UNITED959

This runs through my head frequently these days as more and more of my flights require bright green luggage tags and lack a proper cup table in First Class.

Like any occupation, your skills improve as you do it more. The younger pilots cannot be "faulted;" 99.9% of them do an awesome job. Unfortunately, there simply is no substitute for experience.

I've said this before: despite the many shortcomings of UA from a service perspective when compared to its European and Asian counterparts, I'm convinced they still have the best trained pilots (as well as other U.S. carriers) and I feel much safer on a UA plane.

1111

Looks like it did indeed change his life forever.

kenhawk

I wish Capt. Taylor the best in finding a new career and hope that his stress issues ease with time. United does have the best pilots and is the main reason that I continue to fly UA.

000123UA

It is curious, because the 744 flight deck has a single analog barometric altimeter as well as the "digital" barometric altimeter on both the left and right seat primary flight displays. In addition, both PFDs show the actual altitude above the ground as measured by a radio altimeter. As part of procedure the pilot not flying would callout "1,000 feet" and "500 feet" and an automated audio callout would occur for altitudes below 100 feet ("fifty", "thirty", "ten"). So, there shouldn't have been any confusion about the altitude. What gets sticky, though, is that if the lateral location of the aircraft is incorrect, then the altitude may not immediately seem inappropriate. That said, on any precision or non-precision approach, a DH (decision height) or MDA (minimum descent altitude) is defined and is set on the EFIS control panel. Now, given the stated conditions, it would be likely that the aircraft was on a CAT III-B approach to 28R, which carries a DH of 50' and a RVR (runway visual range) of 600'. Basically, it is hard to imagine if normal procedures are being followed, how there wouldn't have been a "100" callout at 100', followed by an "approaching minimums" and a "minimums" callout; if at 50' the runway or lights aren't visible, the procedure is to go around. 50' is really close to the ground, don't get me wrong, but a go-around at the DH is a SOP. Now, again, if the aircraft were not where it should be laterally, that 50' DH could be much more dangerous over land, near a bridge, a large boat, etc, etc - so not being laterally in the right place is a serious problem. For lessons on controlled flight into terrain and the role pilots play in maintaining situational awareness, see AA 965. In any case, it is extremely difficult to imagine a scenario where a 744 executes CFIT into the bay with three highly qualified pilots in the cockpit for any reason; there are simply so many redundant systems, many of which are cross checked during an approach as part of SOP.

I want to be clear that I have no knowledge of the specifics of this case, just knowledge of aircraft avionics systems and instrument approach procedures. According to Boeing: I wonder what the statistics tell us about CFIT incidents that occur at the hands of US ATPs – are they low time young pilots, or high time experienced pilots? Or both? The AA reference guide for Latin American flying says:

owendelong

Excellent Post 000123UA

I don't have your familiarity with the 747 specific systems, but, as a pilot with a commercial certificate and an instrument rating, several things about this story don't seem right to me...

1. An ILS is not dependent on a database. It's a pretty rudimentary ground-based radio-navigation aid. Basically, it transmits 4 lobes (two vertical and two horizontal) which overlap only at the very center of the desired course. The aircraft senses these lobes and is able to translate variations in the signal into indications of how far and in which direction from the center of the glidepath the aircraft actually is. Because these lobes spread with distance, the accuracy of the instruments in the cockpit increases as you get closer to the runway, and, ILS procedures are designed with wider "protected" airspace farther from the landing for this reason.

2. To the best of my knowledge, the only database-dependent approaches available are Area Navigation approaches (GPS, INS (RNP), LORAN(deprecated), VOR-based RNAV, etc.). Of these, I believe SFO has only GPS and INS(RNP). I believe the 747 is capable of INS(RNP) approaches, so, a database error would make sense in the case of either a GPS or INS(RNP) approach.

However, in the case of an Area Navigation approach, most pilots I know would at least be monitoring the ILS as a sanity check of the data, so, it seems odd to me that they got "that close" to the water far enough off to be an issue without seeing some indication of a problem. Additionally, I know that the Norcal controllers are pretty good and I've never even made it close to full deflection on an ILS (less than 2 degrees off) before I started getting questions from the controllers.

I hope the lawsuit makes it to court and that there is a public transcript available as I think that is the only way we will ever likely get the complete story from the pilots (if even then).

Somehow, so far, what has been revealed doesn't pass the sniff test for me and provides more questions than answers. I tend to think that a database error isn't unlikely, but, the system is designed with checks and balances in the procedures that mean a pilot shouldn't follow a database error into the ground. Perhaps fatigue and perceptual narrowing were an issue here. I just don't know. It would be really nice to get a detailed first-hand account from each of the three pilots.

ff4lhr

WOW.

travel.flier

Oh my gosh! I'm getting chills just reading this. Scary to think I was supposed to be on a 747 yesterday out of SFO, and this could happen to anyone. I'm not usually scared of air travel but this definitely makes me nervous reading it!

SoManyMiles-SoLittleTime

In addition to the excellent writeup by UA123, and his link to the AA accident in South America, I'll relate a closer-to-home incident.

About 15, maybe 20 years ago, flying east-bound out of the LA area as a private instrument rated pilot, I tuned into the SLI (Seal Beach) VOR, which should have given me a heading more or less of 090 direct to SLI. However, I got a direct-to heading of about 010.

The just-released sectional (VFR navigation chart) had printed the Paradise (PDZ) VOR frequency, some 20 or so miles north, as the SLI frequency.

Had I followed the direct-to heading, I would have been in the LAX TCA, crossing the final approach course. A potentially very serious situation (close, by the way, to the location of the private-PSA mid-air some number of years back).

I reported this in a NASA ASRS report, got a call back, and had a long discussion with the NASA folks on this.

aerokitty

Which is why you should always identify the VOR by listening to the transmitted ID. ... .-.. .. is very different from .--. -.. --.. (actually many VORs now transmit Voice ID as well).

aerokitty

I agree with you. Things don't add up here. If the weather was really that bad, they would have shot the Cat III approach as 123 pointed out. There is no "database" involved as you point out, unless the database was the table of ILS frequencies, DHs to set up the instrument system (vs. reading them off the chart). It is possible that something was wrong with one the frequencies tuned (hence the importance to ID each NAVAID).

Besides, who was cross checking the DME? On a 3 degree glideslope each NM from the airport should be about 300 feet of altitude. If the radar altimeter shows 250 feet and the DME shows 2.3 miles, something is amiss...

We are not in possession of all the facts here...

bschaff1

Wasn't there a flight around that time that also had a similar issue? I recall something that I had heard about an incorrect navigation system a few years ago.

The question I want to know is how come this was the only plane that encountered this issue, when this same data was probably on hundreds of planes that would've also landed at SFO. Also, I'd really like to know how someone overwrote the coordinates for such a major airport without such a change being flagged or caught in a review.

luv2ctheworld

I recall a particularly interesting landing once into SFO... foggy beyond belief and I just kept wondering how the pilots managed to do it. Didn't seem like we could see the water... just kept descending and descending. I felt comforted knowing that there was technology to bring our plane in... reading the pilot's account is a bit chilling.

dhuey

I tend to agree. United and Honeywell have a strong incentive to downplay the seriousness of the incident, and the pilot, in pending litigation against Honeywell, has the opposite incentive. The news report indicated that the pilot claims that he can no longer fly due to post-traumatic stress disorder stemming from the event.

At the end of the report, we see that the reporter still doesn't know what the FAA's take was on this. That to me is a critical piece of information, as the FAA is more or less a neutral party with much expertise.

My completely non-expert sense for this is that it two landing approaches, badly off-course, followed by an emergency landing at OAK on minimal fuel looks pretty serious.

UpstateNY

Reported on CBS affiliate in San Francisco (h/t Airline Pilot Central). Allegedly there was some kind of anomaly in the database of the flight management system, leading the plane to be off the localizer on the ILS approach to 28R at SFO. The captain noticed "something wrong" at the last second and ordered a go-around. After a second approach and a second go-around, the crew declared an emergency and diverted to OAK. Apparently this was written up on FT by a passenger on the flight -- FT gets a little shout-out on the news report.

I don't know what to make about the claims against Honeywell (the manufacturer of the FMS), but great job by the crew.

River in Sight

This is also discussed in this thread.