Robomaster S1 is still very limited in programming possibilities. Despite using the Python language, the robot API is very closed on things that are already ready. What is the possibility of including third party libraries? When we will have the possibility of direct manipulation of the features of the robot, camera, sensors, memory card, etc. And how to implement third party sensors? Third party library?


The robot would be much more interesting if it were possible to include third party library, use arduino and raspberry sensors, write to files, etc. The documentation itself needs to improve. These things would even make the Robomaster S1 not look like an abandoned product.


What do forum colleges think about these limitations on robot manipulation, both physical and logical? Is there a way around such limitations, especially programming limitations?

I second the motion!
The RoboMaster S1 breaks the mold created by every other DJI products who all offer a great, well documented SDK. Even the Tello, developped in partnership with Ryze, has a well documented SDK.
DJI, IMHO, need to correct this with the highest of priorities.

  

Same for me, but at least there is some hope. In my last Firmware comparison I've found some hints in the code that documentation will be updated in the future:

  

TODO LIST:

  

• 1. add framework doc
• 2. add plaintext sdk doc
• 3. add custom UI doc
  

  

Can't say if this is just the DJI internal docs.

  

Cheers

  

Well, there is an external APi coming up. It is just a matter of time. But there will always be limitations imposed by DJI. Some due to legal concerns, some simply because they don't want to give us free rein over the device.

Yes I'm sure the fear is fly-by-night manufacturers doing huge runs of cheap clones then disappearing.  Nobody wants that, it harms even us consumers, and victimizes hapless buyers with broken stuff.

That said, we've been hitting the walls of our programming sandbox and finding it a bit small.  For example while I really like the new audio features offered by the update, but I still can't code my own custom sounds, having no direct API access to the sound chip (built into the ARM processor) itself.

Also I really desire to code for obstacle avoidance and 'bot to 'bot communications.  I hope somehow they find a way to let us do more "real robot stuff".

Any news about the S1 SDK? is it already possible to use the plaintext SDK on S1? I still don't understand why such a simple update, which could even impact the acquisition of new robots, and not abandon the platform, takes so long to become a reality. I hope it evolved once I decided to reactivate my S1 for some educational activities.