Are modern airliner cockpits automated to the point where pilots are no longer able to intervene effectively? Only new research will tell
As long ago as the late 1960s, airline pilots were saying that their job was changing: it was less about flying and more about managing the flight. Automation, in the most modern aircraft at that time, was already capable of taking over the flying task for almost all phases of flight. The Hawker Siddeley Trident was probably the most advanced airliner then, by virtue of the operational test programme that made it the world's first "blind landing" autoland aircraft.
Navigation, however, was still carried out using raw data, presented electro-mechanically and processed in the pilots' heads. Aircraft systems were still, with a few exceptions, handled by a flight engineer (FE) or third pilot, and in aircraft like the BAC One-Eleven, McDonnell Douglas DC-9 and Boeing 737 that did not have an FE the systems were relatively simple and partially automated.
Technology has moved on massively since then and, particularly in the widebodies, systems have become more complex; not only because the aircraft are larger, but also because they have to be operated more precisely, both in navigation terms and to be more efficient. An example of the latter is the tailplane trim tank: not merely a place to store extra fuel, it is used to adjust the aircraft's centre of gravity optimally during the flight to reduce aerodynamic drag. Meanwhile, digital monitoring and control have replaced the FE.
In the early 1990s the US Federal Aviation Administration decided it was time to review the effect of increasing automation on air transport operations. Automation had accelerated with the introduction into service of the Airbus A310, Boeing 757 and the 767 just over a decade earlier, and continued with updated variants of existing types, followed by a new generation of aircraft. Among the early changes the most fundamental were integrated electronic flight instrument systems (EFIS) and digital flight management systems (FMS). At the same time, electrical, hydraulic and cabin environmental control was becoming more automated, so that the number of physical actions required by the crew to manage the aircraft's systems via the overhead panel was reducing. And the new equipment had been in service sufficiently long to make a study of its performance worthwhile.
The FAA led a team of international aviation experts and academics in a major study, and in 1996 published a report called The interfaces between flightcrews and modern flight deck systems. It concluded that modern, automated aircraft were safer than their predecessors and were on the right design track, but they were far from perfect and crews needed more training to manage them, not less. This flew in the face of an initial claim by Boeing that the new airline pilot would be trained on a "need to know" basis, meaning that it was only necessary to know what the automated systems did, not how they did it. No-one has preached that for a long time now.
Incidents like the recent Airbus A340-600 diversion into Amsterdam, about which the UK Air Accidents Investigation Branch has just published an initial report, suggest that it is time to revisit the FAA automation study to see if its recommendations have been applied by the manufacturers, the regulators, and the airlines.
Another automation study may not seem necessary because the first was not only thorough, but it avoided getting bogged down in fine technical detail – it addressed automation and the associated human factors issues generically. But the need for a new study a decade on should not be ruled out. Automated systems management in the latest aircraft is becoming integrated almost to the point of artificial intelligence, and it is normally so reliable that pilots rarely need to intervene. So on the rare occasions when they do, a degree of confusion – often born of surprise – should be expected, especially if they have not been given any of the warnings they would normally expect to receive.
In the recent A340-600 fuel management system incident, not only did the systems faults disable the ultimate warnings – tank low-fuel alerts – but they had deprived the pilots of information about system activity they would normally have seen during the flight. It has frequently been shown, however, that the absence of routine information is a cue that pilots fail to notice.
With the A380 and the 787 on the way, the European Aviation Safety Agency and the FAA must revisit the 1996 study on human factors and automation, but also determine whether there is a need for further research.
Source: Flight International
