The headline in The Daily Mail says, "... desperate pilot landing at Austin airport tells passenger jet below it to abort takeoff because they're using the SAME runway." This is a great headline because it truly was only the Fedex crew that prevented a disaster and not the ATC system that is supposed to prevent collisions like between the landing fedEx jet and the departing Southwest jet.
Usually when these events break through to the public's awareness, an FAA spokesperson appears and says, Safety was Never Compromised. It's a cliche. This was a total system failure. These airplanes were not separated by any good fortune of serendipitious timing. The only thing preventing another Tenerife was the FedEx crew's situational awareness and the breath of god.
American aviation is a system with different parts and priorities, checks and balances, and really quite a bit of public transparency. This system, like all systems, can be studied and improved. The people who study the American ATC system have been shouting for at least twenty years that the next major airplane disaster will look like a particular scenario. This is going to be the Next Big Thing.
It looks like this: Two big jets. One of them is supposed to use a runway for takeoff or landing. The other plane will either cross or use the runway, and they collide. This is the nightmare scenario of American aviation, this is what the safety analysts tell anybody who will listen, this is the thing that keeps people awake at night.
Technology progressed, the internet arrived, and now if a Tower has an event over the weekend, guaranteed on Monday morning the phone rings and Headquarters says, "Hey I'm sure you're already looking at Saturday's event with United 123, call me when you get a solid handle on it".
This Austin situation is awful. As bad as it gets without body bags. The phrase, "pink mist" which was popularized in a 1999 ATC movie to refer to the clouds of airborne body fluids is not misplaced.
Aviation these days is quite public. Various Flight Tracker websites offer the public a view of the airplanes. There are ATC radio fans who put receivers on their houses and stream the audio online at sites like LiveATC.net. The airplane transponders, which used to send position information to the radar site, now transmit really detailed info into the public realm.
So the re-enactions, the tapes, the transcripts that we see - they're all ersatz wannabees, using readily available hobbyist info of unofficial provenance to paint the picture for the public, before the government agencies are anywhere close to making a public disclosure. This is a good thing which keeps good people honest.
As in any breaking news event, we tend to focus more on what the media shows us - oooh, bright shiny object - than to what's missing. The Austin airport does not have ASDE / AMASS gear which would have rang an alarm about the occupied runway. Even though they have 767s and 737s, Congress did not see fit to authorize funds for the Austin AMASS. I bet they will now.
This brings us to the cost-justification of saving lives. JFK deserved AMASS. Austin didn't. You may have heard of a Vision Zero philosophy in eliminating vehicle-driven deaths; proponents argue that there are no cost-justified levels of acceptable death. They say, it can never be ethically acceptable that people are killed or seriously injured when moving within the transport system. There is an Austin Vision Zero program.
The Tower controller cleared Southwest for takeoff when the inbound was three miles out (roughly). Usually, in nice weather, this could work. If the controller had said, "cleared for takeoff no delay traffic two mile final", that would have been even better.
I think the buried clue is the very low visibility. You hear the Tower controller reciting RVR numbers, "runway visibility range" touchdown 1400 (feet), midfield 600 (feet), rollout 1800 (feet). That's not much at all. 600 feet is the minimum requirement for planes with special equipment and crews with special training. This is the sort of sensitive operation that invokes the concerns about 5G phones interfering with radar altimeters.
With an 600-foot midpoint RVR, the departing captain taxies carefully out to the centerline, makes a full slow ninety-degree turn, and really really lines up. Takes a look out the window just to check for deer or vehicles, and then gradually applies takeoff power. It's a completely different takeoff experience, and it takes a lot more time. An experienced tower controller would know that.
Armed with the same misinformation that you have, I offer these thoughts:
- Was the controller a low-experience tower controller?
- Has the controller ever done same runway, arrivals and departures in very low visibility before?
- Has the controller ever sat in a jumpseat or simulator to see the different performance in different conditions?
- Was the Cab Cordinator or Tower Supervisor position staffed?
ATC is a system, not an individual feat. So lets look at system effects:
- The event happens at 6:47 am. Was the tower team all present? Was this a trainee and a distracted instructor?
- Why were they arriving and departing on the same runway when parallels were available?
- Did anybody brief the tower controller on what to expect and what to watch out for?
- Were positions combined? Were controllers working more than one job? Were they fully staffed?
There was a similar event at JFK a few weeks ago, between Delta and American. One was taking off, the other crossed the runway downfield. We hear that the ASDE / AMASS worked well. The tower controller urged the departure to cancel takeoff clearance. They stopped before hitting the crossing jet. Disaster was averted.
Let me say this: If these two events in two weeks are a trend, it's going to be a terrible year. Pete Buttigeig, who is notionally in charge and responsible, needs to get ahold of this.
And finally, we're in the media awareness timeframe for the police killing of Tyre Nichols in Memphis. Media awareness begins at the killing and ends at the funeral. Can anybody doubt that police killings are a safety issue for black Americans? Is there pressure for results, courtesy quality assurance, political influence, overtime constraints, and a code of Omerta? Of course. There is no functional civilian oversight, and people fly Blue Lives Matter flags.
Police departments are failed public safety systems. People will say, But not all cops are racists. And of course that's true. Put any group of people together and they're not a monolith.
The phrase, Not all cops.... recognizes that in fact, some cops are bad apples, prone to beating and biased against black people, and they walk around with authority and guns. And the good apples stand around and let it happen.
Would we tolerate an ATC system where "not all controllers" are bad apples, indifferent to safety? Where the good apples get busy with their coffee and scones while the rookie puts two planes together? I don't think so. But these cops are going to kill more Americans than the airport disasters.
Why do we tolerate a police system - ostensibly a public safety system - that kills more Americans than aviation does, with some cops walking around indifferent to safety? And yet we're petrified about two airplanes getting too close.
Couldn't be that the cop's victims and the passengers are from different socio-economic groups, could it?
Posts
8 comments:
Very good article. I would like you to comment on 2 thins that I think are critical for this systemic issue:
1- The mistake was not only to clear the Southwest to take off with the FedEx on a tight 3-miles final, with little timing. That may have been very poor judgement. But I think that allowing a vehicle (including an airplane) into the ILS protected area in such low visibility conditions with a plane conducting a CAT-III ILS/AUTOLAND approach inside the Final Approach Fix is a black-and-white "thou shalt not do" thing
2- Questioning the practice of "in-advance" landing clearance, where a plane can be cleared to land on the expectation that the runway will be clear, even if ATC plans to clear several other folks to take off, land, or cross the runway before the cleared-to-land guy lands, is not even an afterthought in the USA. However, this practice is FORBIDEN in most of the world. In most of the world, an airplane can be cleared for the approach, and to continue the approach, but CANNOT be cleared to land until a) the runway is clear, b) nobody else is cleared to use the runway, c) nobody else will be cleared to use the runway until after the landing. The difference is that in the US style, it requires ATC to cancel the landing clearance and instruct a go-around if things don't go as planed, and that requires not only the tower not to make mistakes but also to be aware that things are not going as planned, which in low visibility conditions can be impossible. However, in the "rest-of-the-world" style, the plane is NEVER cleared to land until it is, and then the runway is theirs and only theirs. If they don't receive the clearance to land in time the pilots MUST go around because they are not cleared to land. In the Austin example, using the rest-of-the-world practice, even with the tower making the mistake of clearing the Southwest for take-off when they did, they could have not clear the FedEx to land until the tower saw in their radar screens (because they didn't have visual contact due to poor visibility) that the Southwest is airborne, and the FedEx, not being cleared to land, would have had to go around by standard operating procedure, not by exceptional situational awareness and brilliant judgement of the pilot as they did. In other words, with things as they are in the US, FedEx would have been perfectly legal to lad on top of Southwest. With the procedures used in te rest of the world, it would have not. And do you know why the US takes this additional risk? o you know what we get in exchange? What is the benefit in terms of costs and fluidity of air traffic? Well if you do let me know, because in may assessment there is none. In other word, it is a totally unnecessary risk to assume.
Another example of this risk was Delta 1086, and MD-80 that lost directional control on the roll out landing in La Guardia and went off the side of the runway. During the accident (where nobody resulted injured) the electronic equipment of the plane were damaged and the pilots could not advise the tower of what had happened. The tower could not see the runway due to low visibility. Another plane had already been cleared to land on the same runway (since before the Delta landed). By sheer luck, an airport vehicle was near the place where the accident and advised the tower. Only then the tower advised the other landing plane to go around. If it wasn't for the luck that the airport vehicle was there and had the presence of mind to advise the tower, the second plane would have landed with the Delta stuck next to the runway edge. However, with the procedures used in the rest of the world, the second landing plane could be cleared to continue the approach but could not be cleared to land until the Delta confirmed that they were clear of the runway. So even if the vehicle didn't see the accident and the tower didn't tell the second plane to go around, they would have still gone around.
Gabriel, thank you for your well-informed and written comment. To be brief: Yes, and Yes.
First, the Tower at AUSTIN erred in putting the departure into the ILS Critical area, if that's what happened. Second, there's a technique (not a requirement) that some employ in low-viz conditions that only one airplane is ever cleared to use the runway. Clear the departure for takeoff, when they get airborne then you clear the arrival to land. To old school types, the runway clearance really means something, it's the controller's bond that the runway now belongs to the airplane, and issuing half-dozen landing clearances diminishes that value.
Especially with a mix o arrivals and departures, having only one airplane cleared at a time is a great technique. Not a requirement. Thank you.
Just a thought … I’d wager that there is no issue in America that has been talked about more than police brutality, over the past 10 years. And we have seen vast improvements in that time.
It is not that we tolerate it, but that it is a hard problem to solve. When we find out we have a bad ATC, they are gone. Same with cops, these days. And surely we can do more, but your words imply that we could just solve the problem by, what, caring more?
"Couldn't be that the cop's victims and the passengers are from different socio-economic groups, could it?" I don't understand that statement at all. Lots of different socio-economic groups fly, if you have been on a flight anytime lately. Nevertheless, if you only watch mainstream media, you might be left with the impression that police killings of unarmed blacks is the only issue. In reality, police killings of unarmed people not committing a crime of any race is very, very tiny, only a handful in the past few years. Police killings of armed people most often happens during or after commission of a crime, and racially is in proportion to the criminal (jail) population. Whereas, hundreds of police of all races, but mostly white, died at the hands of criminals in the past few years. But, narrative.
Reminds me of the near catastrophic miss at SFO in 2017:
https://simpleflying.com/air-canada-flight-759/
Very interested to see if GBAS for controlling both movement & coordinating landings become a thing.
One thing about "a few bad apples" is the rest of the statement: "...spoil the whole barrel"
Oh, and the error: a dirty little secret in ATC right now is that they're terribly understaffed and tired at a lot of facilities.
During the initial pandemic, to be blunt, they had it easy; flight schedules were slashed, and controllers at most facilities were working a super-cushy schedule designed to limit their mixing (and thus limit breakouts of the disease): 5 shifts in a row on, 10 shifts in a row off.
If a crew had a breakout, then they'd shift things around, and sometimes controllers would wind up working 10 on and 5 off for a while, but mostly they did 5-on-10-off for quite a while.
Then... traffic returned, but the issues around Covid exposure and cases continues. Plus the hiring pipeline was shut off for basically two years, and they were already behind as the wave of controllers hired in post-strike have all punched out (like me!)
So now nearly everyone at my old facility is working 3-4 shifts of OT every month.
Now, do we know this was a factor in Austin? Nope, not at all. But would it be surprising to learn? Nope, not at all.
I appreciate your related comment about policing as public safety, bad apples and all.
Both the two near-collisions on active runways, and the rolling disaster of police frequently shooting suspects (well over 1,000 fatal cop shootings per year in the US. No other nation has our numbers, or rate per population) are signs of systemic failure.
I think one of the reasons police shootings don't rise to the attention of plane crashes is scale. We tolerate 10s of thousands of auto-related deaths per year (including a rising, horrifying trend of people running over kids with their own massive SUVs because they can't see down past the enormous, high hood & grill) because the deaths look like a localized trickle. I see the WI programmable info signs that say "237 deaths so far. Slow down!" and it doesn't have the emotional impact of 237 people perishing in a moment in a fully packed airliner.
I'm not sure what it is in our nature - or in how we're normalized to accept it. But it impacts our muted response to police overuse of a gun as a first resort, too.
Post a Comment
Comments and Feedback? Love that stuff. Please leave your thoughts in the box below--