15 July 2005

Commercial Air Travel and Security

The commercial air travel market in the United States is a kludge. It gets millions of people from point A to point B on medium to long range trips faster than driving. It has a remarkable safety record. It is still possible for middle class families to afford to travel by air. It achieves this without and at times, in spite of, sensible planning.

Airline Economics

It has been said that the commercial airline industry as a whole has not made a profit, in the aggregate, since it came into being in the 1960s. Certainly, many airlines have gone bankrupt along the way, and no industry in the United States is currently more fragile. Almost all of the "legacy airlines", United, Delta, Northwest, U.S. Air, and American have either gone bankrupt or are in serious financial trouble. A few "discount airlines", Frontier, Southwest, and JetBlue, for example, seem to be holding their own with a different business model. But, the industry's future is uncertain.

It appears the a big part of the problem is in the pricing structure of the legacy airlines, which have fed overly high operating costs. Legacy airlines have relied on monopoly pricing at airports they control and ridiculously high fares for business travelers who must travel during the week and often on short notice, rather than on more budget conscious, pre-planed vacation trips. Another part of the problem may be that airlines generally are growing less competitive vis alternatives, like teleconferencing and driving, as prices have remained high, while the effective speed of the trip, in terms of miles per hour from point of origination to destination, has declined.

Another issue is whether the old hub and spoke system continues to make sense. Direct flights are considerably faster than a hub and spoke flight. Yet, in a market with dozens of airlines, we need a hub and spoke system so that each airline can aggregate enough passengers to make a flight worthwhile. But, with the advent of regional jets, and if airlines were permitted to coordinate so that only a single airline served routes that could support direct flights with regional jets if consolidated, but not if divided amongst many airlines (with accompanying price regulation to prevent abuse of monopolies on routes), many more cities could be connected by direct flights, greatly reducing travel time for many passengers.

For example, on a flight from Denver to Northwest Airlines hub Minneapolis that I was on recently, there were enough passengers heading to both Washington DC and to Milwaukee to support separate direct flights. But, in all likelihood because of the need to keep the flight to the hub full, which it would not be full if direct flights to those destinations existed, about half the people on the plane had to endure time consuming transfers and a significant deteur from their ultimate destinations.

And, if their were limits on monopoly pricing, I would have taken an existing direct route to my destination, rather than an indirect route to a more distant airport at half the price.

Travel Times

One of the ironies of commercial air travel is that the flying part has become a relatively small part of the process. On my most recent trip back from my father's house, I spent eleven hours in transit, only four of which were actually in flight. The experience is not atypical. (If I had taken a direct flight at twice the price, the time in the air would have been only about two and a half hours, but non-flying time would have been cut by only about an hour).

If I want to fly somewhere, it takes me about 45 minutes to drive to the airport (from central Denver), longer during rush hour, and I need to add an extra half hour if I want to use cheaper shuttle parking at the airport at $5 a day, rather than the $7 a day parking immediately adjacent to the airport (which is sometimes full). The current norm, particularly if you are travelling with children, is to arrive at the airport itself about two hours in advance, to allow you to get your boarding passes, check your baggage, make it through a security check point (check point delays are discussed here), take a light rail shuttle to my terminal and arrive their at least a half an hour prior to your flight for boarding.

Delays for mechanical problems, weather, and who knows what else, are not uncommon. If the flight is not a direct flight, a layover can last anywhere from half an hour to several hours, and there is a good chance that your luggage won't make it if the gap is short.

Upon arrival, it typically takes at least half an hour to collect your baggage from the baggage claim (not infrequently damaged), another half an hour to get ground transportation (be it picking up your car upon your return home or arranging for a rental car), and another half hour to an hour and a half to get to your final destination on the ground.

Thus, a typical flight takes about five hours in addition to any time spent actually flying, if all the stars are aligned in your favor. The actual time spent in the air within the continental United States ranges from one to five hours for a typical flight. A flight with a connnection (typically as part of an airline's hub system) will typically almost double the time spent from the first takeoff to the final landing. In other words, trimming a couple of hours off boarding and departing times saves as much total trip time as running the Concorde at supersonic speeds on every route.

By comparison, when my I was in law school, in Ann Arbor, Michigan, my wife lived in Chicago and we frequently took Amtrak to visit each other. Amtrak is hardly the model of efficiency either. It almost always takes longer than travelling by car (outside essentially one route in the Northeast Corridor, which not coincidentally is also almost the only route in the system that breaks even economically vis operating costs), and while it sometime is faster than a Greyhound bus, it is often slower (albeit, more comfortable), in part because the train travels slowly and in part because it makes so many stops. But, the collateral time involved in an Amtrack trip in absolutely trivial by comparison to a flight (which is why rail proponents emphasize the potential of high speed rail for medium length trips). The station at each end of our trips was about fifteen minutes from the point of origin and the final destination, respectively. You need to arrive at the station about ten or fifteen minutes before your secured departure, and your baggage is unloaded in about ten minutes -- and our baggage was never damaged on any of our frequent trips. The total time spent off the train at both ends of the trip combined was less than an hour. There are no security checkpoints. The baggage handling system is the equivalent of what is know in the airline industry as a "gate check". The fact that passenger trains are quieter than airplanes and involve less infrastructure allowed stations to be located closer to population centers.

Security and Risk Analysis

Amtrak probably has insufficient security. Certainly, a London Underground or Madrid style bombing could happen on Amtrak (or just about any commercial bus or passenger train, intra or intercity) in essentially the same way, while an event of that kind would be much more difficult to orchestrate on an commercial flight. Whether it makes sense to devote the effort to airline style security, however, is another issue. The other way to bedevil a commercial train is to cause it to derail at any point along the hundreds of miles of track it must follow. The example of a Los Angeles man who left his car on the tracks in what was probably an aborted suicide attempt, causing dozens of deaths in a derailment, shows just how easy it would be to cause an accident like that intentionally. Why add half an hour of delay or more per trip to avoid one fairly sophisticated problem (a passenger based bombing of a train), when there is no practical way to address the greater threat of a non-suicide bomber with far less sophisticated means interfering with train tracks to cause a deadly derailment? Also, on Amtrak, checked baggage is on a different railroad car than passengers, so a heavy explosive device in checked baggage would probably do far less damage than it would on a plane -- so, even if Amtrak instituted searches of carry on baggage, it isn't clear that searches of checked baggage would be that valuable.

But, the security surrounding air travel is probably overkill in many important areas (while remaining insufficient in others). Are we really meaningfully safer because we now prevent non-airline passengers who have gone through a security checkpoint from escorting passengers to and from the gates where passengers board and deplane? Are we really better off because we have zero tolerance for small pocket knives? Does it make sense to treat unattended carry on baggage as a menace when all carry on baggage has already passed through a security check point?

Not every abandoned security measure reduces safety. How many people think we are really less safe because we no longer ask passengers when checking in if the are carrying a bomb?

The biggest hole in the current regime design to prevent hijackings and bombings of planes is that not all checked baggage is searched or otherwise scanned for threats. Checked baggage can be larger than carry on baggage, which is uniformly searched, and a cell phone or even a key fob designed to remotely unlock a car would be perfectly capable as functioning as a trigger in the hands of a suicide bomber.

Yet, according to yesterday's U.S.A Today, the government is seriously considering spending $1 million on each of the 6,800 commercial U.S. planes to mount an anti-heat seeking missile laser turret to repeal shoulder mounted missiels, a total cost of $6 billion or more, to deal with a threat that has never manifested itself in the United States, although backers claim that such missiles have been used against civilian planes 35 times in history, killing about 500 people.

How do you weigh a risk like that? On one hand, $6 billion seems like a good deal compared to the several hundred billion dollars a year we spend on "national defense." It is far more likely that a shoulder fired heat seeking missile will be used against a U.S. commercial aircraft than it is that the U.S. will find it prudent to launch an all out nuclear missile attack on Russia, a fool's errand to which we have devoted far more of our nation's tax dollars.

Indeed, the risk may even compare favorably to the expense we devote to preparing every commercial aircraft for water landings. It appears that commercial airplanes have "ditched" in water just three times in recorded history. Once in the time period from 1957-1979. In the second:

an Ethiopian Airline Boeing 767 had to ditch off the coast of the Comoros Island in Eastern Africa when it was hijacked and it consequently ran out of fuel. The ditching sequence was probably not well executed, possibly because of interference by the hijackers. It cart wheeled and broke into a few sections. Despite this blotched up ditching, 48 out of the 175 passengers survived. The life vests they were wearing certainly saved their lives.

In the third:

On the 16th of January 2002, a Boeing 737-300 belonging to an Indonesian Airline had both its engines flamed out - a term to describe that the jet engines had failed. It happened as it commenced its descend to 9000 feet through thunderous clouds that were filled with rain.

The crew then tried to relight the engines but it failed to revive. . . . When the engine failed, the Captain maneuvered the airplane so that it could glide at an optimum speed of around 240 knots. This would cause the airplane to lose height rapidly at about 3000 feet per minute. He then attempted to make a forced landing, but preferred to ditch into water if only he could locate the sea. As the sea was out of reach, he decided to ditch on a river instead.

During the forced landing process, the Captain tried to decelerate from 240 to 150 knots by use of the flaps, but the hydraulics were not available to power the action. . . . Luckily, the ditching was very well executed and the Boeing 737 came to a stop, floating near the side of the river. . . . In this accident, 23 people were injured in the plane carrying 54 passengers and a crew of 6. One stewardess died when she was drowned in the river.

Note that in the Ethiopian Airline incident, doors that prevent entry to the pilot's compartment (implimented in the U.S. fleet after 9-11) might have been equally effective. In the Indonesian Airline incident, some form of crash landing on land (admittedly possibly causing more injuries) might have been possible.

But, consider the huge amount of effort (elaborate warnings at the start of every commercial flight), and expense (fitting every seat with floatation devices and most doors on commercial planes with life rafts) that has gone into preventing this remote risk. Would the money have been better spent on better instruments at airports for bad weather and night landings, or short distance radars to prevent planes from colliding on runways in fog and darkness? Would we be better off developing a plane sized parachute for commercial airliners (something which already exists for at least one model of general aviation plane) than anti-missile lasers or water landing gear?

The Baggage Problem

The biggest problem overall with the commercial air travel process is how it handles baggage. Baggage accounts for a large part of the delay in air travel.

Checking in baggage, and retrieving baggage accounts for a large part of the pre-flight and post-flight delay in a typical air trip, which is why business travelers typically do all that they can to take only carry on luggage (in addition to their desire to protect their luggage from the vageries of baggage handling processes resulting in lost or damaged luggage on a regular basis). Checked bags, as previously noted, are also a security risk, because it is hard to adequately screen them all.

Carry on baggage poses its own problems. It accounts for much of the time required to get passengers boarded and deplaned. It accounts for much of the delay in getting people through security check points (compare the delay at an airport security check point to that at a court house where people must pass through metal detectors but are less likely to carry lots of large parcels). It leaves passengers with baggage under their seats cramped and uncomfortable. Baggage falling from overhead compartments cause a significant number of work related injuries for flight crews and passenger injuries, although most are minor. And, security efforts notwithstanding, weapons and other dangerous articles do manage to slip through the screening process and get onto planes from time to time (as audit checks of security measures prove on a regular basis).

In contrast, getting people onto planes is a relatively easy process. Walking someone through a metal detector can be fairly quick. Boarding passes can be dispensed by kiosks or printed out at home on a computer. Matching photo IDs to boarding passes once at a security checkpoint takes just a moment. As aircraft evacuation tests have shown, people can get on and off a plane quite quickly if they don't take their baggage with them.

The Baggage Solution

To solve both the security and delay problems associated with baggage, it may be necessary to fundamentally rethink the whole idea of luggage in air travel.

One approach would be to "gate check" all luggage. Instead of limiting carry on luggage, everyone would take their luggage all the way to the gate, which is the way many small airports operate. But, it isn't clear that this would scale well.

Another approach would be to remove most luggage from the security appartus all together.

For example, suppose that every commercial flight had two planes, one with just a pilot and co-pilot to carry the baggage (or perhaps even just a single pilot and a remote operating system for emergencies), and another to carry only the passengers, who would not be permitted to carry large carry on items. The baggage plane might leave fifteen or twenty minutes ahead of the passenger plane, so that the baggage could be waiting for passengers at the gate when the passenger plane arrived. Baggage going on the baggage plane would not require any security screening. Parachutes and/or ejections seats for the two crew members on the baggage plane would be cheaper than an elaborate screening process for every bag, and destroying a plane full of clothing has far less terror causing appeal than blowing up a plane full of passengers.

Baggage planes could be adapted to minimize baggage handling. Now, luggage is (1) accepted and in some cases security screened, (2) sorted, (3) put on carts, (4) rolled in the carts to the plane, (5) put in a luggage hold under the passenger compartment, (6) unloaded from the plane onto carts and then (7) unloaded from carts onto a carosel at the baggage claim area. Instead, with specialized planes, passengers luggage could be (1) put onto carts at check in, (2) rolled cart and all onto the baggage plane, and (3) rolled off the baggage plane into a baggage claim area where arriving passengers would pick up their baggage directly from the cart. Baggage going on connnecting flights could be placed on different carts than baggage going only from point to point. The luggage would be handled less often, so there would be less risk of damage or loss. Now, there is waiting time for the baggage to be unloaded, then the baggage would be unloaded when the passenger plane arrives. Now, each piece of luggage would need to be screeened, but under this plan, none of it would be (which would also faciliate courier service and a decentalized luggage pickup system where some passengers would drop off luggage, for example, at a rail station to the airport, rather than at the airport itself). The need to match bags to passengers, which makes the baggage process far more complex, would also end.

The fuel requirements of the two plane might not be that different from that of the single larger plane that carried both passengers and baggage. The savings on the security and delay end of the operation might make up for the cost of two more crew per trip. One would have to run the numbers, but that kind of rethinking, could make us both safer and make the whole air travel process much faster on a point to point basis.

One could also imagine redesigned passenger only planes, and new airport gates to accompany them, being designed to routinely board from more than one door, to improve turnaround time at the gate. If the boarding process could be cut from say, 35 minutes to 15 minutes per flight, due to a reduced volume of carry on luggage and the use of more than one door to enter the plane, this would add up. If the new system saved passengers an hour and a half in the airport per trip, and the savings from the security system paid for the increased costs of operating two planes per flight, the effect would be equivalent to increasing the speed of the average flight from about 500 mph to Mach 3.

1 comment:

Andrew Oh-Willeke said...

It appears that a commercial flight from Tunisia to Italy made an emergency water landing yesterday (i.e 8-6-2005). About 19 of the 39 people on board lived.