The bottom line
In a word, money. Airlines' single largest cost is staff wages, which consume half of a typical airline's annual revenue—far more than fuel, aircraft leasing or servicing costs. Pilots are the most costly airline workers to employ. At a typical airline, industry insiders say,
97 of the 100 most highly paid employees will be pilots. #'s 98, 99, and 100 however make more than all the pilots combined.
Pilots are well rewarded, and with good reason: people trust them with their lives. Captains on long-haul flights are the most highly paid. According to Aviation Information Resources, a pilot-recruitment firm, the most senior captains earn an average of $178,000 a year. Some airlines pay far more: the most generous is Delta, which pays its top pilots $248,000.
dated?
There are other costs too, including per-diem and overnight expenses. Pilots expect to stay in fancy hotels when stopping overnight between flights. And why not? Before they take hundreds of lives in their hands, pilots ought to have had a good night's sleep. Safety considerations, which preclude pilots working too many consecutive hours, impose other costs. Airlines maintain elaborate rostering systems to ensure that there are enough pilots ready for duty, since each one can only work a few days a week. When a flight is delayed, the pilot may find that proceeding with the flight will require him to work more than the permitted number of hours; so a new pilot must be found, or the flight cancelled.
Automating flights completely would, of course, do away with these costs. Auto-pilots are amenable types: they never get tired, demand pay rises, go on strike, or
drink too much.
(Nice touch there) It would require lots of expensive new systems, but if it could be made to work, the cost savings would be enormous: billions of dollars a year for a large airline. Even partial automation, perhaps involving shifts of pilots looking after several semi-autonomous aircraft at once via remote control, would save money. Again, Global Hawk points the way: a single operator can control two or more aircraft from the same command console on the ground. And research by Boeing, which is building an unmanned fighter aircraft for the American air force, suggests that operators could efficiently control up to four aircraft at a time.
None of this is likely to happen any time soon, of course. But the technological trend is clear. Craig Mundie, chief technical officer at Microsoft, the world's largest software firm, estimates that passengers will routinely travel in pilotless planes by 2030, a claim he has backed up with a public bet to that effect (see
www.longbets.org/bet/4). For his part, Mr Mitchell believes it is unlikely to be before 2050, but that it will happen eventually.
In the meantime, there may be nearer-term opportunities for unpiloted planes. Northrop Grumman is examining commercial uses for unpiloted aircraft including surveillance, communications—a drone aircraft flying over a city could act as a low-altitude communications satellite—and even crop-dusting. In November 2001, Boeing set up a dedicated unmanned-systems unit, which will initially concentrate on military aircraft, but with the aim of transferring the technology into civil and commercial aviation.
Assuming they can prove themselves in small commercial applications, the next big step would be to use unmanned aircraft for freight.
Federal Express has discussed this possibility with Mr Mitchell. If unpiloted freighters then became commonplace, the way would be clear to consider passenger flights. And at that point, advocates of unpiloted planes would face their greatest challenge: convincing the public to get on board.
That could be a tall order. For while it would be possible to make a case for pilotless planes on safety grounds, this would mean pointing out some of the hazards associated with piloted flights. No aircraft maker or airline is going to do that.
Yet more than half of air-travel deaths are the result of “controlled flight into terrain” (CFIT), which is industry-speak for “the plane was working perfectly, but the pilot flew it into the ground”. CFIT accidents cause the most casualties because few passengers survive them. People are far more likely to survive if a plane overshoots a runway, or its landing-gear collapses, than if it flies into a mountain. Of the 18 fatal air accidents in the first half of 2002, nine are thought to have been due to CFIT, resulting in 397 deaths.
Other forms of pilot error can also be fatal. On July 1st 2002, two planes collided over Germany, killing 71 people. The airborne collision-avoidance system worked perfectly, but human error, both in the air and on the ground, meant the collision happened anyway. It would be foolish to suggest that unpiloted planes would never crash or go wrong. But a few years of data from unpiloted freight flights would make meaningful comparisons possible, and might favour more automation.
Another problem with advocating unpiloted aircraft as a means of reducing pilot error, however, is that it would mean drawing attention to the realities of air safety—a subject the industry seems keen to avoid, with the tacit approval of passengers. No large airliner has ever made an emergency landing on water, for example. The engines, which hang in “pods” beneath the wings, make such a landing almost impossible, notes Mr Jackson. So the life jackets, with their little whistles and lights that come on when in contact with water, have little purpose other than to make passengers feel better. Yet airlines and passengers seem happy to pretend that there is some point in having them.
Similarly, there is abundant evidence that rear-facing seats are safer than forward-facing ones in the event of an emergency landing. But any airline that made all its seats rear-facing would probably go out of business, because most people prefer facing forwards. Only soldiers, who have no choice, routinely travel in aircraft with rear-facing seats. The rest of us, it seems, put comfort above safety, or would rather not be constantly reminded of the possibility of an emergency landing. (the Economist)
