© Provided by La Tribune "It is estimated that autonomous flight would save aircraft manufacturers, their suppliers and airlines more than $ 110 billion per year. Savings from flight optimization, capacity increased forecasting, lower insurance costs and pilot training "(Steve Ganyard, former Marine Corps fighter pilot).
Confronted with the economic devastation caused by Covid-19 on the commercial aeronautics sector, limited by the state of the art of current technologies, and exposed to the development of "plane-bashing" particularly in Europe, aircraft manufacturers are today on the defensive. In this context, a daring strategy would be to accelerate the inevitable substitution of pilots by algorithms, helping to revive the world economy with cheaper, more reliable and safer flights.
For commercial aeronautics, financial resources are a first challenge: the return to normalcy of air traffic postponed for several years, the enormous loss of income and the chain effects of the debt created by the Covid-19 will have an effect. undermines both R&D budgets and investments. Aircraft manufacturers, airlines, and ultimately the ecosystem as a whole, will for a long time have neither the funds for development, nor the demand in terms of passengers, nor even the technologies, to launch a new aeronautical program as envisaged. before the health crisis.
Unmanned aircraft: savings of 110 billion per year
In the world before, new production methods, increased propulsion efficiencies and the use of better materials generated such improvements in the efficiency of commercial aircraft that the $ 20 to $ 30 billion in investment necessary for the design of new models until their entry into service were justified. Today, and for the next ten to fifteen years, existing and emerging technologies can do little more than generate marginal improvements to engines and airframe. While waiting for the hydrogen airplane revolution, led by Airbus in particular, the levers for lowering costs should be looked elsewhere.
It is also not unthinkable that airlines are drawn to a costly 5% improvement in engine efficiency when the price of oil hovers around $ 40 a barrel. In the absence of clear improvements in efficiency and a faster return on investment, airlines will be reluctant to purchase such "new" aircraft.
Where to find substantial efficiency gains if aeronautical technologies stagnate for ten to fifteen years? The answer is in the cockpit. Autonomous flight is estimated to save aircraft manufacturers, their suppliers and airlines more than $ 110 billion annually. Savings from optimization of flights, increased forecasting capacity, reduced insurance costs and pilot training.
In 80% of crashes, a pilot error
The growth generated will be much greater than the savings made. Commercial aviation is a trillion-dollar industry with over 10 million associated jobs worldwide. The overall decrease in the number of pilots will be more than offset by the improvement in the global economy brought about by autonomous flight.
Dragging its feet towards autonomous flight can have serious consequences: designing software systems to address human-made deficiencies rather than progressing towards autonomy has already proved fatal. The Boeing 737 MAX's faulty MCAS system is a sad example of misguided automation, not failing range: in attempting to recreate a human "feel" , Boeing has attacked the weakest link in the human interface -machine, de facto opening the field to human error which cost the lives of 346 passengers.
More generally, the frequency of accidents on commercial flights (although very rare) is not decreasing on average. Boeing estimates that human piloting error is the cause of more than 80% of crashes. A proactive and phased approach to autonomy – first a full cockpit, then single-pilot, before eventually achieving full autonomy provided by air traffic control only – is paradoxically necessary to achieve higher levels of safety. important.
There is obviously a lot of work left to guarantee the safety of the algorithms as well as of air traffic control. However, technological progress is much faster than before, in particular due to advances in the fields of artificial intelligence, machine learning and the generalization of spatial guidance (GPS, star finders). "The only reason why we still have pilots is public opinion," recently confided (but discreetly) a leader of the FAA (Federal Aviation Administration, the American equivalent of the European Aviation Safety Agency). This “public opinion” can however change if we know how to create a climate of confidence by demonstrating the safety, redundancy and safety of the autonomous aircraft. This will require considerable awareness-raising and educational efforts.
Airbus recently proved that the taxiing, take-off and landing phases could be autonomous, with a first phase of conclusive tests carried out in January 2020 with an A350. Airbus plans that the first planes to fly with a single pilot in the cockpit will enter service by 2030.
Boeing is a leader in the military field, notably with its “ Loyal Wingman ” (an advanced combat drone). The US Air Force recently simulated an aerial combat between an algorithm and an F-16 pilot. Result without appeal. Algorithm: 5 – Pilot: 0. The same advances exist in the field of cargo: two start-ups in Silicon Valley have operated small autonomous cargo planes, with full integration of air traffic control systems on the ground. The CEO of FedEx is now convinced to use this type of solution to serve certain particularly remote areas of the United States. The paradigm shift towards autonomy has already started. The time has come for the pilots to leave the cockpit.
(*) Steve Ganyard is a former Marine Corps fighter pilot, with more than 4,200 flight hours on 12 different types of tactical aircraft. Today he is one of the advisers of Avascent, a strategic consulting company in aeronautics, space and defense.