Traffic forecast

AI has provided EUROCONTROL forecasters with a set of highly capable new tools in their predictive analysis toolbox.

Background

Air traffic forecast is reaching new levels of uncertainty with the industry itself becoming less predictable, more prone to sudden swings in industrial, economic and environmental turbulence. However, despite increased volatility, accuracy of forecasts has made small but almost constant improvements. This is primarily due to a better understanding of the air traffic dynamics key patterns, which provide indications of future demand but also to the use of Artificial intelligence (AI) techniques.

How it works

The forecasting system is based on a wide number of different components. The STATFOR forecasters isolate the trends, the relationships between the different factors into components, and ultimately improve the quality and performance of the final forecast. When they’re forecasting air traffic movement in Europe, they’re dealing with relatively small numbers of countries, airports, aircraft operators and aircraft types - which means that one event or one decision by a carrier can significantly change the traffic in several countries. However, statistical techniques are reaching their limits and human analysis remains essential.

AI machine learning approach analyses thousands of different types of exogenous data, more than any analyst could make sense of. The machine learning system selects the most relevant data sets.

Benefits

With new AI algorithms using more input into the calculations than just the core gross domestic product (GDP) data results have been very positive: the tests performed on seven traffic flows across the North Atlantic showed that ML process enabled the reduction of the median absolute error, by between 8.5% to even 71% for some pairs compared to the STATFOR median absolute error, for a specific year.

Carefully selected AI technics relying on good datasets and deep knowledge of the local situation can further improve forecasts. Applying these principles, STATFOR forecast experts were able to continuously improve their forecasts. In 2018, the median error per year was small, at 1.1%, and the median absolute error per year was 2.1%, one of the lowest errors ever recorded by STATFOR. The 2018 forecast was 104% more accurate than the same-as-last-year growth baseline.

As turbulences within the Global air travel industry are predicted to grow over the coming years, the importance of identifying and exploiting new AI techniques – and then learning how best to apply them – will also increase.