Torqued: Aviation Safety and ATC Fatigue

 - December 1, 2015, 3:39 PM

All of us in the aviation safety business–and who isn’t involved in safety if they’re in aviation?–talk a lot these days about human factors. Human factors are involved in the overwhelming majority of aircraft incidents and accidents, by some estimates more than 80 percent, as aircraft and their components have become more reliable and less prone to causing catastrophic failures. The equipment has become less vulnerable to failures, but human errors and lapses have proved stubbornly difficult to anticipate and mitigate or eliminate. Much of the conversation these days involves the effect of fatigue, perhaps because a fatigue connection has been cited as a contributing factor in a number of tragic accidents across modes of transportation.

So, with all this concern about fatigue I was surprised to see that a recent National Air Traffic Controllers Association (Natca) media blitz concerning air traffic controller staffing shortages focused on their potential for creating more airline delays. This effort failed to mention any safety impacts, even though union officials claimed that the staffing shortages have led to six-day work weeks at some of the busiest facilities in the U.S. (Atlanta, Chicago and New York among them). Curiously, these union officials were quoted by numerous media sources as saying that in spite of the six-day weeks, the decrease in staffing was not causing a safety problem. Say what? Routine six-day work weeks and no concern for fatigue? That seems rather odd.

Fatigue and its effects on human performance have been an enduring concern. As I wrote about earlier this year, the NTSB pointed to crew fatigue as a contributing factor in the UPS accident on approach to Birmingham, Ala. on Aug. 14, 2013. The A300 crashed just short of the runway, killing both pilots. The Board determined that the probable cause of the accident was “was the flight crew’s continuation of an unstabilized approach and their failure to monitor the aircraft’s altitude during the approach, which led to an inadvertent descent below the minimum approach altitude and subsequently into terrain.” The Board concluded that the crew’s fatigue contributed to the errors made on that early-morning approach to land.

Another aviation accident I wrote about involves the finding of the UK Air Accidents Investigation Branch that maintenance errors caused fan cowl doors on both engines to detach from an Airbus A319 on takeoff from London Heathrow Airport, damaging the airframe and a number of aircraft systems. A fuel leak from one of the damaged engines started an external inflight fire. Fortunately, the crew landed safely and there were no fatalities or serious injuries.

Once again, investigators listed fatigue as one of the contributing factors in the chain of events that led to the mechanics forgetting to close and latch the engine cowlings properly. While the report states that the mechanics’ work schedules complied with legal requirements, it also finds that their performance “may have been compromised by fatigue, induced by the significant level of planned and overtime working undertaken before the overnight maintenance shift” when the engine cowlings were left unlatched.

Accidents in other modes have also brought focus to the fatigue factor. The recent appearance of comedian Tracy Morgan on Saturday Night Live reminded me of the tractor-trailer accident he was involved in on the New Jersey Turnpike that left one person dead and Morgan severely injured. The truck driver had been on duty 13.5 hours of his 14-hour work day, and awake for 28 hours. Not surprisingly, the NTSB found fatigue a factor in the deadly collision. Commuter train accidents, including several fatal ones, have also focused on fatigue as a causal or contributing factor.

NASA, FAA Study Six-day Work Week

With fatigue a concern throughout the transportation industry, it made no sense that the union would blithely dismiss six-day work weeks as having no safety impact. I was curious to see what basis the union had for saying that controllers’ routinely working an extra day did not affect safety, especially since a NASA report prepared at the FAA’s request had found a number of concerns related to controller fatigue. The NASA study was prompted by an NTSB recommendation to the FAA in 2007 that the agency “work with [Natca] to reduce the potential for controller fatigue by revising controller work-scheduling policies and practices to provide rest periods that are long enough for controllers to obtain sufficient restorative sleep and by modifying shift rotations to minimize disrupted sleep patterns, accumulation of sleep debt and decreased cognitive performance.” 

The NTSB’s recommendation resulted from its investigation of the August 2006 crash of a regional jet in Lexington, Ky., when the aircraft took off from the wrong runway, crashing into a perimeter fence and killing 49 people on board. The Board concluded that the controller on duty was likely fatigued at the time of the accident, although it was unable to determine that the controller’s fatigue was a contributing factor. Nonetheless, the Board’s concern for the impact of fatigue on safety prompted the recommendation. 

With a little digging–the FAA had kept the 2012 report secret until this past summer–I found a copy of not only the NASA report but also a cryptic FAA cover memo prepared two years after the study was completed. The cover memo is dated July 8, 2014, from the FAA’s v-p for safety and technical training to the air traffic COO. The memo states that although the FAA paid NASA $1.2 million to “conduct an independent, baseline assessment of fatigue in the air traffic controller workforce,” in the end the agency did not accept NASA’s report, which reached numerous conclusions and made a number of recommendations regarding air traffic controller fatigue related to their work schedules. It did, however, use some of NASA’s study to prepare its own fatigue risk management system in conjunction with the controller union.

The FAA’s memo states that it did not accept NASA’s report because “the academic approach used by NASA did not sufficiently integrate an understanding of the air traffic 24-7 operational environment with a scientific approach. As a result, some verbiage in the report contained judgmental rather than quantitative results–often based on sourcing opinion versus science.” I’m not completely sure what that means but it sounds like NASA just got badly dissed by the FAA. I reached out to the FAA, Natca and NASA for comment but none of them responded.

In any event, reading through the 225-page report–which looked to me to be exhaustively and scientifically researched, including an extensive field study–I came upon some interesting information related to six-day workweeks. First, the study found that of all the aspects of controllers’ jobs included in the research, schedules were the number-one complaint and the one controllers felt contributed most to their fatigue. Six-day work-week schedules were the least satisfactory. They also appeared to be the most troublesome from a safety perspective. According to the report: “Respondents working this schedule reported having a higher percentage of operational events in the previous year than those on any other schedule (32 percent versus 17 percent)” although the percentage that reported constant six-day schedules was small (4 percent). However, the percentage that reported working a six-day week their last week of work was 14 percent.

The study concluded,  “Fatigue was associated with working six-day schedules with only a single day off.” However, the study left unanswered whether alertness was affected by working six-day schedules. NASA recommended investigating the impact of six-day schedules on controller alertness.

To date, it does not appear that those additional studies have been done by or for the FAA. It would certainly seem to me that the biggest concern with staffing shortages increasing six-day work weeks at our busiest facilities should in fact be safety. And whether the fatigue produced by those schedules does, in fact, affect controller fatigue.

The question I have is why doesn’t the FAA want to find that out? And why is the controllers’ union letting it get away with not finding out?