Reflections On An Intimate Working Relationship

So I'm suddenly down to my final few legs on the DC-8. The phase-out of an aircraft fleet must be accomplished gradually, with the aircraft getting mothballed in the desert a few at a time. The first step of this process has been to reduce the work assigned to the fleet, leaving us a bit fat on crewmembers relative to the actual staffing needs. And thus many of us who used to fly set schedules have instead been assigned what is called a reserve schedule. Usually this means sitting for your designated days in an apartment or, if you live in domicile, at your home waiting for the phone to ring with an assignment (technically, you can sit reserve wherever you like, provided you can be at your assignment within 90 minutes of the phone call). But with the reductions of flying on the fleet, the phone doesn't ring very often.

There are far worse things, of course, than to be paid to not work. But one of the few negatives in this situation is that one becomes a bit rusty. In some ways, flying knowledge is a bit like memorizing your friends' phone numbers. When you use them all the time, they exist in some kind of half-long-term / half-muscle memory state; they're readily at your fingertips, but not actually that firmly lodged in memory. Flying involves reams of little facts and procedures and rules, details that begin to fade with just a little inactivity. I've been flying very little now for six or eight months (which is not to say I haven't been away from home quite a bit, sitting in my crash pad in Kentucky), and I can attest to this rustiness business firsthand. But now, as I begin to shift my focus to a new airplane, I'm spending my final week on the DC-8 with a Montreal assignment. And these previous months of relative inactivity enable me to look back on the DC-8 with a bit of freshness.

(Airplanes in mothballs in Roswell, NM in 2006. These are preserved to fly again if necessary.)

Flying an airplane is a unique occupation. It's machinery operation at its core, with a bunch of practical knowledge things tacked on--rules and regulations and airport operating procedures and protocols and company procedures and so on. Maybe the closest analog most of us have is driving a car, but imagine something about five times as complex and then imagine training intensively to do this task with a high degree of proficiency and then doing it all the time--for a daily living, and not just as an afterthought to get to something more important. So a pilot develops a relationship with the machine to a degree that's odd in most other professions. This is partly what makes pilots as a group such machinery geeks (chicken or egg, you decide).

And that contributes to this time being a bit poignant for me. Not only am I ending this relationship of nearly seven years, but there's the added angle that how we do things on an old airplane like the DC-8 is different from a newer airplane, and these old ways are just about gone from the industry. A good analogy here (I'm sure I've made it before) is the rapid influx of diesel technology in railroads in the 40s-50s. Relatively quickly, the steam technology which played such a huge role in America's industrialization and its push to modernity, and which provided good, solid careers for untold thousands of workers, went the way of the dodo. For every railroad in the country, over a pretty short span, a crew brought the last steam locomotive into the shops and dropped the fire for the last time before the equipment was towed away and scrapped. It was a paradigm shift.

(The DC-8's three-person cockpit.)

On the DC-8, one of these things is the flight engineer. All large airplanes used to have at least three people in the cockpit, and the early jets were all this way: the Boeing 707 and 727 and 747, the Lockheed L-1011, and the Douglas DC-8 and DC-10 (among others). The two pilots cope with the flying duties, and the flight engineer runs the airplane mechanically--fuel and pressurization and electrics and hydraulics and heating and cooling, etc. Additionally, the engineer is tasked with keeping an eye on the front-seaters to catch errors and assist with radio calls and checklist duties, etc.

At my own company, the DC-8 is the last three-person cockpit on the property. When I came here in 2001, we had three different fleets on the property that needed flight engineers: the 727, the classic 747 (we've since added newer 747-400s, which are a two-crew airplane) and the DC-8. The 8s are all that are left, and now the end is nigh for these. I began my career here with three and a half years as a DC-8 flight engineer (we referred to the task as "plumbing"), plus a short stint plumbing on the 727. So I learned firsthand of the old school at the very tail end before it vanished. (I'm actually quite pleased to have a flight engineer's certificate--it's a separate license from a flying license--and, though I hated it at the time, having engineered on two different airplanes by different manufacturers gave me better insight to the philosophy behind the job.)

Checklists are a good example. On any modern two-pilot airplane, the checklists while the aircraft is on the ground are generally called for by the captain and read by the first officer. In flight, the checklists are called for by whichever pilot is flying and they are read by the non-flying pilot. Checklist protocol--the call-and-response--is different on a three-person airplane: the responses must come from three people instead of two. Like the two-person airplane, some checklists are called for by the captain, others by the pilot actually flying the leg; but most of the checklists are actually read by the engineer, and space must be given for a third person to speak their piece and accomplish their tasks (there's generally a lot more to do on an old airplane). This three-way interaction is unique, and it always takes people new to the airplane a while to adjust their routines to accommodate the extra person.

[Aside] The core idea here fascinates me, this business of figuring out what tasks need to be performed for a large airplane to fly and how the systems and cockpit are designed to accomplish these things. On one extreme, military airplanes used to fly with two pilots plus an engineer plus a navigator plus a radio operator and all manner of load specialists. On the other hand, through automation we could probably do away with pilots altogether, or certainly all but one (indeed, we have unmanned aircraft performing many military tasks already, though passenger flying is another matter). So what we end up with is a compromise. It's worth remembering that someone had to design this human / machine interface, that the tasks we perform and the displays we use and the controls we manipulate and switches we throw are the result of an intense engineering effort; these things could be quite different. I've read discussions about this with the early fly-by-wire Airbuses (the A-320 & 319); the airplanes were so automated that pilots could spend almost the entirety of a flight doing nothing but watching the airplane to make sure something didn't malfunction. A little research showed pretty quickly that in order for a pilot to respond effectively to a mishap or an emergency they needed to be kept engaged and in the loop. And so a workload was "built in" to the airplane's design, like a complex version of a railroad dead man switch. With railroads, the switch was just there to ensure the operator was not asleep or dead--the train would stop itself without intervention within a specified interval; in the airplane, the tasks are more designed to keep the pilot in the loop on several fronts: communication, navigation, aircraft mechanical status. (I can see my wife's eyes rolling so far back into her head at this little rhapsody that she's going to need surgery to see straight again. I raise the white flag: I'm hopeless.) [/Aside]

(The Mighty Mad Dog. I don't know what that screen-thingie is by the co-pilot's window. It's not one of ours.)

So that--the three-person cockpit--is one big difference about the DC-8. The other is automation. Virtually every modern airplane is flown through programming, and over the past two decades the pilots' tasks have become much more biased toward management of the automation (there are many good reasons for this, but it's another long post in itself). While our DC-8s have a good navigation box and a good autopilot--this is not a luxury, by the way, but a really essential part of flying; the airplane is seriously restricted in what it can do if the autopilot(s) is inoperative--it simply doesn't have the capability of a newer airplane. As one specific example, the DC-8 is almost unable to navigate vertically. It can follow a lateral course nicely over the ground (though not with the precision of a newer airplane), but it does not have the capability to specify that it wants to be at such-and-such a point at a specific altitude and / or airspeed (we can obey most such commands from ATC, of course, but it requires the pilot to figure out how to make it happen). In addition to the the navigation box needing to think in three dimensions--which ours will not--this feature requires the computer to have the ability to control the airplane's thrust: if the autopilot points the airplane's nose down for a descent and doesn't retard thrust, the airplane's speed will quickly build to a dangerous level (and, conversely, if the nose is pointed up for a climb and power is not added, the airspeed will quickly drop and the airplane will stall and fall from the sky. This is bad). This feature which allows the airplane's automation to add or subtract thrust is called the autothrottle system, and the DC-8 doesn't have it. The end result of this is that the DC-8 gets hand-flown quite a bit of the time, which is very rare in the airline world today.

I'm trying to savor all these things a bit in these final days, knowing that I'll never have the chance again. Friday night as we blasted off from Louisville and headed up to Montreal, I thought of all this as I looked at the DC-8's old round dials and heavy controls polished smooth from decades of use. Old seats and small windows and very heavy flight controls--all so different from a newer airplane. A few minutes earlier the captain had called for engine starts, which I accomplished, like a thousand times before, with the flight engineer calling out oil pressures and starting duct pressure readings over my shoulder, and me timing the start and reading back N1 and N2 rotations and announcing "fuel on... flow... lightoff." The captain coordinated with the mechanic over the headset and talked to ground control--and, of course, taxied the airplane while the engineer and I started the remaining engines (what a luxury to have four!). It's an interesting little ballet, a bit of controlled chaos that is specific to this airplane and this circumstance.



I'll have to start the engines in an MD-11 as well--three of them; still a luxury!--but it's a quiet, solitary, mostly automatic task. Push a button and make sure nothing screws up while the start sequence runs. No one looks over my shoulder and joins me in the effort. I welcome the move to a more modern airplane and all the amenities it offers, but I'll definitely miss the third crewmember in the cockpit. As a flight engineer myself I corrected enough mistakes by the front-seaters, and as a front-seater I've had enough of my mistakes caught and corrected by the engineer, that one wonders how we'll get along without the third set of eyes. I guess that's where that machine / human interface design business comes into play.