Holding on the runway is common enough. You’d be told to “lineup and wait”.
As @Richard_Pearse says, holding in takeoff position itself is normal.
As you suggest one of the changes 10-15 years ago in the USA when we switched to the ICAO “Line up and wait” terminology is that if ATC is waiting for any delay that’s not obvious, tell the pilot what it is and about how long. I do not know whether that extra info is an ICAO standard. I doubt it, just due to the simplified rote-ified nature of ICAO English procedures which are intended to work when English is not the native language of the pilot, the controller, or both.
Unrelated to that:
Upthread in the context of another runway incursion near miss I made the comment that (good) pilots try to piece together a mental mockup of the relevant traffic situation from what they hear on the radio.
e.g. If you’re on final and you hear somebody cleared to “line up and wait”, that’s a clue that you should register. Namely that the runway is now obstructed and that situation needs to resolve before you get there, whether you can see anything yet or not.
The gotcha is that pilots can only gather that info from what’s happening on the frequency they’re talking / listening on. The moment you change freqs, your picture mostly obliterates and you need to start building a new one from almost zero.
As applied to the Haneda crash …
From the ATC analysis I read, JAL was cleared to land before the Coast Guard (“CG”) checked in with Tower. So assuming typical pacing & radio congestion, Coast Guard would never have heard JAL was out there. If the weather was nice, JAL should have been visible to CG, at least until CG entered the runway and turned their tail towards JAL. But … Per the 2019 Haneda chart I was able to find online, the angle that C5 enters the runway puts the approach area where JAL was behind the FO’s right shoulder. Making it likely that their last opportunity to see JAL was even earlier.
From JAL’s POV they’d have heard other airplanes told to hold short at the threshold and CG told to hold short some ways downfield. Which sounds, and is, completely normal to them.
When mixing turboprops and big jets, it’s common that they enter runways downfield a bit, happily take off on the remaining shorter length with less waiting, then once airborne quickly turn away to the side to get out of the way of the faster traffic launching next. Likewise on landing they often vacate the runway quickly due to their short stopping distance and end up with a slightly oddball taxi route to avoid getting face to face with big planes taxiing towards the takeoff end of that same runway. All normal stuff done everywhere every day.
A consequence of turboprops entering the runway downfield is that they’re not sitting where a landing jet is naturally looking for stopped traffic. A pilot who’s overly focused on just the touchdown area might not notice the smaller plane sitting farther down the runway. It’s easy to see familiar expected anomalies (e.g. airplane sitting at the approach end) and harder to notice unexpected anomalies (airplane sitting well down field) when workload is high and/or visibility is low. Darkness certainly limits the visibility of obstacles, net of whatever lights the CG may have had on. But busy airports are seas of blinking and steady lights and moving and stationary lights and sometimes there’s just not enough contrast.
At night there’s another and bigger issue. The ~3000 feet closest to the approach end of a big jet runway is very well lit with in-pavement lighting. It’s a carpet of light to land in/on. It’s lit well enough that an airplane sitting there is obvious as a dark airplane-silhouette in the midst of a dense surface of lights. In this case the CG was sitting about 1000 feet past the end of all that lighting. Which means the JAL pilots had to see a dark object with limited lights on it, sitting in the dark area just past a brightly lit area. Tall order the way human eyes work.
Bottom line:
So there’s lots of little holes here the the layers of swiss cheese that all had to line up just wrongly to crunch two planes & kill 5 people.
Which of course is totally the norm in aircraft accidents. One mistake is mostly tolerable - the overall system redundancy will trap it before the situation turns ugly. One mistake with 5 more unfavorable factors piled on is not tolerable; the redundancy has been eroded to nearly zero. Then we’re just relying on luck. Which given the huge number of operations per day worldwide, is not a reliable partner. Or rather it’s a partner that is statistically reliable: statistically reliable to fail us regularly.
Oops. Sucks to be them.
This guy needed a lot of anger management or some kind of intervention. And is it common for a pilot to be cleared to carry while flying a plane?
See dedicated thread here:
In a perfect world all the communication would be on 1 frequency so everybody was on the same page but even a small airport would be overwhelmed by such an arrangement. That ship sailed in the era of the DC3.
I’ll ask again, do commercial planes use their taxi/runway lights while moving on the ground? If yes is this visible from above when on short final or is it washed out by all the surrounding light at the end of the runway?
Yes they do but they are forward facing lights. The Dash 8 was facing away from the A350 so its taxi light wouldn’t have been visible. All you have are the nav lights (green/red on right/left wingtip and white at the tail) and the beacon/anti-collision light. That sounds like a lot of lights but the touchdown zone of a runway at a large airport like Haneda has a lot of lights so any on the Dash 8 will have been hard to spot unless you know to look for it.
By (US) regulation position lights are required from dawn to dusk on small planes and all the time on large planes. The anti-collision lights which are rotating red beacons and or red strobes must be on when engines are running or are about to be.
After that you get into pilot technique and or company procedure. Which also depends on what ights are instaled on that type and how much fine-tuned control the pilots have over what’s on or off. As a general rule:
-
On older planes the taxi light is generally mounted on the nosegear & swivels with it. While moving during taxi turn on the taxi light. Turn it on just before releasing brakes as an “I’m about to start moving” signal and turn it off immediately once you’ve stopped & reset brakes, or while moving as a signal to someone you’re supposed to give way to that you acknowledge you see them and will stop before you get to where the two paths intersect.
-
Some airplanes have white wingtip and tailtip strobes that are independent of the red beacons / strobes. If installed these are generally turned on to enter or cross a runway.
-
Each model has different variety of additional lights. Typically there is a separately controlled runway turnoff light on each side which aims about 45 degrees off to each side illuminating the area you’ll be turning into if you start now. If installed these are generally turned on to enter or cross a runway.
-
In addition there are 2 to 4 forward facing landing lights which point straight ahead and are insanely bright. Depending on the airplane there may be one switch, 4, or some kind of rotary switch for different combos of what’s on and what’s not. Do not use the bright landing lights willynilly lest you blind any other taxiing airplane you happen to be pointed near. Do not turn them on to enter or cross a runway. Once actually aligned with the runway and cleared for takeoff, switch everything to full bright, release brakes & away we go.
As @Richard_Pearse said, the lights that are real obvious are forward facing. While JAL was approaching from the rear.
An issue with more modern airplanes is they tend to have one large LED array in the wing roots and that’s it. All the various lights regardless of where aimed or how bright are just different sections within that array. So it can be hard for anther pilot looking at that from the forward quadrant to accurately tell which lights are off or on and do divine the crew’s intent from them.
As well, because these things are matters of company or individual technique, any signal you think you divine from the lights you think you see must be taken with a lot of skepticism until it proves out. And most of that signalling is only meaningful among slow-moving airplanes on the ground.
I’m a month late to the party, but this was a phenomenal read, and I’d suggest it to anyone interested in aviation or in any sort of regulated manufacturing environment. My quality manager will be telling a (VERY) brief synopsis in our next town hall to drive home why our systems and policies matter - though in medical manufacturing, a defect gone wrong tends to only harm one person at a time.
Which brings up questions for me. Do most airliners have ADS-B in? How common is it overseas? It is cheap enough that many US hobby airplanes have it, shows ADS-B traffic (which in the US anyway would be everybody at a class B airport like Haneda). I know I usually see traffic on there before I can pick them up visually. Provides one more chance to “see and avoid”.
Sometimes the problem isn’t detected for a while, though. Therac-25, et al.
In software, the bugs often live at the interfaces. That’s where mismatched assumptions are least likely to be detected. I find it interesting the “bug” here was also at an interface, but an unusual one–the mismatch between two different ways of specifying a coordinate system. Hard to detect since it’s the same thing, in principle. But the allowable error was very different.
US has a B-1 bomber crash at Ellsworth Air Base. Crew ejected safely. No other details. Weather was crappy with icy fog.
ADS-B in isn’t common on airliners. There is TCAS which displays traffic on the navigation display but the TCAS display isn’t always reliable in azimuth as it doesn’t use a GPS position to determine a potential conflict, instead it uses a rate of closure algorithm.
A bigger problem is that the smallest range scale on the nav display is typically 10 nautical miles and so even if the target position is highly accurate, you’re not going to be able to differentiate between an aircraft on the taxiway vs the runway.
Finally I gather the Dash 8 didn’t even appear in the usual ADS-B feeds so I don’t know that it would’ve been displayed on an ADS-B in unit anyway.
Two rankings of airline quality were recently published.
This one ranks 11 US airlines according to 14 different quality criteria. The data is presented in the form of one of those annoying slide shows, but in a nutshell, Delta was rated best and Southwest by far the worst. When my son was doing a lot of international travel he was very fond of Delta.
Here’s a different survey that ranks the 10 biggest North American airlines just in terms of on-time performance. Once again, Delta is #1. Disappointingly for this Canuck, Air Canada is dead last. The AC prez blamed bad weather in June and July for some of the delays, as well as staffing shortages, and promised to do better – and indeed they were much improved later in the year. Westjet did better but not by much – it came in at #7.
Honestly, though, I’ve had other issues with AC over the years. More than on any other airline in my experience, flight attendants tended to reflect a culture of entitled complacency. There were always exceptions, of course, and unsurprisingly, the good ones are found much more often in business class and first class.
Of course if the thing you are manufacturing is medicines, not machines, you can certainly harm a lot of people all more or less simultaneously with defective pills and potions. Whether they are harmful or just ineffective.
Here’s a recent blog post on both by one of my favorite bloggers: Compounded (And Counterfeit) Semaglutide | Science | AAAS.
As @Richard_Pearse says. Very rare today.
In long-winded greater detail …
There are experiments afoot now in the US to add ADS-B to airliners. With the intent of letting crews manage their own spacing to follow traffic ahead to the runway. Assuming the experiments are fruitful (read “saves money”) it’ll be 10 years before it starts rolling out US-wide and it’ll be 20+ years before it’s well-nigh universal.
While you’re in flight, TCAS filters out what it perceives to be traffic on the ground. No ground traffic is displayed at all. Otherwise every airport would simply be an impenetrable blob of dozens of TCAS targets’ symbols. Every one of which you’re converging towards laterally while closing in relative altitude. TCAS firing a proximity alarm over that would be bad. So TCAS is, by design, 100% useless for a landing aircraft to detect an occupied runway.
The converse is not true. You can use TCAS on the ground to see airplanes approaching the airport. The collision alarm feature is disabled though. It and ideally you too are watching, but any action is 100% on you.
As @Richard_Pearse also said, the resolution is not great. When holding short where you can visually see the final approach area to busy parallel runways you can look out the window and easily see and separate the two streams of airplanes heading to the two runways. Meanwhile on the TCAS display you can maybe resolve them into two streams, but it looks a lot more like one lumpy stream. Given just a single target out there, you’d be very hard-pressed to know whether it’s coming to your runway or the adjacent one until very very late; like too late to taxi away to avoid a collision.
Another obstacle is that all the TCAS displays are optimized for inflight use. So you have a lot of “view” on the display forward of you. At minimum range you might be able to “see” TCAS targets 10 miles dead ahead, 5 miles due left/right, but only 1/4 mile behind. Sitting on a runway, the threat is coming from dead astern and that’s where your display range is shortest.
On-time performance is subject to considerable gamesmanship. Cancellation performance, also called “completion factor”, are subject to insane levels of gamesmanship. Baggage (mis-)handling stats are probably the least gamed, but there’s still some room.
Back in the ~1980s the US DOT mandated that statistics be collected in a certain fashion. The detailed rules were a very crude first draft effort that haven’t been updated since. With the predictable outcome.
The whole effort is borderline useless, and especially as to ranking the top 3 or 4. Any carrier can “buy” their way up that ranking a few places. If they think the extra expense will deliver enough extra reputation that in turn produces enough extra revenue to pay for it. Or conversely they can consciously forego the expense and settle a bit lower in the ranks.
I assume you mean ADS-B In is rare, because ADS-B out is required in class A-D.
ADS-B in is a wonderful thing (I love it) - but seems to me it has the same issue where you going to get a blob of targets when near a busy airport – or is it good enough to id an airplane on the landing runway vs one on a parallel runway? I’ve seen airplanes on the runway on foreflight but I fly (mostly daytime) VFR (at usuall unbusy airports) so by then I’m mostly using mk 1 eyeballs.
Brian
PP-ASEL
Thanks for the correction / amplification. ADS-B in is what I was talking about.
If one had ADS-B in traffic data available on board, a “runway occupied” warning could be programmed. Each ADS-B out message includes the lat/long of the transmitting airplane to pretty tight tolerances. Enough that in principle receiving airplanes could geo-fence their runway of intended landing and be warned of intruders with a tolerably low false-alarm rate.
Which would work great in Class A-D airspace and for airplanes flying to/from airports that are fully characterized and programmed in the aircraft’s FMS or equivalent. Which encompasses the vast majority of airliner ops, but not necessarily the vast majority of lightplane/bizjet ops.
I use Avare for the display, I zoom in when at the airport or in the pattern and zoom out for cruise. Zoomed in I can tell planes on the runway vs. on the taxiway. Also just day VFR at smaller airports, but lots of training traffic most days. Don’t have a parallel runway situation to have experienced that.
The irony is that we’re talking about $2000 worth of existing equipment and software using an ipad that commercial pilots already use to replace paper. But to put a certified system in the cockpit is a long process as LSLGuy described.
This is what general aviation has access and is using now. I suppose it would be major no-no to bring your own equipment on board but it reminds me of the early days of handheld GPS units. Commercial pilots would bring their own units onboard when crossing the ocean because it was way more accurate than the system in the cockpit.