AirSim Drone Racing Lab

  • Ratnesh Madaan ,
  • Nicholas Gyde ,
  • Sai Vemprala ,
  • Matthew Brown ,
  • Keiko Nagami ,
  • Tim Taubner ,
  • Eric Cristofalo ,
  • Davide Scaramuzza ,
  • Mac Schwager ,
  • Ashish Kapoor

Download BibTex

Autonomous drone racing is a challenging research problem at the intersection of computer vision, planning, state estimation, and control. We introduce AirSim Drone Racing Lab, a simulation framework for enabling fast prototyping of algorithms for autonomy and enabling machine learning research in this domain, with the goal of reducing the time, money, and risks associated with field robotics. Our framework enables generation of racing tracks in multiple photo-realistic environments, orchestration of drone races, comes with a suite of gate assets, allows for multiple sensor modalities (monocular, depth, neuromorphic events, optical flow), different camera models, and benchmarking of planning, control, computer vision, and learning-based algorithms. We used our framework to host a simulation based drone racing competition at NeurIPS 2019. The competition binaries are available at our github repository.

Harnessing high-fidelity simulation for autonomous systems through AirSim

Robots and autonomous systems are playing a significant role in modern times, in both academic research and industrial applications. Handling the constant variability and uncertainty present in the real world is a major challenge for autonomous systems as their areas of usage expand. Recently, machine learning techniques, such as deep neural networks, have shown promise as building blocks for improving robot intelligence, and high visual and physical fidelity simulation has the potential to address the needs of data-driven autonomy algorithms.

In this webinar, Sai Vemprala, a Microsoft researcher, will introduce Microsoft AirSim, an open-source, high-fidelity robotics simulator, and he demonstrates how it can help to train robust and generalizable algorithms for autonomy. He will explain the features of Microsoft AirSim while giving an overview of some research projects that have benefited from AirSim, particularly focusing on robotics and how these algorithms are trained with simulated data but are capable of working in real life. He will also introduce AirSim Drone Racing Lab, an enhancement of AirSim aimed at enabling robotics and machine learning researchers to tackle the specific domain of autonomous drone racing.

Together, you’ll explore:

  • How simulation can address the needs of data-driven autonomy algorithms
  • General features and usage of Microsoft AirSim
  • How robotics research projects have employed AirSim for training AI models capable of sim-to-real transfer
  • How you can get started with the AirSim Drone Racing Lab and use it to generate data for perception, planning, and control algorithms for autonomous drones

Resource list:

*This on-demand webinar features a previously recorded Q&A session and open captioning.

Explore more Microsoft Research webinars: https://aka.ms/msrwebinars (opens in new tab)