Friday, May 25, 2007

Stick and Rudder

Never let anyone tell you that today's automated cockpits make basic flying skills any less neccessary. As an example, take a look at these two extraordinary videos of B757s landing at Tegucigalpa, Honduras. On the second video note just how far out they are when the radio altimeter makes the "100" call. American runs eleven 757s in here each week.

View from the ground

View from the cockpit

Friday, May 18, 2007

Risk Management or Micromanagement?

I've found that the biggest difference between general aviation and airline flying is the level of standardization. As big and complicated as a Megawhacker must seem to the average GA pilot, I suspect your average Seneca flyer would be comfortable with it after a few hours in the sim. What most GA pilots could not easily adapt to, however, is the regimentation of flow patterns, checklists, memory items, limitations, standard profiles, and emergency procedures. As a CFI, I had to hound students to consistently use a simple checklist. The airlines will not hire you unless they believe you have the ability and discipline to follow established procedures every flight.

It wasn't always this way. During the airlines' early days, each captain was truly the master of his ship; things were done his way and copilots were expected to adapt. World War II provided the impetus for change as the demands of war required thousands of pilots to be trained quickly. Standardization of procedures substituted for experience in providing the consistency the military demanded. The advent of jet aircraft necessitated the airlines' adoption of the military's methods.

Today, the Captain is still nominally the master of his or her ship - but they'd better comply with the Flight Standards Manual, or else! There are certainly a good many advantages to this approach. First Officers no longer have to adjust to the whims of each Captain and Captains don't have to guess how to avoid falling astray of the chief pilot or check airmen. We all know what is expected of us. Having everyone on the same page allows flight standards managers to more effectively manage risks. Air travel has never been safer.

If everything in aviation were neat, tidy, and clear-cut, complete standardization over every aspect of flight would be utterly acceptable, even desirable. The reality is that aviation is fluid, ever-changing, and sometimes unpredictable. It punishes those who confine themselves to a small box of rote procedures with no room for variance. There was a NASA study that found that around 80% of airline emergencies involved situations for which there is no checklist. This demands crews that are adept at gathering available information, critically thinking through the options, and making smart decisions without worry that their actions will be questioned by those safely esconded in their cubicles. It requires experienced pilots who've seen a strange thing or two before and know how to bend established procedures to fit the situation. A culture which dictates every action for every phase of flight and demands blind compliance is not the best environment to produce such pilots.

The inability of many flight standards managers to realize that procedures cannot and should not be written for every conceivable situation has produced a disconnect between those who write the procedures and those who fly them. While flight standards people often have encyclopedic knowledge, many line pilots consider them to be out of touch with the realities of the line. We call them "bean counters." We are particularly vexed when they consistently take quantitative approaches to risk management which ignore the real-world consequences.

One example of this at my own airline came over a change in landing procedures. I've written before that the Megawhacker is a harder-than-average airplane to land well, given a six degree maximum pitch angle for tailstrike avoidance. In the past, we had the option to do landings at Flaps 15 or 35. F15 is the harder of the two because your angle of attack on approach puts you closer to maximum pitch on touchdown. Furthermore, the situations that require F15 landings are often marginal ones: Cat III approaches or heavy, hot & high landings at mountainous airports. In the past, many line pilots (myself included) made it a point to practice one or two F15 landings per trip to keep ourselves sharp in case we had to do one in marginal conditions.

Now, we've never had a tailstrike, but other airlines have, and it's a very costly occurrence that management would like to avoid. The flight standards people decided that the best way to reduce our exposure to the risk was to reduce the number of F15 landings we did by restricting when crews could do them. They issued a FSM revision that made F35 the standard configuration and permitted F15 landings only when flown by the Captain using the Heads-Up Guidance System. Now, many first officers, some of whom had been flying the Megawhacker since we first got them, took this as an insult to their abilities. The company countered that they merely wanted crews to use the HGS, since it provides tailstrike protection in its symbology, and this is on the captain's side only. But the revision also restricted F15 landings to Cat III approaches during less than Cat I weather, or when performance required it - in other words, under marginal conditions only. Yes, they were reducing the number of F15 landings, but they were ensuring the remainder were flown by pilots who might've not done one in months.

In the wake of our 16C overrun incident, the flight standards people realized that the revision had unwittingly prohibited Captains from practicing Cat III approaches for proficiency. So now they're planning on revising the FSM once again, to require three Cat III practice approaches every 90 days! It is this sort of reactive micromanagement that makes line pilots disdainful of the "bean counters." If a line pilot were writing this section of the FSM, it would say:
The Captain will use discretion in selecting landing configuration to be used, taking into account weather conditions, aircraft weight, performance requirements, currency, and any other factor he or she believes could affect the safety of landing in any given configuration.

That's it - easy to use and it directs the Captain to be conservative while acknowledging that the Captain is best suited to choose the safest course of action. Of course, it would give the company lawyers less specific things to accuse the Captain of when they're trying to distance the company from his actions after an accident.

I fully recognize the benefits of having flight standards to follow and of having people dedicated to managing our risk exposure. I just wished they realized that micromanaging the cockpit has its own negative consequences, that Captain's authority is a positive thing that should be reinforced, and that asking the pilots who fly the line every day before making revisions would result in much better procedures that don't attempt to do a Captain's job for him.

Monday, May 14, 2007

There But For the Grace of God...

I've been waiting to get more of the story before writing this post, since it's a bit of a sensitive subject around here and one that the company would probably rather I not blog about. I think there are some related safety concepts that are worthwhile to discuss, though. Now that I've heard the Captain's side of the story, I'll relate some thoughts on my company's overrun incident on Rwy 16C at Seattle a few months ago.

In the last post, Fred commented that pilots seem to have an "inordinate attraction to analyzing accidents and incidents." He's right. There's probably a bit of a voyeuristic attraction and there's occasionally some armchair quarterbacking involved, but most pilots regard accident analysis as an integral part of their professional education. For me, the greatest value lies not in picking out the crew's major mistakes but in identifying the familiar elements - the situations I can see being in or the actions I could see myself taking. It's the perfect antidote for invulnerability and complacency. In this particular case, it's easy to see that a crew ran a near-STOL airplane off the end of a 9426 foot runway and dismiss it as nothing more than a monumental screwup. Closer inspection shows just how it could happen to you or I.

The incident took place on a foggy Seattle morning with Category III ILS approaches in use. All Cat III approaches at my company are hand-flown by the Captain using the Heads-up Guidance System. The runway visual range (RVR) was hovering near the approach minimum of 600 feet. In fact, the RVR at the far end of the runway, which is not controlling, was lower than 600. The crew planned on rolling out to Papa taxiway [TAXI CHART], which is significantly further down the runway than where we typically exit under normal conditions. Due to the poor visibility at the roll-out end of the runway, the crew missed Papa and subsequently decided to exit at the last taxiway at the runway's end, Quebec. They didn't see this taxiway until the last moment and only then realized they were still doing about 50 knots. There was nothing to do but throw the props into reverse and stand on the brakes. The airplane ran over a couple of threshold lights before coming to a stop, but nobody was injured. The airplane did suffer significant damage, and the incident was a major embarrassment to the crew, the company, and the FAA.

I'm sure many of you are wondering what in the world they were doing at 50 knots in such thick fog. What few realize is just how few visual cues there are use in estimating speed in this kind of visibility. Both of the crewmembers were peering outside for the first sign of taxiway Q, not at their instrument indications. They had no idea they were still going so fast. One major psychological factor was having traffic not too far behind them on the ILS. Missing taxiway Papa likely exacerbated the situation by hurrying the crew to get off on Quebec rather than slow down to find it.

Those who know anything about the Megawhacker will likely question why the crew planned to exit most of the way down a 9426 foot runway when we normally exit in half the distance without trying. This is where it gets interesting.

At the time of the incident, Seattle's SMGCS program was in effect, prompting crews to use the correct Low Visibility Taxi Chart. The chart in use showed taxiways P and Q as the only "low visibility taxi routes" exiting Runway 16C. For years, the FAA had maintained - and the company trained - that these routes are mandatory, so that when low visibility operations are in effect, one must roll all the way to P or Q. In the Megawhacker, this means adding power after touchdown to keep up your speed. In the aftermath of this incident, the FAA is saying that the crew was misinformed and it's actually permissible to exit on any lighted taxiway! Mind you, this is not what the FAA or the company taught in the past. I know Captains that were given demerits in the sim for exiting on the "wrong" taxiway after a Cat III approach. To be sure, the crew was at fault for hurtling past the end of the runway, but in my opinion it was the FAA that put them there in the first place. I have yet to hear an explanation of just what exactly a low visibility taxi route constitutes.

This is what is known as a "system error" - where poorly thought out regulations, procedures, or training leads to or contributes to an accident or incident. The classic systems error accident was the American DC-10 that crashed on takeoff after shedding an engine that was installed using a "money-saving" procedure that turned out to unapproved by McDonnell-Douglas. The cause-and-effect relationship isn't nearly so direct here, but I have no doubt that the system error was a contributing factor. For me, the lesson to be learned here not a new one: Procedures are not the end-all and be-all of piloting. They are set down by humans and are therefore fallible; they can even lead you down the wrong path. I think Earnest K. Gann said it best: "Rule Books are paper. They will not cushion a sudden meeting of stone and metal."

When you look at the factors that came together to cause this incident, you can see that while responsibility ultimately lies with the Captain, this is something that could've easily happened to any of us - there but for the grace of God go I. Cat III approaches at Seattle are a pretty common occurance for me, as is pressure to get off the runway for the next guy. Although this time nobody was hurt, the incident will follow the pilots through their careers. It should provide ample motivation to read the regs and procedures with a critical eye for well-intentioned dictates that reduce the margin of safety, become aware of the built-in "gotchas," cry foul when necessary, and exercise Captain's authority when in the left seat.

Wednesday, May 09, 2007

Back to School

I just got done with three days at the schoolhouse for my annual recurrent training. Yes, I'm aware that it's a little strange to do ground school after my trip to the sim. I certainly wouldn't plan it that way, but the schedules often conflict and force the company to do it backwards. And really, rather little of the information presented in ground school would help in the sim anyways. Even when an attempt at a good honest systems review is made, there simply isn't enough time to really bring you up to speed. You need to study on your own before going to the sim.

Each day started at 8am and lasted until 5pm, with a one hour lunch break plus lots of coffee breaks to keep everyone awake. Caffeine consumption is a major part of airline life under normal circumstances; during ground school, it doubles. I normally won't drink the coffee on the airplanes (I'm a bean grinding, french pressing coffee snob) but will mine the grindy dregs of the breakroom pot in an effort to keep from falling asleep in class.

It's not that all the information is worthless, there are definitely some gems that get passed along each year. There's just a whole lot of useless crap that goes with it, most of it mandated by the FAA. Systems review actually makes up a pretty small portion of the material, and that is so crammed that little new knowledge is gained. I will say that the part-time instructors (part-time line pilots) and the full-time instructors who have time on the line are far superior to the instructors who've never flown the airplane. We have one instructor who was paralysed in a motorcycle accident and no longer flies, but is particularly excellent at relating the systems to our actual line operations. He makes the Megawhacker electrical system a joy to learn, and that's saying something.

One thing that's sure to be discussed at any recurrent ground school is any accident or incident that happened during the preceding year. In this case, the big one was an incident we had a few months ago at Seattle; one of our Megawhackers ran off the end of runway 16C after a Cat III approach, for a variety of reasons that I'll save for another post. The interesting part is that the captain involved was actually sitting in this class, which made discussion of the incident rather awkward. He eventually did share his perspective on the event; it was rather gut-wrenching. I've flown with him several times and found him to be a good captain and a nice guy. Although the captain is ultimately responsible for his actions, there were some training and checking issues on the part of the company and FAA that I think strongly contributed to the outcome. Like I said, I'll expound on that in a future post.

We also had several bigwigs throughout the flight standards and training departments in to talk to us about what's coming down the pike. Our manager of turboprop flight standards was there to inform us that in response to the overrun incident, the company is drafting a bulletin that requires all pilots to adopt the landing technique of slowing to 50 kts immediately upon touchdown and then taxiing to the appropriate exit at that speed, no matter how long the runway or how nice the weather. I think a lot of pilots are going to have trouble with that one - not so much that it's a horrible idea as that it's the continuation of a trend in micromanagement we've been seeing over the last several years. You'd be shocked at the detail our flight standards manual goes into on how exactly they want us to land the airplane in different situations. Late last year they prohibited making Flaps 15 landings except during Cat III operations or if performance requires it, and only the captain could land Flaps 15. That caused an uproar among the pilots, first officers especially. I'll actually address this in another post in the near future, but suffice it to say that the pilots made their views known to the manager of turboprop flight standards. He didn't care, of course. That's okay - seeing him dodge pointed questions was more entertaining than usual Recurrent Ground School fare.

Anyways, it's all over now. I have two days of reserve to finish out the week, and then it's the new bid, in which I have a regular line. I had an easy time of it during my two bids on reserve - too easy, in fact. I only flew 6 days this month. My per-diem check is miniscule, and Dawn's sick of me moping around the house all week, so I'm putting myself back to work. I have layovers at a few places that are brand new for Megawhacker pilots, so that should provide yet more fodder for future posts.

Friday, May 04, 2007

Still Employed

Two constants of an airline pilot's career are training and checking. You undergo training when you're first hired, when you upgrade, any time you transition to different equipment, plus the occasional refresher. And then, you're always being checked to make sure you paid attention in training and haven't forgotten the material since. Most pilots that I know prefer training to checking. With training you get the pleasure of learning something new that might prove useful on the line, and there's not quite the same sense of your career being in jeopardy that comes with a proficiency check.

I've never failed a checkride in my life, and I still don't like them. I know that if I fail, the retraining and rechecking is going to be a pain, plus there'd be the stigma and self-doubt that accompanies a failed checkride, so there's a fair amount of pressure. The added stress actually improves my performance, I think, but it doesn't make me like the experience any better. Fortunately, first officers only have to take proficiency checks once every two years. It'd normally be every year, but the FAA lets us substitute training for pro checks every other checkride. The actual flying involved is actually much harder than a procheck (see my last example), but the atmosphere is much more relaxed and the emphasis is on learning.

Shortly before my recent training-in-lieu session, the company issued a bulletin titled "Training-in-lieu Completion Standards," which made it clear that TIL was no longer the non-jeopardy event it had traditionally been:
"Although a Training in Lieu (T/I/L) event is not considered a formal evaluation, the completion standards for the event are the same as for a Proficiency Check. A pilot must demonstrate the same level of performance in each maneuver that is expected on a Proficiency Check in order to complete the session with a MEETS
STANDARDS grade...At the end of the T/I/L session, each pilot must have demonstrated the same level of performance/judgment required on a Proficiency Check."
The bulletin caused a stir among the pilot group, especially those of us scheduled for TIL in the near future. What did it mean? Would the instructors really be holding us to ATP standards on outlandish scenarios like the ones we'd faced on previous training sessions? And why the change?

The backstory leaked out. Apparently a crew had showed up for a training-in-lieu session completely unprepared - no studying, flying skills rusty. Because training-in-lieu had a reputation as a "free pass," they essentially blew it off. The instructor was incensed and graded them "Does Not Meet Standards;" the crew protested that it was training, not checking, and therefore they couldn't fail. The company issued the bulletin to clarify the matter and to warn crews to take the TIL sessions seriously. Really, most of us already were; you don't want to make yourself look like an incompetent boob in the sim even if your career isn't on the line. It only took one lazy crew to screw things up for the rest of us - and I'd be one of the first guinea pigs for the "new" training-in-lieu.

It turns out that I had nothing to worry about. The session was just like my last TIL, except the scenarios weren't quite as outlandishly complicated as last time. The flying was a little harder than your standard proficiency check, but the instructor made it clear that we weren't expected to fly perfect on the advanced scenarios; these really were for the purpose of learning. The captain and I both flew fine and I came out of it feeling like I'd added something to my proverbial bag of tricks. With another pass, I'm still employed for another year. Yay! :)