Fatigue in the Cockpit

July 27, 2006
The NTSB cited pilot fatigue as a factor in three recent major aircraft accidents.

At 8:36 p.m. I cross into airline pilot hell, which I call the Land of Naugahyde. It's a little room deep in the bowels of Newark airport, far below the corridors of the passenger terminal. But it's close enough to hear the banging and clanking sounds of all the activity above. The problem isn't the noise. It isn't the rows of simulated leather recliners lined up on both sides of the dark room, with a single harsh fluorescent light casting shadows, or the sticky sweaty smell that comes from lots of pilots having shared the same recliners, blankets and pillows while they attempted to rest here in the terminal. No, this is airline pilot hell: trying to make yourself fall asleep in the early evening when you're not tired, while knowing that in an hour or two you'll have to wake up and fly all night.

After an hour stuck to the recliner I give up and get up. I splash water on my face and pretend it's morning and I'm a regular person waking up for work. In reality, I'm going to push back from the gate just before midnight and land in South America just before dawn. By the time I shoot an instrument approach through the clouds, flare the airplane just above the runway and cut the power, I will have been awake almost 24 hours.

I know that tonight my brain is going to work slowly, my over-40 eyes will not be able to focus well in the dark, the two of us in the cockpit will make mistakes, and we're going to try all night to keep them from affecting the safety of our flight. It's an uncomfortable feeling to know all of this ahead of time, lying awake in the Land of Naugahyde. Luckily, I also know that I've flown this kind of red-eye flight many times before -- about once a month -- and it worked out fine.

The schedule tonight will be six hours from pushback to shutdown, about seven and a half hours on duty for our crew, including the time to plan and prepare for the flight. That does not even come close to pushing the regulatory limits for flight and duty time; many other schedules that airline pilots fly are much more conducive to fatigue -- but I don't fly any of those.

We lift off on time and head offshore over the Atlantic. Level at 36,000 feet, the autopilot is holding the course and altitude. The cockpit settles into quiet as the velvet night casts its spell. I tilt the radar down a couple of degrees. Are those radar reflections simply thunderstorms, or the Bahamas? They're the Bahamas, and we pass over in smooth air.

About three in the morning, I'm getting sleepy. I'm staying awake with coffee, slurping down the fifth cup of the flight. When the acid hits bottom, my stomach sends up the no-more-coffee signal. I drink some more.

Controlled naps

Several years ago, NASA researchers found that pilots did a much better job on the critical approach and landing parts of the flight if they alternated taking controlled 20-minute naps in the cockpit during the quieter, cruise portion of the flight. Despite the scientific evidence, the Federal Aviation Administration has not acknowledged or formally accepted this practice. If I took a nap on a flight with an FAA inspector riding behind me on the cockpit jump seat, I would expect to have my pilot's license lifted for "careless and reckless operation of an aircraft." More significantly, broader FAA action to update the rules about crew rest in accordance with current knowledge about fatigue has also been held up for years. Meanwhile, more than 20% of pilots' reports to the aviation industry's confidential safety reporting system raise the issue of fatigue, and the National Transportation Safety Board cited pilot fatigue as a factor in three recent major aircraft accidents. But throughout all the years of airline flying, pilots have taken matters into their own hands and taken naps during the cruise time on these all-night flights. We know how important the last few minutes of the flight are, and we want to have had a little rest before undertaking them.

Mind is barely ticking

Approaching the coast of Panama, the radar picks out some thunderstorms, and I curve us about 20 miles to the left to avoid them. The plan works, and the flight continues smoothly for the folks sleeping in back. I curve us back on the original course and all is routine again. Our jet's nose is pointed out over the Pacific now, and the view through the windshield is once again pitch black. We turn the cockpit lights way down and enjoy the constellations and the Milky Way arching overhead. Meanwhile, it's that time of night when I have to take my glasses off to read the small numbers on the chart. It's time to turn up the lights. And as we continue south, periodically reporting our position to the next country's air traffic controllers, my mind is barely ticking. There are navigation, weather and fuel concerns to think about, but it's almost easier to let them go. I recognize the mental fog of fatigue and try to push myself to think them through.

'Microsleeps'

Sleep researchers have hooked fatigued pilots up to EEG sensors while in flight. The pilots' brain waves showed they were experiencing "microsleeps" while flying the approach to landing -- that is, the pilots were repeatedly falling asleep for periods of less than a few seconds, without being aware of it. It's not inspiring to think about what else they might not have been aware of while microsleeping. This certainly suggests that the aviation system's regulators should approve the concept of controlled napping in the cockpit during these night flights.

Safe on earth

As we descend next to the Andes Mountains tonight, the coffee and adrenaline -- and the brief rest my eyes took en route -- all combine to keep me reasonably alert, and the approach and landing are fine.

As someone doing this job, I can tell you that these flights are safe enough, as safe as they can be given that the pilots are staying up all night. I would put my family aboard an all-night flight without hesitation, with or without the pilots being allowed to nap on the flight deck. But fatigue inevitably cuts down the margin of safety. It would be safer for passengers and crews if the FAA rules allowed a good idea such as controlled napping. Even more so, though not so much an issue on this flight, it would be safer if the FAA's established limits for pilots' duty time and rest time better reflected the mental and physical needs of humans operating on the back side of the 24-hour clock. I think about all this as the sun rises in Ecuador and we get to the hotel, eat breakfast, go to bed and try to catch up on sleep all during the day. As the sun sets again, my night flying is done, but this evening, pilots all over the world are crossing into the Land of Naugahyde before they have to fly all night.

Ben Berman is a U.S. airline pilot flying the Boeing 737 and a former chief of major investigations of the U.S. National Transportation Safety Board.

: Limitations for international flights FAA regulation 14 CFR 121.481 pertains to U.S. carriers on international flights. It's notable that this regulation never mentions the time of day at which flights occur.

(a) A certificate holder (airline) conducting flag (international) operations may schedule a pilot to fly in an airplane that has a crew of one or two pilots for eight hours or less during any 24 consecutive hours without a rest period during these eight hours.

(b) If a certificate holder conducting flag operations schedules a pilot to fly more than eight hours during any 24 consecutive hours, it shall give him an intervening rest period, at or before the end of eight scheduled hours of flight duty. This rest period must be at least twice the number of hours flown since the preceding rest period, but not less than eight hours. The certificate holder shall relieve that pilot of all duty with it during that rest period.

(c) Each pilot who has flown more than eight hours during 24 consecutive hours must be given at least 18 hours of rest before being assigned to any duty with the certificate holder.

(d) No pilot may fly more than 32 hours during any seven consecutive days, and each pilot must be relieved from all duty for at least 24 consecutive hours at least once during any seven consecutive days.

(e) No pilot may fly as a member of a crew more than 100 hours during any one calendar month.

(f) No pilot may fly as a member of a crew more than 1,000 hours during any 12-calendar-month period.

Pilots report incidents

According to NASA, airline pilots cite fatigue as a factor in roughly 1 in 5 incidents reported to the confidential Aviation Safety Reporting System (ASRS). Excerpts from pilots' reports:

*Dulles, Va., April 2004: "I got a stick shaker (flying too slowly) at 3,000 feet. ... Contributing factors: ... Fatigue: unable to sleep on ... rest breaks. Unable to nap on layover before the flight."

*San Francisco, March 2004: "The flight crew missed several radio calls, had difficulty recalling clearances issued just minutes before, and failed to make a cabin security announcement in a timely manner. ... Each of the crew members felt tired. I am writing this report to bring attention to the serious dangers of operating on the back side of the clock without enough time for proper rest."

*San Francisco, April 2004: "We crossed an active runway without clearance. Contributing factors ... fatigue -- the late night departure for a back side of the clock flight."

*Oslo, Norway, March 2004: "Fatigue was the primary factor leading to this event. The duty day started late night ... in Cologne, Germany. Both crew members had experienced multiple circadian rhythm disruptions during the previous few days contributing to cumulative fatigue. After flying an uneventful leg to Malmo, Sweden, the crew had to wait for about two hours (in the cockpit) before departing for Oslo, Norway. ... (Later in that leg) leveling at 3,000 feet (in preparation for Oslo landing) ... (air traffic control) requested an altitude check. We immediately realized we had not reset the altimeters. The aircraft was at 2,400 feet."

*Santiago, Dominican Republic, December 2003: "We had been on duty for 16 hours and were on the final phase of the approach. ... We got a ground proximity warning ... and climbed out immediately. ... If we were not so tired, we would have caught the minimum altitude. ... We came very close to being splattered all over the mountainside."

*San Francisco, January 2004: "We aborted takeoff at low speed due to a takeoff warning horn. We cleared the runway and then noticed the flaps were still in the up position. Apparently, I had failed to drop the flaps. ... Fatigue may have contributed to my performance, due to the fact that I had flown 93 hours in the last 30 days."

*Los Angeles, December 2003: "We were both fighting fatigue during the flight, consuming many cups of coffee and fluids. ... At this point, we had been on duty 11 hours 13 minutes ... in my exhausted, fatigued state, I went numb to situational awareness (after touchdown). Before I could realize what I had done, Runway 25R was approaching and the boundary ... had been crossed. ... If ever in my life I could rewind 10 seconds of my life and do it over again, it would be the few seconds I experienced yesterday."

Compiled by Ben Berman