Guy Leitch
The pilot shortage will hasten the move, which many see as inevitable, to single pilot cockpits. And here’s a thought – single pilot flightdecks are much the same as no-pilot flightdecks.
To deal with the shortage, airlines are making profound changes to pilot employment contracts. In South Africa, including increases in recruitment age – in some cases to 67. In the USA the much-protected minimum requirement of a 1,500 hr ATPL to get into the right seat has quietly been dropped.
Also in South Africa, we are seeing massive recruitment drives to suck up South African pilots for the world’s airlines. Many are finding the attractions of a functional government, a safe environment for their families, and large, sometimes tax-free, dollar-based salaries, very attractive.
But pilot recruitment strategies that poach from other airlines, or extend retirement age, or reduce entry requirements, are still not enough to meet the shortfall caused by the loss of up to 40% of the pilots due to Covid, retirements and cutbacks by the airlines.
At the same time there has been an exponential rise in Artificial Intelligence systems. And so there has been increasing noise about a move to single pilot cockpits. The idea that technology has advanced far enough to make pilots redundant fascinates the general flying public, not to mention all the professional pilots who may one day lose their jobs to a computer.
Basic automation has been making the skilful parts of flying an airliner easy. Autoland has already been around for fifty years, and it is reported that pilots spend on average just six minutes hand-flying the plane on each flight. And now satellite communication (think Starlink) will make it possible for someone on the ground to step in and take control of an airliner if the ‘fleshware’ pilots are incapacitated.
It’s claimed that the Airbus A350F was designed with single pilot operation in mind. This will not come as surprise to pilots used to the overarching role of computers keeping pilots from doing stupid things – in contravention of the Airbus “Flight Laws’.
In the endless Boeing vs Airbus debate, ‘real pilots’ prefer Boeings with their direct controls, while the ‘accountants’ prefer the Airbi that use computers to keep the pilots in line. An SAA captain who moved from flying Boeings to Airbus said, “I took my screwdriver and disassembled the Airbus’s side stick – and wasn’t surprised to find that it isn’t connected to anything at all; no wires, no cables, no nothing!”
He was of course only joking, but it’s important to understand that Airbus is designed around the automation system – and not the other way around. This makes the pilot increasingly redundant. So the bean counters are increasingly calling for singe pilot cockpits.
The concept of a single pilot airliner is at least forty years old. In the 1980s Boeing showed a single pilot flight deck 737 mock up. The first and most obvious objection then, as now, is what happens if the only pilot gets sick? Boeing calculated that for flights up to four hours, the statistics on crew incapacitation were well within the 10 to the -6 limit required. But the incidence of pilot incapacitation cannot be ignored and, as we have seen, pilots are getting older.
Airline pilots and their all-powerful unions heap scorn on the idea. They point out that all airline pilots would have to be rated as commanders, and ask how they would achieve that competency without first being co-pilots? And, if there’s no co-pilot, who would do the paperwork?
Another concern is that autopilots have to have an override mode to trip them out. An un-disengage-able autopilot is the thing of nightmares and B movies as it power dives the aircraft to its doom. Think Boeing MCAS) . But if the single pilot has to go to the toilet, he must know that George (or Otto Pilot) is not going to trip out if it encounters a sudden gust or gets conflicting information from a frozen pitot.
Systems such as Garmin’s Autoland can already land single engine planes and the rise of AI cannot be underestimated for flying an aircraft with pilot incapacitation.



One of the enduring problems with single pilot ops is who keeps the pilot awake on long hauls? There have been numerous incidences of both pilots having fallen asleep and missing reporting points, and even the top of descent. In the days of DC-8s, KLM pilots had an egg timer in the overhead to wake them up. Today, if nothing has been touched for twenty minutes, Boeings put up an EICAS message that reads; ‘Pilot Initiated Event required’. If there’s no response, the caution alert sounds. After another minute, if still nothing has been done, the Master Warning screams.
What about bodily functions? Some scatological wag suggested that pilots should not have to actually get up to go to the toilet. Since she is alone in the cockpit, it should be possible to make a pilot’s seat that, with a flick of a switch becomes a toilet so she can go there and then. This would probably suit women better, as no real male pilot would sit to pee.
A key requirement for single pilot ops is satellite-based communication with aircraft and this is now available, cheaply and easily through Starlink. If only we had it in South Africa.
The military is far ahead and uses satellite comms to fly UAVs on the other side of the world. So it’s entirely reasonable for an airliner to have just one pilot. When she needs to go to the toilet, or have a rest then, like a UAV, the flight can be managed from the ground. Instead of having a second pilot costing millions a year, the flight could be monitored by a cheaply trained flightsim enthusiast in Mumbai, earning R200,000 per year – and looking after many flights at once.
Single pilot ops with a backup on the ground will require huge regulatory changes, as well as new technology for the ‘ground pilot.’ This would open the doors to fears that the systems used by the ground pilot to fly the plane remotely could be vulnerable to cyber-attacks to remotely hijack a plane. So they wont even need suicide bombers.
Perhaps the real problem holding back pilotless airliners is that the passengers would not be happy. But it’s not unreasonable to expect that they could soon get used to it. We have learned to accept driverless trains, and it wasn’t so long ago that skyscraper lifts had operators. The aircraft hardware and software has built in redundancy, but could it ever be okay if the ‘wetware’ (the human brain, which is the weakest link) – has no backup? There has to be redundancy, so to have single pilot operations there will have to be aircraft that can fly with no pilots at all. Ipso Facto – Pilotless planes.
This redundancy can come from the explosive arrival of Artificial Intelligence. Sir Tim Clark, the head of Emirates, says, “Artificial intelligence will likely be along for the ride in every flight deck.” Clark has said that he doubts passengers will accept a pilotless flight deck. “You might see a one-pilot aircraft, but they will be a long time coming.”
Can an airliner be flown on a fully automated basis? Clark said, “Yes, the technology is right up there now, but in my view, there will always be somebody on the flight deck.”
The idea that passengers may be so committed to cheap tickets that they would accept just one pilot may be fallacious. Using round numbers, assume a Second Officer has a cost to company of R2 million and flies 50 long-haul sectors per year, that is R40,000 per sector. If that is distributed amongst 300 passengers, that is R133 per ticket. On a fare that could easily cost R10,000, this is barely 1%. I reckon that given a choice between one or two pilot airlines, passengers will be happy to pay R150 for the comfort of a second pilot. Same as we are happy to pay petrol pump attendants.
In the final analysis it’s worth remembering that, no matter how smart AI gets, human experience and judgement, and the subjective skills of CRM, all add up to the aphorism that pilots are not paid for what they do: they get paid for what they can do when the electronic brains get confused.
So, I don’t see single pilot ops, and its natural conclusion, pilotless planes, happening any time soon.