There is a great deal of dialogue about the use of automation in aviation. Lately, some of the talk is focusing on the failure of automation and the role it played in the crash of Asiana 214, the Boeing 777 that failed to make runway 28L at San Francisco International. The point behind having automation in the cockpit include reduction of pilot workload and safety to name just two.
According to a study by faculty of Embry-Riddle Aeronautical University and The University of Central Florida, (Managing Automation in the Cockpit, by John A. Wise, Donald S. Tilden, David Abbott, Jennifer Dyck, and Patrick Guide, 1994) automation in fact does increase safety and efficiency. It also introduces new problems in the cockpit. Through surveys and studies administered to professional flight crews, the authors discovered that while there were improvements in some areas, automation also introduced other problems, most notably, in the human interaction with the systems.
Another paper written by Major Wesley A. Olson of the United States Air Force and presented as a Wright Flyer Paper (Identifying and Mitigating the Risks of Cockpit Automation), notes that automation does provide the promises of safety and efficiency, but at the cost of “mode confusion, errors of omission, and automation surprises.”
Other studies identified the sophistication of certain technologies as an additional problem area with automation. A number of pilots report some systems as “generally non-intuitive for pilots.” There was also an indication the onboard computers were inadequate for performing the tasks requested by pilots of the hardware/software. Another difficulty pilots face is the non-standard installation and operational requirements of different systems. In the 1960s, studies noted a similar problem with instrument panel standardization and in 1969, the industry adopted the standard T arrangement resulting in the “six pack” for instrument installation in new general aviation aircraft. Like instrument panels, we need to develop a “standard T” for functions of various automated systems in different cockpits.
In the course of any discussion of aircraft accidents, the focus quickly falls to the failure of cockpit automation or the possibilities of the crew’s misunderstanding of how the systems functioned. This can result for any number of reasons; lack of education in the system, improper use of the system, or complacency on the part of the pilot(s).
There is no place in any cockpit for any of the above. Companies must make sure their flight crews know all the systems installed in the airplane, they must know exactly how to operate each system, and they must always guard against complacency on all levels.
One accident I occasionally discuss is the loss of John Kennedy, his wife, and her sister. While Kennedy was training for his instrument rating and owned a rather sophisticated single-engine aircraft, the accident report noted he did not know how to use the autopilot. Because of this, it is speculated that he failed to engage the autopilot, which, in the lowest degree of automation, would have allowed him to keep the wings level.
Regarding the second aspect of the use of automation, the more a pilot knows how to use the technology, the higher measure he or she will have regarding personal confidence. After learning about the system, a pilot must be able to know how to use the automation confidently. If a pilot lacks knowledge of the system, that’s one thing, but pilots also need to know how to use the system. And again, with confidence.
The third aspect, complacency, is a bit more difficult. How do we keep complacency out of the cockpit? Pilots get tired, they don’t eat well enough, they drink sodas and coffee instead of water, which leads to dehydration, and they become bored monitoring the same systems and instruments for long periods. It takes a lot of discipline to stay on top of this game.
In the early press releases and the speculation as to the cause of the accident of Asiana 214, many have focused on the automation issues. Some news organizations reported the flight crew was flying with the auto-throttles engaged and they believed the system would maintain airspeed. There is also speculation regarding the failure of this system.
It will be a long time before the National Transportation Safety Board (NTSB) releases the complete accident report. In most cases, reports concerning air carriers take more than a year and as long as 18 months to finish and made available to the public.
I don’t know the cause of the accident of Asiana 214. Over time, causal factors will come out and recommendations made. I do know how complacency affects many of us. I know how automation affects me and the role automation plays in my complacency and the discipline I must exercise to stay sharp while flying.
Also, after studying hundreds, if not thousands of accident reports over my lifetime, I see another trend developing that we, in the industry, seriously need to address. And that is the failure on the part of humans in the area of basic airmanship and judgment.
-30-
©2013 J. Clark
Note: Email subscribers, please go to my blog to view vids
Automation can be seductive. Just when you think it is reliable and foolproof, it bites you in the posterior. I seldom used auto throttles for approach, unless the weather dictated an autoland, because the slightest pitch change to maintain glide slope caused un-necessary power changes. However, I think everyone uses them for takeoff because once the power is set it remains constant and smooth. My bite on the posterior came after takeoff on a clear day when I called for climb power on the MD-88. The copilot punched the mode button and the left throttle reduced to climb EPR and the right throttle went to idle. Even with my hand on the throttles, it took me by surprise. I disconnected the auto throttles and pushed it back up, but not before the airplane sensed an engine failure and shut down both air conditioning packs and sounded a gear warning. The outflow valve closed and caused a pressurization surge that I’m sure the pax enjoyed as much as I did. With the auto throttles off, we continued on our merry way with no problems. Later the copilot told me that the maintenance log showed a history of auto throttle problems. Using my finely honed CRM skills, I inquired as to why he didn’t think that would be a pertinent subject for discussion prior to takeoff (the maintenance log is actually a checklist item). I guess we both learned something and reinforced the aviation motto, “In God we trust, everybody else is suspect.” That includes the automation.
Harrison, wow, what a story. I absolutely hate surprises! We had auto throttle in the A-7, but we hated to use it. Designed for those times you might bring a crippled airplane back to the boat and you were unable to use your left hand… Invariably, you would go a little high at the ramp, the auto throttle would pull off too much power, the jet would then settle through the glideslope and hit the deck just beyond the ramp, and then taxi up to engage the one-wire. So much easier to do it by hand…