Microsoft Expands AirSim AI Simulator to Include Autonomous Car Research

Home > News > Content

【Summary】Microsoft’s artificial intelligence simulator AirSim which was originally designed for drones, has been expanded to include autonomous car research.

Original   Eric Walz  ·  Nov 27, 2017 2:05 PM PT
author: Eric Walz   

REDMOND, WA — Microsoft's artificial intelligence simulator which was originally designed for drones has been expanded to include autonomous car research. The company has announced that its Aerial Informatics and Robotics Platform called AirSim, will be used for testing the safety of autonomous vehicle artificial intelligence systems.

The updated AirSim platform provides a detailed three-dimensional urban environment where conditions can be varied to allow safety features to be fully tested. This includes having operating traffic lights; aspects of an urban landscape including parks, lakes, and construction sites.

The AirSim platform solves for two problems: the large data needs for training, and the ability to debug in a simulator. It will provide realistic simulation tools for designers and developers to seamlessly generate the copious amounts of training data they require.

In addition, the platform leverages recent advances in physics and perception computation to create accurate, real-world simulations. Together, this realism, based on efficiently generated ground truth data, enables the study and execution of complex driving maneuvers that might be time-consuming or too risky to test out in the real-world.

Microsoft's AirSim is a open source research project for testing the safety of artificial intelligence systems and was made available earlier this year. It provides realistic environments, vehicle dynamics and sensing for research into how autonomous vehicles that use AI that can operate safely in the open world. AirSim creates a training environment in which it's possible to simulate mistakes, without the real-world consequences.

The update extends the system to include car simulation, which will help advance the research and development of self-driving vehicles. The latest version is available now on GitHub as an open-source, cross-platform offering.

AirSim Adds Additional Support for Drones

The updated version of AirSim also includes many other features and enhancements, including additional tools for testing airborne vehicles. Microsoft has made it easier for people to simulate flying drones by adding a built-in flight controller, called ‘simple_flight', that simplifies the drone setup process. This allows rapid experimentation with control and state estimation algorithms without requiring expensive debugging and development in the embedded world.

Testing Self-Driving Cars in Simulation With AirSim

Developing algorithms for self-driving cars is an expensive proposition. It requires infrastructure to build expensive hardware platforms, large amounts data, and the ability to quickly test and benchmark results.

Microsoft's goal was to make these various aspects of developing self-driving cars available to a broader group of researchers by providing an open, community-driven platform for testing those algorithms. The new version of AirSim includes car simulations, new environments, APIs to ease programming and ready-to-run scripts to jump start autonomous driving research.

AirSim comes with a detailed 3D urban environment that includes a variety of diverse conditions, including traffic lights, parks, lakes and construction sites. Users can test their systems in several types of neighborhoods, including downtown, semi-urban, vegetation and industrial environments. The simulation contains more than 12 kilometers of drivable roads spanning more than 20 city blocks.

Unreal Engine Plugin

AirSim has been developed as a plugin for Unreal Engine, a popular tool for video building video games, simulations, and visualizations. This means that the car simulation is decoupled from the environment it runs in. Developers can create an environment for specific needs, such as a city or rural road, or choose from a variety of environments available online, and then simply drop in the AirSim plugin to test various self-driving algorithms in that environment. AirSim extensibility also allows researchers and developers to incorporate new sensors, vehicles or even use different physics engines.

Turnkey AI research

AirSim also provides developers APIs that can be used in a wide variety of languages, including C++ and Python. This makes it easy to use AirSim with various machine learning tool chains. For example, Microsoft's Cognitive Toolkit (CNTK) can be used with AirSim to do deep reinforcement learning. Microsoft also sees significant opportunities with running multiple instances on AirSim with Microsoft's Azure cloud computing platform to scale up the training for modern, data-hungry machine learning algorithms.

AirSim is now available as compiled binary release, which means users can now download and start calling its Python APIs to control an autonomous vehicle in just minutes.

As a company, Microsoft is fortunate to have great support and ideas from a growing community of AirSim aficionados all over the world. This includes internal teams, such as the Microsoft Garage team, as well as a number of others in the AI community. In future releases, Microsoft hopes to add new sensors, better vehicle physics, weather modeling, and more detailed realistic  driving environments for testing autonomous vehicles.

Prev                  Next
Writer's other posts
    . Related Content