In the way they experienced and remembered it. Which might be far from what really happened. The stress they remember and the stress their voices do show might be totally different. If they felt this was a very dangerous situation, they would have decided for a go-around. As they acted wrongly, they may remember wrongly, even if they are most sincere with the investigators. A CVR remembers accurately, neutrally and with a correct timestamp, humans don´t.

Pukingdog, Brilliant!

Portmanteu:
Indeed, the CVR is primarily the voice of the dead.
Secondarily used for cases like this.
(Mr CEO in Quatar has other plans for the future that he has experimenting with, but lets leave that one for later, but it is relevant!)

As NTSB goes , yes they shall indeed ask a lot of questions!
But do not for one second assume that they get the correct answers in a proper timeline.
I shall back off on my rather harsh description of the crew, but once parked and debriefed OPS AC should have told them to secure the CVR.

When debriefing the Crew and later the wet debrief at the hotel, after a quick call to the Union boss, I think the gravity of the case starts to sink in.

If OPS did not just DH them back to base the next day, they flew active.
Regardless, would you not think they spent a minute or 3 to go over what happened?

If the standard 2 hrs CVR was saved NTSB has a starting point at least 30min before TOD , possibly more than 1 hr. The brief would have been there.
Now they have a crew that have had a week or more to communicate to interview. With no facts to back it up.

I am not saying the crew will come with false statements or keep important details to them selves if not asked.
After all this is NTSB not the Police. And they do not know if NTSB can recover
some or all of their last conversations ( unlikely, but).

I rest assured that their right are taken care off by AC and their union.

Now , there is monumental events in aviation that leads to positive improvement or re -focus on basic skills.
This is one of them.

Volume
You beat me to it!
Have done a few checklists 2 times: " Did we do XYZ Checklist?" FO:" Yes! I think so!"

Even when the pilot's report is identical to the CVR, the CVR is a better source to cite when arguing for changes in regulations, procedures, equipment, etc.

I notice an interesting pattern in the second-by-second analysis shown in earlier posts. The pattern may indicate that the AC pilot(s) may have initiated the go around before the time stamp indicated.

Elevation above ground has already flattened out near 130 feet as the plane overflew UAL1. The next five seconds of data would be consistent with the plane's nose beginning to lift and the tail continuing to drop. For about four seconds there is more tail drop than nose lift, presumably, then the full climb is achieved about two seconds after the indicated marker.

PukinDog,

Very well detailed and thought out post.

To add to this, 28R has the FMS/RNAV visual, but it also has a pure visual approach, (that does not align with centerline) which several other aircraft preceeding them had used.
Should the FMS on the ac be off, the visual approach is still there, if nothing else, as you very well illustrated, the visual clues to back up the automation.

With the FMS visual procedure, the airline and the individual crew must be authorized to use, so there is a bit more involved in training, and hopefully the briefing.
That big white cross at threshold, that is not difficult to miss...

http://i1108.photobucket.com/albums/...ps7bd67c11.gif

Looks like we also need this on the taxiway.

I fully agree except the most important point has been left out. Runway 28L is NOTAMed as Out of Service. So there will only be Runway 28R available. Both crew members minds now expect only to see one runway.

Similarly had the Tower (or a previous arrival controller) just added runway 28L is out and unlit, expectations would have changed.

But all the holes in the cheese lined up. Including worst circadian time at the end of a long day.

I am inclined to think they were not aware it was out of service. This led them to think the lit runway on their left was 28L and 28R had to be further over to the right which was where they headed. How they could do that in the absence of any approach lighting is the mystery yet to be solved.

Is AC 'dispatched' with a briefing or is it where the crew print out an armful of paper, have a coupe of minutes to review the salient items, load fuel and blast off. During the cruise they have time to read through the NOTAMS. How are they in Canada? Is it easy to pick out these critical items or are they hidden in the acres of garbage we are often presented with in Europe? Either way it was not a short flight. Only the crew will know.

From the FAA:

SMES (Surface Movement Event Service): ASDE-X/ASSC data, OOOI events (Spot Out, Off, On, Spot In), Category 11/Category 10 position reports.

ASSC improves surface surveillance and situational awareness in all kinds of weather. ASSC is similar to a prior system deployed in the U.S. called the Airport Surface Detection Equipment, Model X (ASDE-X), which is deployed at 35 airports.

In October 2016, ASSC will be operational at San Francisco airport. Over the next few years, it will be implemented at:

https://www.faa.gov/nextgen/programs/adsb/atc/assc/

It appears that the system referenced for the aircraft location and altitudes was the first of its type installed by the FAA.... Did this new system provide warnings of the conflict?

Since the new and improved system does not appear to have detected the conflict, what about its predecessor which is deployed a minimum of 35 airports in the US, and an unknown number of airports worldwide?

here we go, a few more holes in the Swiss Cheese ...

EDIT: 35/36 airports for ASDE-X.... http://saab.com/saab-sensis/air-traf...gement/asde-x/
Following the success of the ASDE-X program that covers 35 airports in the USA, the FAA decided to extend the technology to nine additional airports: Anchorage, Andrews Air Force Base, Cincinnati/Northern Kentucky, Cleveland, Kansas City, New Orleans, Pittsburgh, Portland, Ore and San Francisco.

The ASSC system deployment includes multilateration sensors, data fusion, conflict detection and alerting, and the same HMI already in use at the 35 ASDE-X sites.

Still does not explain where this "to the foot-precise" altitude or height comes from. ADS-B/mode S rounds to the nearest 25-feet, surfaces movement radars are primary/2D radars. FDR - RadAlt seems a very likely source, but this data shouldn't be available to Flightaware and neither so soon ...

Regarding the deconflicting, depends what they had in mind when designing the system, rwy incursions, etc. yes, wings clipping at the gate, most likely not, twy incursions 'from above'... probably neither....

The graphic was from FlightAware using data from the FAA's Surface Movement Event Service

Yes, this system is designed to alert to conflicts from traffic on approach and ground operations.

Please read the links provided showing the capability and expectations of the system before responding. (data is used from the terminal radar system)

The cost-effective alternative to the ASDE-3/AMASS capability, referred to as ASDE-X, is one of the first new runway safety program technologies aimed at improving ATCS situational awareness by providing tools to supplement their tasks (McAnulty, Doros, & Poston, 2001). The data that ASDE-X uses comes from a surface movement radar located on the airport traffic control tower or remote tower, multilateration sensors, ADS-B (Automatic Dependent Surveillance-Broadcast) sensors, terminal radars, the terminal automation system, and from aircraft transponders. By fusing the data from these sources, ASDE-X is able to determine the position and identification of aircraft and vehicles on the airport surfaces, as well as of aircraft flying within 5 miles (8 km) of the airport (selectively up to 60 nmi).

As many are aware, I have not been a proponent of the ADSB or Mode S for this type of use, due to the accuracy issues, especially on descent.
That being said, this is the data that the FAA system provides, by whatever means.
That also being said, that is my question on validity of the alert capability of the systems already in operation.

So you are saying terminal radars provide this accuracy? If publicly available, thanks in advance for providing souces/links so I can learn. Technology moves fast, my knowledge stops at the 25feet rouding for mode S fed SSR's or ADS-B dataframes precluding foot-precision. But as stated, always keen to learn...

ASSC certainly looks promising but a couple of issues....

To take full advantage of it, aircraft will need both ADS-B in and out. Airlines are fighting the out mandate. No anticipated mandate for ADS-B in. That may be market driven with numerous programs for reduced separation based on having both ADS-B in and out.

Also, for maximum protection, the aircraft should be the source of any warnings. This is possible with ADS-B (I served on a ARINC/MITRE committee researching this some years back). The problem with a ground (ATC) based system is the inherent delays associated with:

(1) recognizing the issue/conflict
(2) issuing the warning to ATC
(3) delay in interpreting the warning
(4) issuing a warning to aircraft involved (with the potential for transmission being "stepped on" as in Tenerife).

Dibo, no worries, please keep in mind that is what THEY say the systems are doing.

Shore Guy. I hope you enjoyed the mind numbing experience on ADSB committees as I did. In reality, I really dont see why they are flogging this antiquated system, especially with all of the limitations. The broadcast string is too short to have a viable encription, so you will never, ever see ADSB-In. Boeing swore to that.
We trialled ADSB-In with test flights into Brisbane, and the tested many potential capabilities, and with what we were able to do, it scared the hell out of them.
The ability to input data to the FMS should scare anyone. Do it with the charecture string available in the ADSB tranmission to provide any encription, lemming talk.

That being said, the tech is going to the Ku band transmissions through uplinks to/from the ac. Many new aircraft already do this, such as the 787 variants and most new 777 variants. Boeing doesnt monitor 40,000 datapoints live on the ac through ADSB.
By the time ADSB is sorted, if ever, it will be like comparing AM radio to Broadband Wifi.

With remote tower operations, SWIMM, and other acronyms making headway, it will be interesting to see how the ASDE-X/ASSC system failed to alert an aircraft centered on the taxiway for at least 4nm affects current and near term operations.

Underfire and all,

Yes, I felt the same during my committee work. When I initially became involved, my thoughts were my kids will think this is pretty neat.

That became modified to thinking my kids kids would think this is pretty neat.

SOOO many potential downfalls to ATC using this as sole source global ATC infrastructure. It has some positives, but the FAA seems driven to hang onto this concept since (ancient history) CAPSTONE.

Does the world really want to hang the entire worldwide ATC system to a low signal/easily jammed source of navigation information (GPS, etc.) with no backup?

I think not.

And with the bad guys still out there, a ground based system (radar) will be required near population areas anyway. Whether in the US, it is budgeted under FAA or DOT, no difference. It will be necessary.

Data links, bandwidth, etc., etc. still major issues.

Remote areas (Western Australia, Hudson Bay, etc.) already using it with benefits. Worldwide satellite com will help with Oceanic applications.

Bottom line: not ready for prime time. I could go on, ........

So far, the FAA has no plan to decommission TRACON ASRs.

As to ASRSs (center primary/secondary long-range radar) those belong to the military, and will not be going away anytime in the foreseeable future.

In absence of information from CVR, crew interviews etc., this seems most likely. Otherwise it's hard to see how they ignored the mass of approach lighting a few hundred feet to their left. Still leaves the question of how they approached a major international airport at night without any approach lights on the 'runway' and carried on. The circadian time seems probably very important. And as many have already alluded to, seeing what you think you should see.

When all is said and done, hopefully a wake up call for everyone, that any crew, from anywhere, can make potentially disastrous errors on a nice easy VFR to a big airport.

I mentioned the use of ADS-B technology in an earlier post on this thread, but didn't explain how I thought it could be used. Perhaps the presentation on Day 2 at 16.20 (ATSA-SURF - Enhanced situational awareness under adverse weather conditions) will help to explain what I mean:-
Joint Fifth ASAS-TN2 Workshop and Second FLYSAFE Forum ? ASAS TN

I know these are advanced concepts but I remain convinced that the display of other traffic using ADS-B In and Out, together with data linked warnings and taxi instructions will be the way forward as traffic levels rise and to help resolve mis-perceptions of the type that may have occurred at SFO.

One day soon I hope that a technological solution along these lines will become the norm.

There is an important difference, between knowing it was out of service (logical level) and being aware that it would simply not be there (subconsious level). They probably knew it was out of service, their subconsious expectation however was that it still would be visible, which it probably was not. They were very well aware, that they had to use the right of the two runways, they were not aware that what they saw were 28R and Taxiway C, and not 28L and 28R.

Definitely. The more tired you are, the less you think and the more you act intuitively.

The NTSB wasn't as sympathetic as that. See attached.

Aterpster and all,

One of the big selling points by the FAA for ADS-B was to be able to get rid of the "expensive radars".

Not going to happen.....

Also, I believe that a backup source of navigation will ultimately be needed to provide a secure system. Enhanced Loran C comes to mind.

Imaging the consequences of a worldwide hack/shutdown of GPS......

Oh but it did happen. The DoD is paying for the 'expensive radars' which are now not on the FAA books.

What evidence of fatigue do you have for the SFO incident. Are you aware of their schedule or sleep patterns? I think not. I suspect that you are just making an assumption based on the time that the incident happened. But perhaps the pilots were well rested prior to the flight.

There seems to be an automatic assumption by some that if any incident happens at certain times or on a long flight, that fatigue is a factor, perhaps in an attempt to deflect attention away from the possibility of crew error.

As of right now, most of us have no idea on whether the crew were fatigued so we should not be posting in a manner as if it was likely. The reality is that it is a possibility along with many other possibilities and that's all.

The other reality is that there are flights all over the world that are long haul and at strange hours. Yet they are completed without incident. Misidentification of airports and runways happen on a regular basis to fatigued and non-fatigued crews.

As we can see in the UPS cause that was posted, the captain had some deficiencies that were seen in training and resulted in the accident. I'm sure they were tired, but so were probably 25% or more of UPS pilots in the system that night and every night.

That presentation could be given today, given what was shown and what has been done. Sorry to say that was done in 2007, so in 10 years, nothing has happened.

ADSB signal consists of 112 bits, that is it, you barely get the root and the ac id in that, with coordinates, so forget about any encription, just not enough room. Even the other information is pretty granular.

ADSB may have use for ground movement, but in reality, you will never ever see ADSB-In on a commercial aircraft. Military has it, but it is built behind a very secure encryption system which is not available commercially.

Jammedstab, I think a distinction needs to be made here between fatigue and circadian rhythms. When I mentioned the "C word", I intentionally did not accuse the crew of being fatigued or even tired. They may have been, but we have no info on that.
What we do know is that it was nearly midnight local time, and the flight was originating from a place where it was then 3 AM. Contrary to your suggestion, we can guess a few things about our crew. For example, they were not based in San Francisco. Also, it was probably between 3:30 AM and midnight on their base time.

And here's the thing about Circadian lows: they're not based on fatigue, but rather what time of day your body thinks it is. Industrial accidents have their highest rates between 2-4 AM and 1-3 PM, and that's with people who are well rested, sober, and in the time zone.

Sure, most crews pull it off just fine. Most times, you'll get by without the last five minutes of final reserve fuel too.

Part of it is self-discipline, experience, and proficiency. I am speaking of this crew, rather the fatal UPS crew.

I flew many lousy night pairings with TWA before the duty/rest rules got a bit better.

As tired as we may have been on a multi-leg flight from LAX to the east coast on the back side of the clock, we managed to get ourselves alert for the terminal area business at hand, whether it be CAVU or bare minimums.

Don't really know. Perhaps the lack of automation forced us to be more alert. The human-factors folks probably don't have a real good handle on it.

Also, commuting from a distant location to begin a night flight should be illegal.

Having had the domicile I lived in closed causing me to commute, I'll be happy to agree with you when the company purchases my house and provides moving expenses to a domocile within my time zone. While I'm senior enough to bid around the trips you mention, not everyone is.

I have no problem with commuting provided it is subject to FAR duty time and rest regulations.

I was forced to commute from LAX to STL for several months. I always went over the day before and spent a night in a hotel before taking my pairing the next day.

That's not the way it works for most commuters. They ride the red-eye then report for their pairing a few hours after arrival, at most. Domiciles open and close. Same happens in other industries. That's a personal/employee/employer problem that should not be transferred to the flight deck.

Also, these days many commute because they chose not to move to the domicile in the first place.

There was a time living within "reasonable ground travel" of the domicile was mandatory. Why I never applied to NWA.

TWA never had that. But, before the early 1970s the jump seat couldn't be used for commuting. So, only the senior guys with term passes could commute.

Circa 1970 the LAX chief pilot issued a letter that all LAX pilots had to live within 50 miles of KLAX, The union got that rescinded in short order. We had guys driving up to 200 miles in those days.

That is the problem, you are just guessing. How do you know that they were not Vancouver based? You don't do you.

And even if they were based in the east, how do you know that they were not well rested prior to departure? You are making an assumption(and y are making the suggestion) that this crew had to be tired. I suggest that you are just making guesses as potential excuses.

I don't know how prevalent long distance pilot commuting is currently, but only a decade or so ago, I knew a DL pilot who would leave his home in southeastern MA (BOS pilot base had closed) about 0930 to catch an 1130 DL shuttle flight BOS-LGA, taxi from LGA to JFK to command the DL flight JFK-SVO leaving between 1600-1630. This would happen 4-5 times a month.

His landings at SVO came at 0200 or so "body clock time" after a pretty long day. There was a rest period during the TATL flight.

We're still conflating fatigue with circadian rhythm. It doesn't matter if the crew was well rested. If the pilots were flying "out of sync" with their biological cycle, then their cognitive performance might be compromised.

There's a lot of research indicating that most of us cannot perform optimally past 2am unless we only work nights for an extended period of time -- e.g., doing night shifts exclusively for several weeks in a row without break.

Yet as soon as we switch to day work (even for just one day) our body clock "resets", and our night time alertness suffers again -- regardless of how "well rested" we might be.

Speaking of ADSB, been hearing rumors that the 2020 mandate in the US is not going to happen, anopther delay will be announced soon (associated with the ATC privitization bill)

Well put, peekay4. Anecdotally, a good many years ago I had a summer job assignment that led to an entire calendar month of midnight-to-8:00 a.m. shifts. Not flying anything - compiling production numbers and various numerical production reports for a BOF (Basic Oxygen Furnaces) steel-making shop. Around sun-up or so each shift, I had to drive around the massive steel mill works, in my personal car, to deliver some reports to other departments. Indeed the body clock does adjust to "steady midnights" (and some mill folk take pride in their ability, or perhaps it's only perceived ability, to adjust as necessary).But even so, driving around the sprawling Plant 2 of Indiana Harbor Works, alert for gargantuan fork-lifts carrying ingot molds for example, as the August sun was rising over Lake Michigan taxed even a 21-year-old's ability to ..... look sharp!

I'd say that most commuters I know haven't ever hopped an all nighter, let alone sign in for a paring after, certainly not "most" as you characterize. Perhaps in the past, not these days.
Either way, unless there's a demonstrated, continious problem, I'd prefer to avoid additional government overreach.

Colgan Air 3407 was just over 8 years ago.

It's next to impossible to gather information as to whether it is "a demonstrated, continuous problem." The commute issue of 3407's F/O was of considerable concern to the NTSB.

Given the considerable changes due to the Colgan crash along with the implementation of FAR 117, there's been ample opportunity to address it if it was deemed thst dangerous. As tragic as Colgan was, it was one accident and there are a lot of other pieces to that puzzle besides commuting.

Still lots of NH MA folks flying out of JFK and almost all are augmented. I would just make sure I had the last rest period before making a landing.