civic

Pushback to the Future: Predict Pushback Time at US Airports

An image of an aircraft being pushbacked by a tug.

Help the National Airspace System (NAS) keep flights running on time! In this challenge you will use air traffic and weather data to automatically predict the time an aircraft pushes back from its gate. Accurate estimates of pushback time can help air traffic management systems more efficiently use the limited capacity of airports, runways and the National Airspace System.

This challenge features two competition arenas:

  • Phase 1: Open Arena is available for all participants to work with the data and test their solutions to see how they fare against others on the open leaderboard.
  • Phase 1: Prescreened Arena is available to prize eligibile participants to build out and submit code for their solutions.
  • Phase 2 is available to prize eligibile finalists from Phase 1. To qualify for a prize, finalists will develop their winning models into a federated learning models.

For this challenge, cash prizes are restricted to individual participants or team leads who are a U.S. citizen or permanent resident and are affiliated with an accredited U.S. university.

Results

The challenge has ended! Phase 1 of the challenge generated over 450 submissions to predict pushback times. The top submission predicting pushback time achieved a mean absolute error of just over 10 minutes, compared to over 19 minutes for the benchmark. Phase 2 saw the top 5 teams incorporate federated learning into their winning solutions.

You can read more about the results of the challenge and meet the winning teams in the Meet the Finalists blog post and learn from their winning code and write-ups in the competition repo!


MEET THE FINALISTS

WINNING CODE AND WRITE-UPS ON GITHUB


The competitions

PRE-APPROVAL NEEDED
civic

Phase 2: Federated Learning

Welcome Phase 1 finalists to Phase 2 of the Predict Pushback Challenge! Now's your chance to explore translating your winning Phase 1 model into a decentralized federated learning model. #civic

13 joined
jul 2023
competition has ended
civic

Phase 1: Open Arena

START HERE! Accurate estimates of pushback time can help air traffic management systems more efficiently use the limited capacity of airports, runways and the National Airspace System. Use air traffic and weather data to automatically predict pushback time! #civic

409 joined
apr 2023
competition has ended
PRE-APPROVAL NEEDED
civic

Phase 1: Prescreened Arena

CALLING PRESCREENED PARTICIPANTS! Accurate estimates of pushback time can help air traffic management systems more efficiently use the limited capacity of airports, runways and the National Airspace System. Use air traffic and weather data to automatically predict pushback time! #civic

49 joined
$50,000 in prizes
apr 2023
competition has ended
$50,000