How Aviation is changing to face the growing turbulence on flights

The interior of the Singapore Airlines plane that suffered severe turbulence — Photo foot, Several passengers collided against the trunk of the Cabin of the Singapore Airlines flight that landed emergency after a severe turbulence last year.

- Author, Philip Maughan
- Author's title, BBC Future
20 minutes

“We saw blood on the roof … it was a total chaos.”

That was part of what a passenger told a flight from the Singapore Airlines airline that was affected by strong turbulence when flying through southern Myanmar in 2024. “There were many people lying on the ground,” he said.

At the beginning of this year, a Boeing 787 plane from United Airlines also suffered strong turbulence while flying over the Philippines. A hostess was shot at the ceiling, which caused a brain shock and a fracture in the arm.

Incidents of strong turbulence such as these are increasing as a result of climate change caused by human activities.

Severe turbulence in clear air (CAT), that is, very agitated air that is invisible to satellites, radars and the human eye, have increased 55% since 1979, when the reliable weather records began, according to an investigation by Paul Williams, professor of atmospheric sciences at Reading University (United Kingdom).

Turbulences are expected to triple throughout the world for the 2050s and have an important impact on East Asia and North Atlantic routes.

This could influence people no longer wanting to fly. Among the most common reasons that people give to justify their fear of getting on airplanes are loss of control and a previous experience with turbulence.

But turbulence, in addition to being potentially dangerous, also involve a cost for the aviation industry, since they cause wear to vehicles and lengthen some flights, as pilots try to avoid them. These maneuvers involve greater fuel consumption and an increase in polluting gas emissions.

Although turbulence is usually more a matter of discomfort than injuries or death, the increase in the volume of chaotic movements in the atmosphere means that airlines, scientists and engineers face the challenge of finding ways to mitigate the problem.

Changes in wings

The cabin of an airplane — Photo foot, Turbulences have increased by 55% in the last four decades.

Flying directly through swirls, vortices and ascending currents with a minimum disturbance requires not only precision engineering, but also many advanced mathematics and an analysis of fluid dynamics. (The air, like water, is a fluid).

The panorama will always be complicated because the fundamental nature of turbulence is that they are chaotic. Small disturbances, from how the wind deviates from one building to the wake of another aircraft, can change the behavior of the currents in the air.

Turbulence Solution, an Austria headquarters, has developed small “flalets” that can be added to the larger ailerons (or flaps) of the air wings.

The flalets slightly adjust their angle to counteract the changes in the air flow based on the pressure readings taken immediately in front of them on the edge of the wing.

This helps stabilize the plane, similar to how birds use small settings of their feathers while flying.

The company states that its technology can reduce the turbulence loads that passengers feel by more than 80 %. So far he has only tried technology in small airplanes, although the executive director, Andras Gallffy, who is an acrobatic pilot, is confident that it can be expanded for much larger airplanes.

“The widespread opinion is that the turbulence can be avoided or accepted and their belt opening up and reinforcing the wing,” he explains.

“But we say that it is not necessary to accept them. Only the appropriate signal is needed. For light airplanes, this has always been a problem, but even for commercial aviation it is becoming more serious because turbulence are increasing.”

A wing of a plane seen from inside the cabin — Photo foot, A solution explored by specialists is to add attachments to wings to better overcome turbulence.

An ideal task for AI

Turbulences can be difficult to understand for humans, but artificial intelligence (AI) may assimilate it more easily.

“Automatic learning is very good to find patterns in large data volumes,” says Ricardo Vinuesa, a researcher in fluid mechanics, engineering and artificial intelligence at the Royal Institute of KTH technology in Stockholm. “Turbulence could be the perfect application for AI.”

In a recent experiment, Vinuesa and its colleagues from the Supercomputing Center of Barcelona and the Technical University of Delft tested an AI system that controlled “synthetic jets” of air in the simulated wing of an airplane.

The AI trained through deep learning by reinforcement, a process in which the model learns by rehearsal and error, similar to how a small child learns to walk. “Instead of making countercurrent measurements, we can use AI to create very precise numerical simulations of air flow behavior based on measurements taken directly in the wing,” he explains.

“And while artificial neural networks (the computational models used by AI to learn patterns) are usually considered black boxes, we use the explainable, which allows us to determine which measurements are most important for the predictions generated by the model,” he adds.

Vinuesa and her colleagues are working with technology companies to continue developing technology.

A plane between clouds — Photo foot, Turbulences mean much more than discomfort for passengers: they can cause serious tensions in the fuselage of aircraft.

Last year, a Caltech and Nvidia team generated extreme turbulences within a wind tunnel to test a system of detection and prediction based on IA for drones, with promising results.

Experts from the NASA Langley Research Center tested a microphone specially designed to detect ultrabajas infrasonic frequencies created by swirls of clear air turbulence at a distance of up to 480 km.

Another approach that has been actively developed since at least 2010 consists in the use of detection and location by light (Lidar) to create a 3D map of the air that surrounds an airplane, very similar to how autonomous cars create a cloud of points of nearby objects and vehicles to navigate its surroundings.

A Chinese study of 2023 proposed a “double wavelength” Lidar system, which, as they affirm, can observe turbulence of light to moderate between 7 and 10 km ahead of the aircraft.

Unfortunately, the lowest density of large -altitude air molecules causes the instruments to be too large, heavy and consume too much energy to be able to use them in current commercial planes.

An atmospheric radar in a pilot cabin — Photo foot, Currently the pilots have several time monitoring resources to detect atmospheric instability.

What is being done now?

The convergence of manufacturing, artificial intelligence and new sensors could transform aviation in the second half of the 21st century.

But what happens today? Before takeoff, the pilots check the weather reports and study the maps of the currents in jet. They consult the flight planning software and check the forecasts, such as the Graphic Turbulence Guide (GTG), to which Paul Williams contributed.

“About 20 years ago we could predict about 60% of turbulence,” he says, “today is more 75%, and I guess my professional goal is to continue increasing that figure.”

When asked what slows down, Williams replies that it is access to turbulence data measured by aircraft. “Researchers have to buy the data, and they are not cheap.”

Thanks to advances in computing, artificial intelligence and the growing number of satellites, weather forecasts are improving, but there is a general lack of wind measurements above the earth's surface.

What we know comes from about 1,300 meteorological balls distributed throughout the planet and the accelerometers of approximately 100,000 commercial flights that cross the skies every day.

Turbulence Aware, of the International Air Transport Association (IATA), analyzes and shares data on real -time turbulence and now uses airlines such as Air France, Easyjet and Aer Lingus.

For passengers, there are more and more applications that give access to data that until now could only see the pilots and flight dispatators, one of which is turbli.

“I use Turbli. It seems quite accurate, considering that they do not know the exact route, so they cannot be 100 % precise,” says Williams.

“But it is a bit like a hypochondriac that seeks its symptoms in Google. I'm not sure it always helps,” he finishes.