post
All the fancy gizmos would be buzzing away and of little use in this hot spot of aerial activity and yet? .... people seem to think they are this wonderful thing that will cover their aircraft in cotton wool. Redsnail hit the nail on the head "good situational awareness and a good lookout" is the most important factor in avoiding a mid-air.
It sounds very noble, man over machine and all that, but has very little basis in fact.

Collision avoidance strategy depends on probability of detection of a conflict in time to do something about it. Take two extremes: you manage to avoid collisions while walking along a pavement in a shopping street reasonably effectively by see-and-avoid. But if I stuck you at the target end of a rifle range and said "keep a good lookout for bullets and avoid them", how effective do you think that see-and-avoid would be?

There seems to be an implict assumption that the human pilot is capable of detecting potential conflicts with sufficient time before collision with a useful level of reliability and all we have to do is be better at it, spend 90% of our cockpit worload resource doing it and all the problems will go away. There's precious little scientific evidence to support that. The anecdotal evidence seems to revolve around the potential conflicts that the pilot did manage to spot, since he has no knowledge of the ones he didn't. And there's a strong argument that the ones that are going to hit you are distinctly harder to see than the ones that will miss, because the former have no azimuthal movement -- they just start as dots and get bigger.

The difference between the bullets and the pedestrians is speed (and angular size, though if I said "150 mm shells" you wouldn't do much better). Speed is a factor in aviation collision avoidance too, which is why we're talking about this in the Private Flying forum. No one has entertained the rather quaint notion that commerical transport aircraft should just see and avoid each other without the help of TCAS for a very long time.

How fast do you have to go before see and avoid becomes ineffective? It seems the recent collision may provide a solitary data point that suggests that it's rather lower speed than many contributors to these threads would like to think.

I'm very content with the notion that the probability of collision enroute is low enough to be an acceptable risk, and that it does not require universal equippage with artifical aids. What frustrates me is the flawed logic of those who revert to the old behaviour of beating the pilot about the head and saying "must try harder next time", and pretend that that's going to fix the problem.
bookworm is offline