If DJI ever comes around and unleashes the RM S1, starting with the ability to organize multiplayer matches into teams, we will be able to start thinking bringing RoboMaster System One events to mimick actual RoboMaster tournaments by having as many RoboMaster robot types as we can.
One of those robot types that is foreseable is the Aerial Robot. This robot could use the DJI Tello EDU as its foundation. Like the RM S1, the Tello EDU can either serve as a WiFi Access Point for a Controller, or as a Node connecting to an external WiFi Router.
The DJI Tello EDU can be programmed in Blocks, JavaScript or Python, the latter being the more interesting for us, for obvious reasons.
It's SDK is fully open and includes many more modules and functions than the RM S1.
DJI Tello EDU SDK 2.0
With its ability to broadcast its video stream, the DJI Tello EDU can integrate into all kinds of AI applications that could run on a Ground Control Station such as S1 Robot Recognition and Team (Friend or Foe) designation.
Another really nifty feature of a RM System One Aerial DJI Tello EDU would be to use the Tello as "eyes in the sky" with its video broadcast being routed by the Referee Station to either one Master Playing Field view monitor viewable by all team members, or as an individual window on a team member's FPV display.

This upcoming week-end, I will bring Chip and Dale and my Tello EDU to a family reunion. We are expecting to have access to a small gym in which I will set-up matches in which I want to include the Tello EDU. Videos to come!

Nvidia has some really good collision avoidance tech recently put into this new flying drone:

https://news.developer.nvidia.com/skydio-2-jetson-tx2-drone

This even comes with a "no crash or your money back" guarantee.

Let's say it would take too much software at this point to integrate flyers into "System One" events - but for sure, the next gen should be able to give us some options.

www.skydio.com

MarkusXL Posted at 10-3 06:10
Nvidia has some really good collision avoidance tech recently put into this new flying drone:

https://news.developer.nvidia.com/skydio-2-jetson-tx2-drone

So COOL!
There was some pretty crazy stuff shown last Spring at the IEEE International Conference on Robotics and Automation (ICRA) in Montreal. Everybody is moving to stereo vision supported by an Inertial
Measurement Unit (IMU) to perform real-time mapping. obstacle detection and avoidance and navigation.
DJI were there and discussed the latest breed of the Phantom 4 Pro and the Mavic Air Pro. The Mavic was shown detecting and avoiding objects as small as branches.

The NVidia family of GPUs and processors were all over the place.

Trying to put in a GPU with hundreds of Tensor Cores is extra challenging for aircraft, as these large silicon arrays consume power and require cooling, which means added mass and less flight time which is bad for aircraft.  

BUT - you can lug around a powerful GPU on an S1, power it, cool it, and still have decent run time.  

The advantages are huge - great frame rate for processing video - so multiple cameras welcome.  With the right software, get collision avoidance, scary fast targeting and aiming, friend / foe detection - I'm really looking forward to it.

Hello and good day rhoude57 - YUL. Thank you for reaching out and for sharing this information with is. Your insights with regards to this matter is greatly appreciated. Have a great programming and a happy driving always. Thank you.