The Open AI Drone Ecosystem

Do you remember when AOL tried with its walled garden to compete with the Internet? Or when Steve Jobs introduced the iPhone and said there will never be an App Store? There are benefits to a closed, proprietary system, but less so in the robotics industry of heterogeneous devices and software, when an open ecosystem is crowdsourced to be greater than what any tech giant can build on its own.

For example, PX4 is a popular robotics platform, with many features. It's focused on human pilots, but we at Astral apply its strengths to autonomous robots. Before flight, you can configure your drone and set the flight path with QGroundControl. And it works for many robots, not just quadcopters, not just flying robots, and not just robots you buy from us. (We also support PX4's competitor open-source platform, ArduPilot).

After the flight, access your drone's flight log to view the power consumption, CPU usage, vibration, temperature, etc. as well as see where your drone actually went.

Even a few proprietary drones, like DJI, added support for QGroundControl, but what if you wanted to fly your drone with your own software, or what if you want to modify the drone that you bought with, I don't know, frickin' lasers? Unless DJI lets your great new idea into their walled garden, you can't. Just give up.

Well what if - dreaming big now - you wanted to run a mobile+cloud+simulator for a drone that you built with off the shelf parts, but first try it out in a simulator, and also use AI for features like computer vision and LLM reasoning? Not unless you have a magic lamp, or Astral. Then don't give up, just go on.

We didn't build this all by ourselves, we leveraged available software and hardware to build a Drone App Store, and an example drone to use it. We added AI before it was cool (so a year ago). We took openness and ran with it. We are your magic lamp. 🧞‍♂️