By providing more accurate digital weather forecasts with longer lead times, the KAIROS project aims to help aviation better manage how weather impacts operations.
It’s no secret that weather can wreak havoc on one’s travel plans. In fact, 25% of delays in the air traffic network can be attributed to weather.
But bad weather’s impact goes well beyond delays and cancellations.
“Bad weather at one airport can have a trickledown effect, quickly impacting flights across the airspace,” says Aniel Jardines, CEO at AI METHODS.
Take for example thunderstorms. Because a thunderstorm can create unsafe conditions, ATM will often rightfully error on the side of caution and close the entire airspace. This in turn translates into delays, rerouting, and cancellations across airports.
But it doesn’t necessarily have to be this way.
As Jardines explains, the problem is that air traffic controllers do not have sufficient weather information as their disposal to make important operational decisions.
“While today’s aviation weather products can often accurately predict the presence of hazardous weather, these forecasts lack the spatial and temporal resolution needed to take meaningful action to mitigate the impact on the operations,” he says. “With climate change increasing the frequency and intensity of hazardous weather, the aviation community faces a difficult challenge in growing an air traffic system that is resilient to weather.”
According to Jardines, having access to higher resolution forecasts that provide more detail on the timing and location of thunderstorms would enable air traffic controllers to make decisions to minimize the impact on air traffic operations.
Artificial Intelligence (AI) could be the key to providing such accuracy.
With the support of the SESAR-funded KAIROS project, AI METHODS, together with a consortium of aviation and technology partners, look to unlock the potential of AI-based weather forecasts for operational benefits.
“By providing more accurate digital weather forecasts with longer lead times, we aim to help aviation stakeholders better manage how weather impacts their operations,” adds Jardines.
Enhanced accuracy, longer lead times
Although a work-in-progress, the project has already achieved some important results.
One of those results is the use of AI algorithms to improve the prediction of convective weather, such as thunderstorms. “We successfully demonstrated that our algorithms offer enhanced accuracy and longer lead times compared to existing weather information,” remarks Jardines.
The project team is currently working on additional AI models to predict such weather phenomena as dust, high altitude crystals, turbulence, and low visibility.
All these models will eventually be integrated into an end-user platform, with demonstrations in an operational setting expected in 2025. The ultimate goal is to reach a Technology Readiness Level (TRL) 7.
The importance of a strong consortium
This success is not to say that the project hasn’t experienced any turbulence itself.
Early in the project, researchers faced several challenges, including an issue with the quality of the data used to create the algorithms. “Observational data on weather phenomena is not perfect, and combining various data from different sources into a holistic representation was anything but easy,” notes Jardines.
The key to overcoming this challenge, and others, was cooperation from across the project’s partners. “We are proud to have support from partners representing all areas related to aviation weather forecasting and who excel at integrating AI-based forecasts into existing operational tools,” says Jardines.
Looking ahead, the project expects to confront a number of regulatory and certification related challenges – challenges that will once again be addressed through collaboration, iterative problem solving, and adaptability.
Digitising Europe’s airspace
Once finalised, the KAIROS solution has the potential to initiate a paradigm shift in the digitisation of European airspace.
For one, it could position AI-based weather forecasting as a crucial component of airspace management, unlocking operational efficiencies and automation opportunities for air navigation service providers (ANSPs), airports, and airlines.
The uptake of AI would also drive the development of a highly skilled workforce, possibly creating new business opportunities and high paying jobs.
“Based on our early and exciting results, the entire project team is confident that our work will have a transformative impact on the aviation industry, revolutionising how weather-related challenges are managed and mitigated,” concludes Jardines.
Check out the KAIROS prototype.
More about the project
https://www.sesarju.eu/projects/KAIROS
