AI Methods advances meteorological data work within the VARIANT project

AI Methods is leading the data acquisition activities of the VARIANT project, laying the foundation for AI-driven micro-weather services for Innovative Air Mobility.

The VARIANT project (Validation of AiRspace Integration Applications and New Technologies), launched by Boeing Aerospace Spain in January 2026, aims to support the safe integration of electric and autonomous aircraft into urban airspace. A core part of this effort is characterizing the weather conditions under which future low-altitude operations will occur. Meteorological factors such as wind, turbulence or visibility, can significantly affect the operations, so reliable and high-resolution weather information is fundamental to the project’s objective.

AI Methods, together with its subcontractor Universidad Carlos III de Madrid, is establishing a consistent, comprehensive data infrastructure to support development of high-resolution weather-prediction models, based on AI models and specifically tailored to low-altitude operations.

Integrating diverse sources of weather data

In recent months, AI Methods has focused on identifying, accessing, and integrating multiple sources of meteorological information, including satellite observations, ground-based sensors, numerical weather prediction models, and operational aviation weather data.

Combining these sources presents an important technical challenge: each provides different types of information, at diverse spatial and temporal resolutions. Integrating them will allow the project to build a more detailed representation of the atmospheric conditions relevant to Innovative Air Mobility.

The purpose of this data is twofold. First, the different identified sources, plus data acquired via local weather sensors deployed for the project, will be classified and form an essential part of the training dataset for the models, helping them uncover the underlying weather dynamics. Second, those same data, in combination with the deployed sensors, will be used to validate forecast variables and to compare and benchmark the different approaches that will be tested throughout the project.

Preparing the ground for AI-based weather services

The meteorological datasets assembled during this phase will feed into the next stage of VARIANT, where project partners will develop and validate AI models and a digital weather service. This will enable the project to move from raw and modelled weather observations toward more detailed, operationally useful information for low-altitude aviation.

AI-enhanced forecasts are expected to deliver higher resolution and reduce uncertainty, making flight conditions more predictable and enabling more accurate real-time predictions of comfort, cost, and safety-margin metrics, benefits that will be valuable for Air Mobility Operations.

For AI Methods, this work represents a key contribution to VARIANT by establishing the data foundations needed to apply artificial intelligence to one of the most variable and critical factors in aviation operations: the weather.

VARIANT is partially funded by the Sub directorate General for Technological Innovation of the Madrid Regional Government.

 [