International Data Battle 2026

📢 6 000 € euros cash prize ! 📢

You are a student in computer science or applied mathematics ? 
Ready for a challenge in data science ?

🚀 Join us at the 5th edition of the IA Pau Data Battle  🚀

➡ 2nd March 2025 to 27th March 2026, 100% remote ⬅

A Verdinum sponsored project

Registration is mandatory so that we can send you the data of the Battle

📌 What is the Data Battle ?

🛠️ Objective : Solve a real problem with data science techniques 

Registered students form teams (3 to 6 people) and confirm their participation before March 5, 2026.

The teams develop a solution and will present it on March 27, 2026.

📌 DB26 Program

🛠️ This year : Data Battle 2026 offers you the opportunity to showcase your skills in the world of meteorology. For 3 weeks,

Your mission : Predicting the evolution of a storm, more specifically estimating the probability of its end, is crucial. In particular, certain sensitive areas, such as airports, must closely monitor the end of storms to resume operations, and every minute counts.

Current technologies already allow for excellent anticipation of the arrival of a thunderstorm. However, determining its exact end time remains complex. Currently, alerts remain active for a fixed duration, typically 30 to 60 minutes after the last lightning strike in the monitored area. For certain sectors, particularly airports, where every minute counts, this method has its limitations. Your objective will therefore be to develop a probabilistic model capable of estimating the actual end of a thunderstorm by analyzing the spatiotemporal dynamics of lightning within a 50 km radius of several European airports.

A second focus is to analyze thunderstorm trends for each airport. Indeed, each location has its own specific geographical and meteorological conditions. An interesting analysis would be to identify the specific characteristics of each location and the main types of thunderstorms present in the data.

Using meteorage data, your objective will therefore be to provide the analyses that allow meteorage clients to minimize their exposure to lightning!

 

📌Our Partner METEORAGE

This year, the company  METEORAGE is the leader of the project at the DATA BATTLE 2026.

Meteorage is a company that, for 40 years, has deployed a network of sensors across Europe to anticipate, prevent, and manage storm risks. These antennas are capable of detecting lightning from hundreds of kilometers away. Triangulation techniques then determine the precise location and date of each strike. This knowledge allows Meteorage to help its clients mitigate the harmful effects of storms and lightning on their operations, enabling anticipation, adaptation, and improved risk prevention.

📌 With which data ?

🛠️ ⚡They consist of the distribution of 230,000 lightning strike points (location and date) within a 50 km radius of 6 airports over a 10-year period. This represents hundreds of thunderstorm events. This tabular data is relatively lightweight, thus avoiding the complexity of Deep Learning applications, while still presenting very interesting data science challenges involving the characterization of spatio-temporal correlations, across various geographically dependent trends, and an objective that is not classification, but the construction of a risk probability. More precisely, it involves:

➡️ Localisation of each strike

➡️ timestamps,

➡️ defined area around 50km of the point of interest (the airport),

➡️ Hundred of storms

How do you define an alert ?

A 20 km radius circle is defined around a point representing an airport. The alert begins when a lightning strike occurs within this zone. The alert ends after the first 30-minute period without a lightning strike in this zone. It is this end-of-alert time that must be estimated to win the data battle.

📅 The program in details (subject to change depending on the number of registration)

✔️ Every Friday : Progress questionnaire. These allow us to track the progress of the teams, and will contribute to the final evaluation in addition to the presentation before the Jury.

✔️ Thursday 12th of Marth 18h : Seminar on the evaluation 

  • lien visio to come

    – How the jury will evaluate and choose the winners

   – How to study the socio-environmental impact of your project

✔️ Every Thuesday 18h : General video conference with the mentors, feedback on Friday’s questionnaire.

✔️ From the 26th of March : No more coding, let’s defend your project and win the prize!

Friday 3rd April 12h-14h :First round of Defence

Friday 3rd April 18h : Final

The AI ​​Pau team will be available to answer your questions via a Discord channel (link).

Follow this link for more details about the program and rules. Please note that registration implies acceptance of these rules.

https://docs.google.com/document/d/1L0O6fRy2URh3hztAtburRgDxLTHs7nUR1LKJHHk-T5g/edit?usp=sharing

🏆 Awards

🥇 3 000 € for the winning entry

🥈 2 000 € for the second

🥉 1 000 € for the third

🎯 Why participate ?

🚀 Boost your CV and win a unique experience

🔍 Develop your skills in AI and data science. 

🤝 Meetup experts and grow your network.

💼 Get ahead and in the top position on labor market.

Programme

10 mars 2025

- 18h00
Lancement du DB25
Présentation du sujet aux équipes, explication des règles et réponse aux questions.

14 mars 2025

- 20h00
Bouclage des équipes
Dernier jour pour inscrire votre équipe (pour les retardataires !) au Data Battle 2025.

28 mars 2025

- 12h00
Rendu des projets
Votre projet, après 3 semaines de travail et de challenge, est à rendre aux jurys !
- 14h00
Poules
Les jurys de poule sélectionnent les lauréats pour atteindre la phase finale !
Étiquettes :

Date

02 - 27 Mar 2026
Ongoing...

Heure

18h00 - 19h00

Organisateur

IA PAU
Site Web
http://iapau.org/

Poster le commentaire

Votre adresse e-mail ne sera pas publiée. Les champs obligatoires sont indiqués avec *