MIT Engineers Build VR System to Train Racer Drones

Engineers at the Massachusetts Institute of Technology have created a virtual reality environment to train drones to fly fast around obstacles.

Dubbed "Flight Goggles," the software causes drones to "see" virtual obstacles in space that is actually clear of physical objects — giving drones a way to fail during flight training without being involved in repeated, time-consuming accidents. It will also offer a testbed for a range of conditions and environments researchers may wish to simulate.

"We think this is a game-changer in the development of drone technology, for drones that go fast," said Sertac Karaman, associate professor of aeronautics and astronautics at MIT, in a prepared statement. "If anything, the system can make autonomous vehicles more responsive, faster and more efficient."

Karaman was inspired to build the system initially by competitive drone racing, a new sport in which humans pilot drones remotely through intricate mazes. Karaman decided to see if he could build an autonomous drone that would beat the human pilots.

"In the next two or three years, we want to enter a drone racing competition with an autonomous drone, and beat the best human player," Karaman said.

Right now, drones are generally trained to fly in enclosed spaces with props like windows to fly through and nets to catch the devices as they lose control. Such a setup is fine for training slow-flying drones, like those used in mapping, but not suitable for training racers, whose flight can vary wildly from small changes in their code. Racers can need repair and even replacement after each high-speed crash, slowing down training and ramping up costs.

"The moment you want to do high-throughput computing and go fast, even the slightest changes you make to its environment will cause the drone to crash," Karaman said in a prepared statement. "You can't learn in that environment. If you want to push boundaries on how fast you can go and compute, you need some sort of virtual reality environment."

The training facility includes a motion-capture system to track the drone's position and orientation as it flies, an image rendering program that can create photorealistic images of physical environments, and tools for beaming the images into the drone's computer as it flies, all to give the drone the illusion of physical obstacles as it flies through an empty room.

Karaman said the system could be used to test out new sensors, as well, by putting them on the drone and seeing if the performance improves or deteriorates in different environments, or to train drones how to fly around humans. For example, a human could walk around in a motion capture suit and that data could be beamed into the drone's computer from a remote location, thus making the person appear near the drone while at no risk of physical harm.

"One day, when you're really confident, you can do it in reality, and have a drone flying around a person as they're running, in a safe way," Karaman said in a news release. "There are a lot of mind-bending experiments you can do in this whole virtual reality thing. Over time, we will showcase all the things you can do."

About the Author

Joshua Bolkan is contributing editor for Campus Technology, THE Journal and STEAM Universe. He can be reached at [email protected].

Featured

  • digital brain integrating legal regulation and security interface

    Agentic AI Moves from Pilot Phase to Production, Bringing Governance to the Forefront

    New research from Caylent, an AI-focused Amazon Web Services Premier Tier Services Partner, found that enterprises are already moving agentic AI beyond pilots and into production environments. At the same time, organizations are putting strict conditions around autonomy, making governance and control the next major challenge for enterprise AI adoption.

  • Silhouettes of business professionals stand against a blurred futuristic city skyline at night, with a glowing digital network data connection

    It's Time for Higher Ed to Get Serious About AI Strategy

    Without a coordinated strategy that involves multiple academic and administrative units across the entire campus, colleges risk wasting resources, duplicating efforts, and ultimately failing to deliver on the promise of deploying technology to improve learning and operations.

  • cloud computing circuit board background

    Report: Cloud Crime on the Rise as Attackers Exploit Trust

    CrowdStrike's 2026 Threat Hunting Report traces a multiyear shift from conventional intrusions toward attacks that exploit trusted identities, cloud services, AI systems and software dependencies.

  • abstract converging light streams and particles

    Nvidia Releases Nemotron 3.5 Lightning Open AI Model

    Nvidia has introduced Nemotron 3.5 Lightning, a new open AI model designed less as an all-purpose answer engine than as a fast workhorse inside long-running AI agent systems.