Japanese carrier All Nippon Airways has implemented a turbulence-prediction service based on artificial intelligence, following several years of testing and validation.
ANA and Keio University formed a Tokyo-based partnership called BlueWX two years ago to specialise in AI-based weather forecasting and advance the prediction service – development of which began in 2019 with trials starting in 2021.
Its model has been built on a decade of turbulence data and refined with “deep-learning techniques”, says the carrier.
ANA claims the service has an 86% accuracy rate and that its reliability has been confirmed in the operational trials which involved 2,500 of its pilots.
The BlueWX service has been formally integrated into its weather-data systems and ANA becomes the first airline in the world to adopt it.
“Turbulence prediction is a long-standing challenge,” says ANA operations division executive vice-president Hiroyuki Kometani.
“Implementation of this new system will ensure a more reliable and comfortable air travel experience for all our passengers.”
BlueWX plans to offer the service to other airlines, through distribution channels. Chief executive Kaz Watanabe says ANA’s introduction highlights the model’s predictive capabilities.
ANA says its pilots expressed “strong support” for its continued use, highlighting its “accuracy against flight reports” and effectiveness compared with traditional methods.
ICAO’s newly-released global aviation safety report states that 95 accidents occurred last year involving scheduled commercial flights, and 32 of these involved turbulence encounters – the highest number of accidents in a single category.
While turbulence only contributed to a small fraction of fatalities, it was responsible for more than 72% of serious injuries.
