# Can a mistake of a single air-traffic controller cause a collision between planes?

I am watching season 2 episode 13 of Breaking Bad and in the episode it shows that:

an air traffic controller, distraught from his daughter's death, causes a mid-air collision, resulting in debris and human remains raining down onto the Whites' residence, as well as the rest of Albuquerque

It is at the very end of the episode and it looks like he has done nothing malicious just forgot to tell something. To me this sounds really surprising that a single human error can cause such damage.

As a programmer, I know that most of the scenes about computers depicted in the movies are complete nonsense. Is this a similar nonsense about aviation or is this really possible?

-
Looks like yes --- youtube.com/watch?v=AZIf_ivdMes –  Tasos Jul 7 '14 at 5:03
and while on the ground too -- youtube.com/watch?v=LepaN0QI-Bs –  Tasos Jul 7 '14 at 5:07
I expect you will like the book Fate is the Hunter –  radarbob Jul 30 '14 at 3:19

I haven't seen the Breaking Bad episode, but in my experience any aviation related scenes in films and series are usually stretching reality beyond the breaking point.

But it is possible that a single mistake can cause an accident. The probability of that happening is extremely remote, though.

Typically there are a number of safety barriers (also called defensive layers) in the system(1) that prevent any single mistake from developing into a collision. However, every barrier has a small probability of not catching the problem (e.g. due to it not being perfect or being in state of maintenance / failure) so eventually a crash could result from a single mistake.

This is often called the Swiss Cheese Model (originally from Reason, J. (1990) Human Error. Cambridge: University Press, Cambridge.)

When an accident happens, sometimes the phrase "all the holes in the Swiss cheese lined up" is heard, and you now know why.

While the first hole might be the mistake of a controller, the other holes may not be mistakes but are certainly rare circumstances and flaws in the system. Accident investigation focuses on identifying these contributing factors.

In the Überlingen accident mentioned by Portree Kid, two immediate causes have been identified(2):

1. The imminent separation infringement was not noticed by ATC in time. The instruction for the TU154M to descend was given at a time when the prescribed separation to the B757-200 could not be ensured any more.
2. The TU154M crew followed the ATC instruction to descend and continued to do so even after TCAS advised them to climb. This manoeuvre was performed contrary to the generated TCAS RA.

The first cause can be seen as a mistake by the controller, however that mistake could only happen because of a number of contributing factors (systemic causes):

• Management and quality assurance of the air navigation service company did not ensure that during the night all open workstations were continuously staffed by controllers.
• Management and quality assurance of the air navigation service company tolerated for years that during times of low traffic flow at night only one controller worked and the other one retired to rest.

In addition to the above, the Short Term Conflict Alerting (STCA) system did not display conflicts to the air traffic controller because of maintenance (not listed as a cause in the report, but discussed in the text).

The second cause is not a mistake by the crew of the TU154M, but a latent failure in the way ACAS/TCAS II was introduced. Never did anyone give guidance on what to do if the controller and the TCAS gave conflicting instructions.

In the report this is identified as a systemic cause:

• The integration of ACAS/TCAS II into the system aviation was insufficient and did not correspond in all points with the system philosophy. The regulations concerning ACAS/TCAS published by ICAO and as a result the regulations of national aviation authorities, operational and procedural instructions of the TCAS manufacturer and the operators were not standardised, incomplete and partially contradictory.

So while the accident was caused by a single mistake from the controller, there were also many flaws in the system, which all together allowed the mistake to develop into an accident.

(1) By system I mean the aviation system as a whole, not only the air traffic control (computer) systems.

-
Another case of a chain of failures. Clapham Railway disaster railwaysarchive.co.uk/docsummary.php?docID=36 –  Portree Kid Jul 7 '14 at 13:19
"Never did anyone give guidance on what to do if the controller and the TCAS gave conflicting instructions. " European (and most anywhere outside Russia) were and are to follow TCAS in such cases, Russian regulations were to leave the decision to the pilot, who typically slavishly (this was shortly after the Soviet era) followed instructions almost without thinking. –  jwenting Jul 9 '14 at 9:57
I haven't heard of the Swiss cheese model (which describes it quite well), but the accident chain is a very common way to describe the same thing. Break any one link in the chain of events leading up to an accident and you prevent it altogether. –  Lnafziger Jul 9 '14 at 11:18
@jwenting Do you have any reference for that? Before 2004, the ICAO requirements did not address the case of TCAS conflicting with ATC. In fact 6 of the 19 recommendations from the investigation are addressing the uniformity and quality of TCAS use worldwide. Also note that a year before the Uberlingen crash, in 2001 a near miss happened in Japan which also involved conflicting instructions between TCAS and ATC. Both aircraft involved were Japanese. –  DeltaLima Jul 9 '14 at 11:49

Looking at this as a deliberate act it is really possible, but it would still take the element of chance to make it work. There is technology designed to prevent such an occurrence:

• ATC software. The picture that ATC sees is computer generated with lots of extra information, and it looks for potential collisions and alerts for them. I believe the US system sends both audio and visual cues called "deals"
• Commercial aircraft above a certain size is almost always equipped with TCAS (Traffic Collision Avoidance System) which will warn pilots about potential collisions and tell them to climb or descend
• The human eye: pilots are often looking out and may spot a potential conflict

So a controller wanting to cause a collision could set 2 airplanes without TCAS on a collision course in bad weather or in the dark, then ignore automated conflict messages and cause a very dangerous situation. That doesn't guarantee a collision as there's a lot of airspace up there, so the controller could do all that diligent work and only get a "near miss".

As for accidents yes, controller errors have led to collisions in the past. It's usually a cascade of errors rather than a single error.

-
At airliner cruise speeds looking out the window doesn't help much. Pilots cannot see up or down, and head-on collisions happen too fast for the crew to react. –  paul Jul 7 '14 at 8:39
@Paul, head-ons are one thing, but collisions in commercial aircraft have been avoided many times by alert pilots when approaching at an angle. Plus, it only take a few feet to avoid a collision. See and avoid is still an important process no matter how much technology there is out there. –  GdD Jul 7 '14 at 9:23
@paul that's not really true. Many cockpits have great FOV though seeing directly down or above+behind can be problematic. At altitude visibility can be tremendous and you can sometimes see planes at quite a distance. Of course it helps if the sun is reflecting off the other airplane or that airplane is seeding a contrail. That gives decent time to react and even if you cant see you an listen to the TCAS which gives ample warning of a potential conflict (TA) before giving a direct command needing immediate reaction (RA). –  casey Jul 7 '14 at 13:48
@casey: I agree that the FOV is usually good, but: In PSA 182 the pilots probably (accodring to the documentary; not sure if the WP page mentions it) reduced their field of view by not setting their seats properly and GLO 1907 did not notice each other in head-on collision in good visibility. –  Jan Hudec Jul 13 '14 at 20:28
If I recall correctly a "deal" is an operational error, the thing alerting everyone to an impending "deal" is called the "snitch" :) –  falstro Jul 30 '14 at 16:09

It not only is possible, but actually occurred over Southern Germany.
Due to another controller resting one controller was watching two workstations. Just before the crash he became aware of the situation an told one plane to descend. Unfortunately TCAS alerted the other plane to descend, too. The first plane ignored the TCAS instruction to climb and both plane collided killing 71 people. The controller was later murdered by a grieving relative.

-
The controller was later murdered by a grieving relative. - Wow, that's messed up. –  Davor Jul 7 '14 at 17:30
The crash was not caused by him, but rather by a plethora of interacting flaws in procedures and equipment. The one blatant error he made (and that under pressure) was to tell the Russian aircraft that the other plane was coming from the right when in fact it was coming from the left, but at that time it was too late anyway. The main factor in the accident was the Russians' failure to follow international procedures for dealing with TCAS warnings. –  jwenting Jul 9 '14 at 9:46
@jwenting of course it wasn't the only factor and it was a string of events that led up to it. That is why it is so important to have procedures and to follow them. Like that TCAS overrides ATC. That said I cannot accept him being killed! –  Portree Kid Jul 9 '14 at 10:38
@PortreeKid of course not, killing him was cold blooded murder, honouring him for that murder just shows how criminal that entire country is. I think one of the results of the investigation was a change in Russian procedures to align their TCAS handling with the rest of the world. Just like a result of Tenerife was to introduce common phraseology and procedures worldwide for runway entry and takeoff clearance. –  jwenting Jul 9 '14 at 11:18
@jwenting: The factor was that the procedures were not established at the time yet. The investigation report does not list this as fault of the Tu154 crew, but as lack of such rule. Of course introducing that rule is in recommendation in the report and it was introduced and in any new accident it would be fault of the pilot if they didn't follow TCAS. –  Jan Hudec Jul 13 '14 at 20:35

This one goes back a ways, but yes. (early 70s) Center split up a flight of 4 F-4s into 2 flights of 2 over Florida, stacked them over a navaid in the holding pattern. Controller confused the call signs and decended the upper flight of 2 through the 2 holding a thousand feet below. One F-4 lost the vertical stabilizer, both pilots landed safely after a nylon let down. The other F-4 in the mid air lost 6 feet of wing and landed safely engaging the approach end BAK-9 arresting cable.

-

Sure. I have been given an incorrect heading to fly that could have been tragic (fly into mountain or another aircraft). Just like pilots, controllers can lose Situation Awareness. Especially given their schedules that have them working crazy shift with minimal down/sleep time.

-