Project AirSim will use Azure to train AI models in a realistic 3D environment
Image: Microsoft through Denis Publishing
Microsoft has unveiled a simulator for training autonomous drones powered by Microsoft Azure to enable millions of complex flight simulations per second.
Project AirSim is the successor to Microsoft’s earlier AirSim and started life as a research project in Unreal, but requires an extensive understanding of machine learning and coding.
The newly announced iteration promises an end-to-end experience that will allow users to use a simple yet powerful tool to train AI models in a 3D environment.
This environment comes from Microsoft’s own Bing Maps and other geographic information sources and can represent specific locations such as New York City or London, as well as custom environments to meet a customer’s individual training needs.
Using Azure to run the simulator allows large-scale testing in the cloud, accelerating customer model training. The availability of pre-trained AI building blocks also promises to accelerate development and prevent vehicle developers from spending time on unnecessary research.
Currently, the project is under development regarding its weather and physics models. For this, Microsoft is partnering with Ansys to implement their high-fidelity physics simulations within Project AirSim.
Software company MathWorks is also contributing to the project, using their Simulink graphical programming environment to help users customize the physical environment.
“Everyone is talking about AI, but few companies are able to build it on a large scale,” explains Balinder Malhi, Project AirSim’s chief engineering officer.
“We have created Project AirSim with the key capabilities that we believe will help democratize and accelerate aerial autonomy, namely the ability to accurately simulate the real world, capture and process massive amounts of data, and encode autonomy without the need for deep expertise in AI.”
Microsoft has indicated that it will work with global civil aviation regulators to decide on Project AirSim’s place in the certification of autonomous systems, evoking the idea of a series of scenarios generated within AirSim that will set a benchmark for autonomous vehicles. could be.
Aerospace manufacturer Bell has already used AirSim to prepare their Autonomous Pod Transport (APT) aircraft for NASA’s Systems Integration and Operationalization (SIO) project, which sought to demonstrate the viability of unmanned aerial vehicles.
Autonomous aircraft are increasingly attracting interest from both the public and private sectors. The British government has announced £105.5 million to be assigned to projects such as Open Skies Cornwall, which will work with the Royal Mail and NHS to make deliveries to the Isles of Scilly using drones, as well as Skyway, which will use drones to survey infrastructure to inform construction projects.
With Project AirSim looking to tackle transportation, supply and air infrastructure inspection, the opportunities for development in the sector seem wide open and promising. It is now available as a limited preview, and access will be expanded in the near future.
© Dennis Publishing