Skip to main contentSkip to footer
Logo IT4LIA AI factory

A prova di AI podcast: 3B Meteo and the AI that sees the rain before it arrives

“A prova di AI” is IT4LIA AI Factory’s podcast about how artificial intelligence is actually applied inside Italian organisations. Every episode, a real case: a concrete challenge, the decisions made along the way and the results achieved.

In each episode, entrepreneurs, decision makers, institutional representatives and innovators share their challenges, objectives and choices, with the support of the IT4LIA AI Factory network.

The episode: 3B Meteo

This episode takes the podcast into the world of weather forecasting. The protagonist is 3B Meteo, one of Italy’s leading companies in the sector, with a team of about 40 people across modelling, meteorology, development and UX design, and media departments.

The conversation features Francesco Bracci, head of the modelling department at 3B Meteo, and Gabriele Franch, researcher at Fondazione Bruno Kessler, together with Matteo Angelinelli, COO of IT4LIA AI Factory and HPC Data Engineer at CINECA.

A product that already worked

3B Meteo already offered a very short-term forecasting service, based on data from a network of radars distributed across the country. Users can follow the movement of precipitation in real time and receive notifications about rain, snow and hail.

The models’ output is then interpreted by the meteorology department, which turns it into the final forecast that reaches users.

The project discussed in the episode starts from a clear goal: having the best possible product. The company chose to use artificial intelligence to improve the performance of an already appreciated service, with forecasts covering from a few minutes up to 60-90 minutes ahead.

How the collaboration began

The collaboration is rooted in a relationship built over the years. Francesco Bracci already knew Gabriele Franch through conferences on radar meteorology and, two years ago, had attended an academy organised by Fondazione Bruno Kessler on the use of AI for weather forecasting. When 3B Meteo came up with the idea of a new AI-based model built on radar data, the first contact was with him, and he introduced the company to IT4LIA AI Factory and its computing resources.

From research to product

A central role is played by the collaboration with Fondazione Bruno Kessler, a research centre based in Trento that supports Italian companies in integrating AI models into Earth sciences and weather forecasting.

Thanks to access to the Leonardo supercomputer, together with the datasets and model architectures made available within the project, the 3B Meteo team was able to train its model in a reasonable time, which would have been unthinkable with the computing resources available in-house.

Fondazione Bruno Kessler also brings a method to this journey. The centre makes datasets and models available, publishes the results of its research and organises training events, such as the annual AI for Weather and Climate school, so that companies find a ready starting point for their own projects. The goal is for those who complete a journey with IT4LIA AI Factory to be able to build and maintain their own systems independently over time.

The episode also looks at what it means for a company to start working hands-on with AI, from retrieving data to training, and at how the dialogue between those who develop the product and those who do research creates a virtuous circle, where every result opens new questions. A step that already looks further ahead: improvements in very short-term forecasting can also benefit other services, such as the one dedicated to hail.

Listen to the episode now

“A prova di AI” is available on YouTube and Spotify. Listen to the episode with 3B Meteo now.

Do you have a case to share or want to understand how IT4LIA AI Factory can support your project? Find out what we can do for you.

90 minutes ahead: the challenge od weather forecasting with AI

A prova di AI podcast: 3B Meteo and the AI that sees the rain before it arrives

90 minutes ahead: the challenge od weather forecasting with AI

“A prova di AI” is IT4LIA AI Factory’s podcast about how artificial intelligence is actually applied inside Italian organisations. Every episode, a real case: a concrete challenge, the decisions made along the way and the results achieved.

In each episode, entrepreneurs, decision makers, institutional representatives and innovators share their challenges, objectives and choices, with the support of the IT4LIA AI Factory network.

The episode: 3B Meteo

This episode takes the podcast into the world of weather forecasting. The protagonist is 3B Meteo, one of Italy’s leading companies in the sector, with a team of about 40 people across modelling, meteorology, development and UX design, and media departments.

The conversation features Francesco Bracci, head of the modelling department at 3B Meteo, and Gabriele Franch, researcher at Fondazione Bruno Kessler, together with Matteo Angelinelli, COO of IT4LIA AI Factory and HPC Data Engineer at CINECA.

A product that already worked

3B Meteo already offered a very short-term forecasting service, based on data from a network of radars distributed across the country. Users can follow the movement of precipitation in real time and receive notifications about rain, snow and hail.

The models’ output is then interpreted by the meteorology department, which turns it into the final forecast that reaches users.

The project discussed in the episode starts from a clear goal: having the best possible product. The company chose to use artificial intelligence to improve the performance of an already appreciated service, with forecasts covering from a few minutes up to 60-90 minutes ahead.

How the collaboration began

The collaboration is rooted in a relationship built over the years. Francesco Bracci already knew Gabriele Franch through conferences on radar meteorology and, two years ago, had attended an academy organised by Fondazione Bruno Kessler on the use of AI for weather forecasting. When 3B Meteo came up with the idea of a new AI-based model built on radar data, the first contact was with him, and he introduced the company to IT4LIA AI Factory and its computing resources.

From research to product

A central role is played by the collaboration with Fondazione Bruno Kessler, a research centre based in Trento that supports Italian companies in integrating AI models into Earth sciences and weather forecasting.

Thanks to access to the Leonardo supercomputer, together with the datasets and model architectures made available within the project, the 3B Meteo team was able to train its model in a reasonable time, which would have been unthinkable with the computing resources available in-house.

Fondazione Bruno Kessler also brings a method to this journey. The centre makes datasets and models available, publishes the results of its research and organises training events, such as the annual AI for Weather and Climate school, so that companies find a ready starting point for their own projects. The goal is for those who complete a journey with IT4LIA AI Factory to be able to build and maintain their own systems independently over time.

The episode also looks at what it means for a company to start working hands-on with AI, from retrieving data to training, and at how the dialogue between those who develop the product and those who do research creates a virtuous circle, where every result opens new questions. A step that already looks further ahead: improvements in very short-term forecasting can also benefit other services, such as the one dedicated to hail.

Listen to the episode now

“A prova di AI” is available on YouTube and Spotify. Listen to the episode with 3B Meteo now.

Do you have a case to share or want to understand how IT4LIA AI Factory can support your project? Find out what we can do for you.