Lockheed Martin, NVIDIA, and the Drone Racing League have partnered to create AIRR: Artificial Intelligence Robotic Racing.

The gist seems to be that the top team to build an autonomous racing drone will win $1 million, and there is an addition $250,000 available to whoever can first beat a human drone pilot in a race.

I had not known about the Drone Racing League, but the videos look pretty cool.

As a person without much hand-eye coordination, it kind of blows my mind that pilots can navigate these types of environments.

On the other hand, as a person with limited hand-eye coordination, it seems like computers would be much better at this than me. Maybe than everybody?

Microsoft is doing a lot of advance work on flight automation, which might be the next big thing.

Co.Design reports on the work Microsoft researchers are doing both in the field and in simulation:

“Software simulators, with realistic physics just like a video game, offer one appealing alternative to real-world data when it comes to training AI. So before Microsoft put its glider in the real-life sky, it trained it to fly by watching hawks inside a simulator. The team built an open-source software called AirSim for its flight experiments, and over countless trials, various algorithms Microsoft developed learned how to fly like a hawk.”

This seems like a smart move by Microsoft, which largely missed the self-driving car goldrush. Instead of being a late entrant into that field, it’s getting a head start in an even more advanced field.

Microsoft’s Seattle location also works better with flight than it does with the automotive industry. Boeing’s Everett, Washington, aircraft factory is the largest in the world, and presumably a large network of suppliers and talent has grown up around that.

Microsoft also has roots in the flight world, with it’s series of Flight Simulator commercial products, and now its open-source AirSim research tool.

Like something out of a dystopian science fiction movie, a team at Georgia Tech and a team at the Naval Postgraduate School have built swarms of aerial drones that are capable of dogfighting each other.

So far, this seems like a “research” project. For example, the drones are not yet equipped to perceive and identify their targets, so each drone just broadcast its GPS coordinates to its opponents, so they would know where to look.

And the drones aren’t shooting real bullets, yet.

But the future of aerial combat is in sight:

Dogfighting tactics have advanced dramatically since the World War I, but the advent of UAV swarms may bring a brand new set of challenges. Unmanned vehicles have freedom to dive, bank, and climb at rates human pilots cannot tolerate. But the real advantage may be in computing power that could track dozens of adversaries — far more than any human pilot could do — and develop new ways to address challenges.

Also, the Navy itself is testing swarms of autonomous vehicles, with governmental acronyms like LOCUST and CICADA. The Navy is primarily building them for purposes of reconnaissance and diversion, or so they say.