Getting A Few Winks

Investigators learned that at 12:04 a.m., American Airlines flight 1012 was being handled by an approach controller at the TRACON, nearing the end of its flight from Dallas-Fort Worth. The flight was instructed to contact the DCA tower. The pilots tried to raise the tower several times as they got closer and closer to the airport. Unable to obtain a landing clearance, they elected to execute a missed approach. They then radioed approach that they had been unable to contact the tower controller, were on a missed approach, and wanted to land. The approach controller then vectored the aircraft for another approach.

The approach controller and the TRACON supervisor tried several times to get hold of the tower controller by both the internal telephone, and by calling the commercial phone number for the tower. They got no answer. The approach controller advised the crew of American flight 1012 that the tower apparently was unattended, and that the flight would be handled as an arrival to an uncontrolled airport.

The flight was again cleared for approach, and instructed to switch to the tower frequency. At 12:12 a.m., the crew returned to the tower frequency, but was still unable to make contact with a controller. The crew made position reports on the tower frequency while inbound, and landed on runway 1.

A few minutes later, United Airlines flight 628T was arriving from Chicago-O'Hare International Airport. The Washington approach controller told the crew about the problem at the tower, and transferred the flight to the tower frequency. The time was 12:22 a.m. The United flight, also unable to make contact with the tower controller, made position reports while inbound, just as would be the case at an uncontrolled airport. Flight 628T landed at 12:26 a.m.

Two minutes later, American flight 1012, on the ground, was able to raise the tower controller on the radio, and normal control tower services were resumed.
Safety Board investigators interviewed the tower controller, other FAA officials responsible for operations at DCA and FAA personnel at the TRACON. The controller had worked at DCA for 17 years, and had 20 years in all with the FAA. He said that he had fallen asleep for a period of time while on duty. He had been working his fourth consecutive overnight shift (10 p.m. to 6 a.m.).

The National Air Traffic Controllers Association's (NATCA) president Paul Rinaldi issued a statement saying, "One-person shifts are unsafe. Period."
In 1989, the NTSB issued three recommendations to the Secretary of Transportation calling for research, education and revisions to existing regulations regarding required rest periods and duty times. These recommendations were added to the NTSB's Most Wanted List in 1990. In 2011, the issue of fatigue still is on the NTSB's Most Wanted List of Improvements.

The Safety Board noted that the FAA convened a work group to research and propose changes to controller-scheduling policies and practices. The group recommended changes involving increasing the time between shifts. According to the NTSB, the recommendations went nowhere because the FAA decided to pursue a more collaborative approach with the controllers' union and establish its own Fatigue Risk Management Office. The office is sponsoring a study of controller fatigue by the NASA Ames Research Center in collaboration with NATCA that will survey all controllers, and will include an on-site component at a select number of facilities to gather data about controller fatigue. The FAA also has training lessons in place for controllers about fatigue.

In a report on pilot fatigue, Dr. Samuel Strauss of the aerospace medicine division at NASA's Johnson Space Center noted that the term "fatigue" is used to describe a range of experiences including sleepy, tired and exhausted. He reported that fatigue can be created by two major physiological phenomena: sleep loss and circadian rhythm disruption. Dr. Strauss notes that the only effective treatment for fatigue is adequate sleep.

As an example, Dr. Strauss says if an individual normally needs eight hours of sleep to feel completely alert, and gets only six hours of sleep, he or she has two hours of sleep loss. Add more sleep loss over subsequent days, and it accumulates into "sleep debt." Laboratory tests have shown that sleep loss of as little as two hours can produce degraded judgment; degraded situational awareness; degraded decision making and memory; slowed reaction time; decreased work efficiency; lack of concentration; fixation and worsened mood. Our circadian rhythm provides two periods

of maximal sleepiness every 24 hours. These are from about 3 a.m. to 5 a.m. and 3 p.m. to 5 p.m. Fatigue just might seriously affect you if you disrupt your circadian rhythm, and add to that disruption the kind of sleep debt that might come from changes in work schedules.

A study done by the Battelle Institute for the FAA noted that as fatigue becomes more acute, so do increased errors of omission, followed by errors of commission and microsleeps. When you're affected by microsleeps, you involuntary fall asleep for anywhere from a few seconds to a few minutes, and may not be aware of it.

Near Collision

Controller fatigue and sleep loss figured in an incident that had the potential for becoming a tragedy investigated by the NTSB. On August 19, 2004, a Boeing 747-400 came dangerously close to flying into a Boeing 737 at Los Angeles International Airport. Asiana Airlines flight 204, the Boeing 747-400, had been cleared to land on runway 24L. Southwest Airlines flight 440, the Boeing 737, had been cleared into position for takeoff on runway 24L. The crew of flight 204 recognized a problem, and executed a go-around. Radar reconstruction of the event found that flight 204 passed over flight 440 by 200 feet. At the time of the incident, a controller change had just occurred, and the relief controller was responsible for handling both airplanes.

Just after 2:51 p.m., the controller who was being relieved cleared flight 204 to land on runway 24L. At this point, the Boeing 747 was 9.3 miles from the runway. About two minutes later, the controller provided a briefing to a controller just coming on duty. After taking over, the first transmission from the relief controller was to flight 440 instructing the pilots of the Boeing 737 to taxi into position, and hold on runway 24L. At that time, flight 204 was 1.81 miles from the runway at 700 feet. According to flight 440's captain, he saw the Asiana Boeing 747 on final approach, but believed that the aircraft was landing on runway 24R. Twelve seconds later, the relief controller cleared Southwest 440 for takeoff. Radar data indicated the 747 was 1.26 miles from the runway. Calculations showed it would reach the runway threshold in about 35 seconds. Data retrieved from flight 440's flight-data recorder indicated the airplane was on taxiway V approaching runway 24L, when the controller gave it the takeoff clearance. This meant that the pilots had less than 35 seconds to taxi onto runway 24L, begin a departure roll, gain speed and lift off before the 747 crossed the landing threshold, which would be impossible. According to the Asiana captain's statement, he observed the Southwest Boeing 737 approaching runway 24L, but believed the airplane would hold short of the runway. Once he recognized the aircraft was entering the runway, he initiated a go-around.

The relief controller told investigators that he believed the 747 was landing runway 24R. When the Airport Movement Area Safety System (AMASS) generated an alarm, the 747 was only about 12 seconds from colliding with the 737, according to the NTSB. When the relief controller recognized the problem, he canceled flight 440's takeoff clearance and flight 204's landing clearance. However, the 747 had already overflown the 737.

The Safety Board found that the relief controller was briefed on the location and clearances for seven aircraft, and noted that seven pieces of information is about the limit that can be effectively retained in short-term memory. Five certified professional controllers and one operations supervisor were working in the tower cab when there would normally be 10 people. Injuries and illness had reduced the available shift staff to five.

The investigation determined that the relief controller had only eight hours off between the end of his August 18 evening shift at 11:30 p.m., and the beginning of his morning shift at 7:30 a.m. on the day of the accident. As a result, the relief controller reported sleeping just "five or six hours" the night before the incursion, and described his shift leading up to the incursion as a "hard day." The Safety Board concluded that his acute sleep loss hurt his cognitive performance on tasks involving working memory and reaction time.

Peter Katz is editor and publisher of NTSB Reporter, an independent monthly update on aircraft accident investigations and other news concerning the National Transportation Safety Board. To subscribe, write to: NTSB Reporter, Subscription Dept., P.O. Box 831, White Plains, NY 10602-0831.