Project AirSim’s high-fidelity environments allow AI models to learn through trial and error which actions to take at each phase of a flight
Microsoft has released Project AirSim, a new platform running on Microsoft Azure that builds, trains and tests autonomous aircraft through high-fidelity simulation.
The project uses artificial intelligence (AI) to generate huge volumes of data. It then uses this information and machine learning technology to train AI models how to react to variables in the physical world – such as weather conditions, battery failure and maintenance of camera quality – whilst autonomous aircraft are in flight.
The project builds on the earlier AirSim project, which was originally created by Microsoft Research’s Ashish Kapoor, that has now been fully transformed into an end-to-end platform that allows advanced aerial mobility customers to easily test and train AI-powered aircraft in simulated 3D environments.
“Our ability to capture data and translate it into autonomy is going to significantly change the landscape of aviation,” said Kapoor, now general manager of Microsoft’s autonomous systems research group. “And because of that we are going to see many more vehicles in the sky, helping to monitor farms, inspect critical infrastructure and transport goods and people to the remotest of places.”
Project AirSim uses data from Bing Maps and other providers to create millions of detailed 3D environments. Microsoft is also working with industry partners such as Ansys and MathWorks to enable users to accurately simulate different weather conditions, changes of force on the aircraft and sensors that an autonomous machine uses to understand its environment.
“We created Project AirSim with the key capabilities we believe will help democratise 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,” said Balinder Malhi, engineering lead for Project AirSim.
Project AirSim offers users access to pre-trained AI building blocks so developers will no longer need deep machine learning expertise to develop advanced models.
“Autonomous systems will transform many industries and enable many aerial scenarios, from the last-mile delivery of goods in congested cities to the inspection of downed power lines from 1,000 miles away,” said Gurdeep Pall, corporate vice president for business incubations in technology and research at Microsoft. “But first we must safely train these systems in a realistic, virtualised world. Project AirSim is a critical tool that lets us bridge the world of bits and the world of atoms, and it shows the power of the industrial metaverse – the virtual worlds where businesses will build, test and hone solutions and then bring them into the real world.”
Airtonomy, which participated in an early access programme for Project AirSim, has already used the technology to help customers to launch remote inspections of critical infrastructure quickly and safely. The drone inspects the asset automatically, captures data and contextualises it at the moment of capture.
Microsoft plans to work with global civil aviation regulators to explore how Project AirSim might help with the certification of safe autonomous systems.
Find out more about Project AirSim.
Share this story