PRC Data Challenge 2025

The Performance Review Commission (PRC) with the support of OpenSky Network aims at engaging with data scientists, even without Aviation background, to team up and compete to solve a (big-ish) Data Challenge of interest to the wider Aviation community using open data and openly sharing final solutions.

PRC Data Challenge

The 2025 edition of the PRC Data Challenge builds on top of the 2024 edition and proposes to build Machine Learning (ML) models to predict the amount of fuel burnt at certain intervals of the flight. The training dataset will consist of flight intervals and corresponding fuel burnt/flown amounts.

The competition runs (in two phases) from 1st Oct till 30th Nov 2025.
The total amount for the first 3 teams is 5000 EUR.
Check the conditions for eligibility for prize and then fill the following team creation request form!

ImportantImportant

And the winners are…

see the dedicated page

ImportantImportant

Full dataset now openly available

The dataset (training and ranking) used for the organization of the challenge is available at
https://doi.org/10.5281/zenodo.19184661

and is described in a data paper openly accessible at
https://doi.org/10.59490/joas.2026.8750

Remember to follow up the data challenge on OSN’s Discord channel #prc-data-competition.
(if not already done, you need to access OSN Discord server first)

Below you can find a couple of plots of flights showing the ACARS trajectory in red and the ADS-B one in blue.

(a) well coverage
(b) less complete coverage
Figure 1: Flight trajectories in the challenge dataset.

This is a data challenge that uses real data with all the inconsistencies, noise, holes of the real world: it is a challenge because it is tough in its goal and in the available data, and not a mere tutorial or synthetic exercise.

Thanks

The organization of this challenge would have not been possible without the creativity and dedication of Junzi Sun, the proficiency of John Fitzgerald and the help of the Friends of OSN, Allan, Quinten, Martin, Rainer, Xavier and Kamil among them.

OpenSky Network is of course the valuable organization without which we wouldn’t be able to assemble this challenge at all.

Disclaimer

Please note that this challenge is a challenge in its own preparation, so everything is experimental and subject to change at any moment at the organizers’ discretion.
We anyway do our best to prepare for an interesting and quality experience.

List of Acronyms

ML: Machine Learning

PRC: Performance Review Commission