Yes, it's a matter of you not being able to see the inner workings of the DJI go software as well as you not having an understanding of how and what communications are taking place between the flight controller and remote control. If you ever had to build a UAV Ground station or antenna tracker to keep your radio antenna and video receiver antenna directly pointed at your drone or aircraft you would understand this a whole lot more.
If the Follow me mode in DJI did work based purely on the GPS signal, I and as well as others would have already made every attempt to figure out how to install or attach a better GPS unit to the remote controller to achieve getting the drone to actually follow while moving fast, but unfortunately no it's primarily using the GPS only for failsafe and updating the home-point as the remote controller is moving.
Sorry I am getting your ARTVA or is it
ARVA confused with other rescue becons and no satellite communication is not very good under the snow or in adverse weather conditions. The ARVA appears to use radio signal antennas to help rescuers locate the beacon, which doesn't seem to me to be anymore advanced than using signal strength to locate your crashed drone via it's remote control signal or RSSI.
Written communications are notoriously prone to misunderstanding about their tone and conveyed sentiment, but I seem to sense a bit of hostility in your answer. If I'm wrong, I'm sorry. If I'm right, I'm also sorry, because it is certainly not the spirit I intended to give to the discussion. What I'm trying to do here is figuring out those "inner workings" you refer to (which may be unknown, but certainly are no black magic and must obey the laws of physics, don't they?) by discussion and comparison of experiences and knowledge.
You say that it's a matter of me not understanding, which is absolutely correct (that's indeed why I started the discussion, with the intent of improving my understing, not just clarify that I don't have one...). And yet, you offer me no possible explanation that can help me (or you) better understand what's going on.
Here are a few facts and considerations I can offer you.
- to the best of my knowledge, there is no way of identifying the direction of a EM source by a single point measurement of the emitted field. I wouldn't know how to measure or calculate it.
- the case of the ARTVA (or ARVA... I also don't know the exact difference) is indeed very different. As I mention in the first post, the ARVA can't identify, from any point, the direction of the source. All it can do is detect the direction of the field lines in the position of the receiver, and follow them to the source through a curved path (see
Redirect Notice, where the source is in the center: from any point in the drawing all you can do is walk along a red line until you get to the center). This is certainly not what you would like your drone to do in "Follow Me' mode (a side consideration: if there was a smart way of directly identifying the direction of the source, maybe they would use it for such a critical application as avalanche rescue...)
- you would agree with me that from a logical point of view, the fact that you (and others) did not try to attach a better GPS unit to the remote controller (which seems hard to do in the first place) is certainly not an explanation for the fact that the Follow Me has to not rely on GPS. You are not able with you decisions to determine how DJI engineers designed the Follow Me mode.
- I don't see why you seem to consider GPS not adequate. The Spark and the Mavic are already relying on their onboard GPS for navigation and stabilization in general. Assuming that the mobile unit attached to the remote has a GPS unit of comparable quality (which I would assume is true for most modern mobile devices), then asking the drone to follow the GPS position of the mobile device must necessarily have performance very close to asking it to over in place. In the latter case is comparing its real-time GPS reading with a previous one, in the former is comparing it's realtime GPS reading with the one provided by the mobile device. Basically the same thing if the precision of the two units is comparable.
The above are my motivations to be skeptical about the use of RSSI for Follow Me, but it may very well be that I'm ignoring something. I would be glad if you could suggest me a possible mechanism (based on what I understand is your previous experience bulding "UAV Ground station or antenna tracker to keep your radio antenna and video receiver antenna directly pointed at your drone") by which the drone could use the radio signal to follow the remote. That would really help me to understand better!
Thanks
Giacomo