Game of Drones – Competition at NeurIPS 2019

Game of Drones – Competition at NeurIPS 2019

About the competition

Push the boundaries of competitive autonomous systems

Game of Drones is a NeurIPS 2019 competition with the goal to push the boundary of building competitive autonomous systems through head-to-head drone races.

The teams will race a quadrotor drone in simulation through a series of gates against an opponent. There will be two drones on the same track at the same time that are allowed to block each other. The goal is to finish the race quicker than the opponent.

The competition uses AirSim, which is an open-sourced simulation platform for robots. AirSim was created to help accelerate research in deep reinforcement / imitation learning, computer vision and robotics for autonomous vehicles.

What’s special about this competition?

Previous drone race challenges, such as the IROS autonomous drone race challenge or the AlphaPilot challenge only need you to race against the clock.  In contrast, here, you’ll have to race against an opponent on the same track. This means you can block your opponent if you are ahead and you have to overtake him when you are behind creating a whole range of interactions that your policy has to handle.

  • To broaden the audience and bring people to drones from all kind of disciplines, all racing will be done in simulation. However, we still require you to solve the challenge of gate tracking from simulated cameras.

    The competition comes in three tiers of different difficulty and level of autonomy.

    • Tier 1: Planning only.
      Given: Ground-truth poses of your and your opponent’s drone.
      Challenge: Finish the track quicker than your opponent, without crashing.
    • Tier 2: Perception only.
      Given: Camera images (RGB from front and downward cameras) and ground-truth pose of your drone.
      Challenge: Finish the track as quick as possible, without crashing.
    • Tier 3: Full autonomy, combined challenge of Tier 1 and 2.
      Given: Camera images (RGB from front and downward cameras) and ground-truth poses of your but not your opponent’s drone.
      Challenge: Finish the track quicker than your opponent, without crashing.

Why should I participate/care about the game of drones?

Besides being fun, this competition provides a fair comparison for the current state of the art.

If your current focus is either more on the planning or perception side, you’ll most probably learn a thing or two. The audience is expected to be large — NeurIPS is the largest AI conference after all, so you can expect to get some attention if your algorithm scores high.

Prizes will also be awarded to the highest performing teams – we will announce them in the following weeks!

Register your team here. Please make sure that your team’s registration form is complete before submitting. Once submitted via your registration form, accepting any changes to registration details will be at the sole discretion of the organizers.

*New* AirSim environments and detailed guidelines for the competition are available here.

Schedule

 

  • July 15, 2019: Release of competition environments for training purposes.
  • August 1, 2019: Release of reference algorithms.
    We’ll release our submission as an opponent in the online interface to benchmark against. Note that the code will not be made available.
  • August 15 – 21st November, 2019: Qualification Round.
    All submissions made in the timeframe above will be considered for qualification.
  • August 31, 2019: Freeze of reference algorithm.
    Last possible change to the reference submission, to allow participants enough
    time to adjust their algorithms.
  • 15th October, 2019: Detailed qualification rules.
    The detailed qualification rules will be released, specifying what exactly has to be passed to be qualified for the placement rounds.
  • 21st November, 2019: End of qualification.
    Announcement of qualified teams for the live challenge, invitation for poster presen-
    tation.
  • December, 2019: Live Competition at NeurIPS 2019.
    All present qualified teams participate in the live competition.

Organizing Team

Who is behind this?

The organizing team behind the game of drones are the developers of AirSim from Microsoft Research and robotics researchers from Stanford who share the vision of a simulator realistic enough to provide a training and testing environment for autonomous mobile robots such as drones or self-driving cars.

Many people are helping to make this happen, but the core team consists of:

  • Matthew Brown, AirSim Technical Development
  • Guada Casuso, Principal Technical Product Manager, Microsoft
  • Eric Cristofalo, Ph.D. at Stanford University in the Department of Aeronautics and Astronautics
  • Darius Garza, UI/Product Designer, Microsoft
  • Nicholas Gyde, AirSim Technical Development
  • Ashish Kapoor, Principal Research Manager, Microsoft
  • Ratnesh Madaan, Research Software Development Engineer, Microsoft
  • Keiko Nagami, Graduate Student, Stanford University
  • Jim Piavis, Principal Program Manager working with the AirSim AI/ML simulation platform, Microsoft
  • Davide Scaramuzza, Professor and Director of the Robotics and Perception Group, University of Zurich
  • Mac Schwager, Assistant Professor and Director of the Multi-Robot Systems Lab (MSL), Stanford University
  • Tim Taubner, Masters student from ETH Zurich currently visiting MSL

For any inquiries please see our FAQ section or send us an email.

FAQs

 

Register your team here. Please make sure that your team’s registration form is complete before submitting. Once submitted via your registration form, accepting any changes to registration details will be at the sole discretion of the organizers.

How do I register?

Register your team here. Please make sure that your team’s registration form is complete before submitting. Once submitted via your registration form, accepting any changes to registration details will be at the sole discretion of the organizers.

How do I submit my entry to the challenge?

We will be publishing the detailed procedure by July 31st 2019.

Do you have some example code?

We will be releasing reference baselines for all the three tiers by July 15th, 2019. You can use these reference baselines as an optional starting point.

What platforms / languages can I use?

Airsim clients can be run on Windows and Linux. The reference baselines will be in Python.

Can I publish results of my experiments?

Yes, we encourage academic publications relating to the challenge. If you do decide to publish results, please cite AirSim platform as:

@inproceedings{airsim2017fsr,
author = {Shital Shah and Debadeepta Dey and Chris Lovett and Ashish Kapoor},
title = {AirSim: High-Fidelity Visual and Physical Simulation for Autonomous Vehicles},
year = {2017},booktitle = {Field and Service Robotics},
eprint = {arXiv:1705.05065},
url = {https://arxiv.org/abs/1705.05065} }

What can I win?

Prizes will be awarded to the highest performing teams – we will announce them in the following weeks!

Where can I get help?

For any inquiries please see our FAQ section or send us an email

Application form

Prizes

Prizes for Game of Drones

Category 1

  • First Prize: Drone kit ($1600)
  • Second Prize: FPV Goggles ($500)
  • Third Prize: FPV Goggles ($300)

Category 2

  • First Prize:  Drone kit ($1600)
  • Second Prize: FPV Goggles ($500)
  • Third Prize: FPV Goggles ($300)

Category 3

  • First Prize: Drone kit ($2500 )
  • Second Prize: Drone kit ($2000)
  • Third Prize:  Drone kit ($1600)

Drone Kit

Drone Kit

Fat Shark Dominator HDO FPV Goggles

Fat Shark Dominator HDO FPV Goggles