Welcome to YuneecPilots.com!

Sign up and join the fastest growing Yuneec drone community.

  1. Texy

    Joined:
    Apr 25, 2017
    Messages:
    68
    Likes Received:
    13
    When flying in CCC, does the controller send the entire flight to the H at once or is it waypoint by waypoint or is it sending constant control data all the time?

    This is important to know before setting a waypoint at the edge of the reception area. When the H sets a heading to the next waypoint it may swing very, very wide sometimes. If I lose reception briefly, will the H continue to the next waypoint, and the next, and so on without failsafe measures taking control or increased risk of flyway?
     
  2. PatR

    Joined:
    May 1, 2016
    Messages:
    2,560
    Likes Received:
    1,009
    Location:
    Home at last
    Traditionally, flight plans have been stored in the auto pilots of autonomous aircraft. I don't know if this is also applicable to the H but I suspect it is.
     
  3. CraigCam

    Joined:
    Oct 1, 2016
    Messages:
    722
    Likes Received:
    251
    Ok.. good question. I have an Autel Amax that does waypoint but it's very clunky and does not make a CCC. You can draw a route with your finger on its screen map however and it will make its own points. That ship will continue to fly the route regardless of connection to radio or video once initialized and sent on its way. It will fly autonomous until its battery gets near LVC and it can calculate how far away it got vs battery time left to know when to abort and return. For sure this uploads to the ship during prep and the bigger the mission, the longer that upload time.

    As for the H...it shows a similar behavior upon beginning of mission. When you slide the start function, you should see the H lights blinking during the upload. It will then fly to it's staring point waiting for the forward stick commands for speed. For sure it's loaded into the H and my guess it's in temporary flash memory that clears automatically when you exit CCC mode. If it losses contact with the H for video that obviously would have no bearing on the route as it flys. I think once the data hand off is complete, the ST is passive and all you get is the telemetry and video data back to you. The fact that you can switch the right sticks to gimbal control makes me believe the CCC flight continues regardless of the radio link because you are not piloting at all.

    Perhaps Tuna will weigh in here as I'm sure he can clarify that is actually safe in flight during a CCC and it won't fly away if there are connection issues because it doesn't care about your signal during the mission. I'm going on observed assumptions for sure. I've yet to create a CCC that goes far away from me and of I did plan such a route, I'd still most likely build it where I can follow close behind and walk with it to keep better VLOS and connection. I'm hopefully optimistic that the H is as smart or smarter than the Autel and that this kind of safety is built in for the obvious reasons.
     
  4. Jimmylee

    Joined:
    Apr 24, 2016
    Messages:
    490
    Likes Received:
    202

    The Typhoon H cannot do true autonomous flights.

    Flight path is sent one by one to drone. If signal is lost drone will wait
    for next instruction or return to home.

    Drones from DJI like the Phantoms or Mavic can do true autonomous flight using third
    party app's.

    Flight path will be uploaded to the drone and will complete that route even if signal is lost from controller.
     
  5. Texy

    Joined:
    Apr 25, 2017
    Messages:
    68
    Likes Received:
    13
    So if I have one waypoint on one side of an obstruction and another waypoint on the other side, but the path leads the H behind the obstruction while on its way between waypoints (with no waypoints behind the obstacle, just flight path) then it should be ok?
     
  6. Jimmylee

    Joined:
    Apr 24, 2016
    Messages:
    490
    Likes Received:
    202
    The H will stop in its tracks when signal is lost and waits for signal to return, it needs constant communication with controller.

    If it does not reconnect in a determined amount of time RTH is triggered.
     
  7. Texy

    Joined:
    Apr 25, 2017
    Messages:
    68
    Likes Received:
    13
    Even if it is between waypoints on its way to the next stop?
     
  8. Jimmylee

    Joined:
    Apr 24, 2016
    Messages:
    490
    Likes Received:
    202
    Yes sir! The controller is telling the H to move forward, there is no memory buffer to complete a command.

    You loose signal and H stops all movement and becomes brain dead.
     
  9. Texy

    Joined:
    Apr 25, 2017
    Messages:
    68
    Likes Received:
    13
    Ok thanks for clearing that up
     
  10. Tuna

    Joined:
    Jun 9, 2016
    Messages:
    1,068
    Likes Received:
    933
    Location:
    UK
    I'm not sure that's true @Jimmylee I flew a mission over some woods this week and at one point the foliage blocked the signal (a lesson for me - plan your vantage point carefully to ensure line of sight) the Typhoon carried on flying the mission and taking photos. The ST-16 is definitely not controlling the mission, which can be proven by the fact that the new Waypoint missions for the H520 don't work properly on the H480 flight controller. However, you're right that if signal is lost long enough it will trigger a mission abort and return to home. That's how most flight controllers (except military ones) work - flying a mission when the operator can no longer safely abort it is illegal in most Western countries. DJI get around it by allowing third party software to do different things, but frankly it's deeply irresponsible.

    @Texy The simple answer is to give your H the best chance to maintain connection - situate yourself somewhere in clear space, keep the H in line of sight and ideally start at the 'center' of the mission rather than at one end where the furthest point is a long way away from you. In good conditions you can get a full mile reception, which is enough to cover a vast area. Brief loss of reception shouldn't stop the mission. Remember to set a good high RTH altitude to ensure the Typhoon can fly over any large obstructions!

    You can solve the 'wide turn' problem by putting in extra waypoints. The closer two waypoints are, the less the 'swing', so if you end a long run with two close together points 'cutting across' the corner, you will get a nice tight turn.
     
  11. Jimmylee

    Joined:
    Apr 24, 2016
    Messages:
    490
    Likes Received:
    202
    The ST-16 definatly is controlling the mission as Yuneec intended. All you have to do is turn off the controller during a mission
    and you will see the H stop in its tracks.
    A very short interruption of less than about 2 seconds you may not notice a stop because of latency.

    This is how DJI's own software works also, they are both staying on the safe side as to not break any rules. (Rules that don't exist yet)

    The fact that DJI released a SDK for there drones makes it much more fun!
     
  12. Tuna

    Joined:
    Jun 9, 2016
    Messages:
    1,068
    Likes Received:
    933
    Location:
    UK
    I believe it stops for safety (no connection to the controller) not because it is not receiving instructions. If you check out Pixhawk and other controllers, they're the same - the flight is run by the controller, but a loss of signal stops the replay for safety reasons. It's also why the curved cable cam is so hard to reproduce - there's no code running on the ST-16 controlling the way it swings through the waypoints, it's all on the controller. That's how RealSense is able to operate - it knows where it needs to travel and all the processing is done on the Typhoon. In comparison, DJIs flight controller is not quite as smart.

    The distinction is important because often you will loose video from the Typhoon long before the controller signal is dropped. So if you fly a mission, it may seem that you cannot see what the Typhoon is doing, but it will still continue (and not RTH) because it's receiving enough of a signal to be confident the controller is still there. Switching off your controller is the same as terminating the signal completely and will stop/RTH for safety reasons.
     
    Graham likes this.
  13. Jimmylee

    Joined:
    Apr 24, 2016
    Messages:
    490
    Likes Received:
    202

    Correct, that is what I was saying in previous messages.
    Loose the control signal (not the video signal) and the H
    will stop and hover and initiate RTH after so long without signal.

    DJI drones operate is exact same way unless you are using 3rd party app's
    through there SDK which Yuneec has not made available.

    I am sure this was done to err on safety.
     
  14. Gurusonwheels

    Joined:
    Jun 10, 2016
    Messages:
    139
    Likes Received:
    29
    tuna why the statement "which can be proven by the fact that the new Way-point missions for the H520 don't work properly on the H480 flight controller"
    they work perfectly fine for me .

    have you tried it ?

    if you follow the instructions i posted and on change the type and do not select the different model it works perfectly fine .
     
  15. Tuna

    Joined:
    Jun 9, 2016
    Messages:
    1,068
    Likes Received:
    933
    Location:
    UK
    As I understand it, it will certainly fly to the waypoints, but won't perform actions programmed for those waypoints - so some of the functionality is there, but not all of it. I've also seen some reports of odd behaviours when certain actions are included. That points to the waypoint missions being run by the flight controller, as the flight controllers on the H920+ and H520 are different from the H480. If it was being run on the ST-16, then you would be able to do all the same things regardless of which model you had connected (and perhaps wouldn't need to select a model, as the flight controller on each machine would just have to do what it was told).

    Unfortunately I can't try it on my controller as the maps aren't working (and I need to keep it that way for debugging things on my app).

    Some of this will become more clear when the SDK is released, which is due fairly soon. I'm prepared to eat my words, but the evidence points to the work being done on the flight controller, which is consistent with the way the industry seems to be going.
     
  16. Texy

    Joined:
    Apr 25, 2017
    Messages:
    68
    Likes Received:
    13
    It seems to me that the logical flow would be like this:

    1. Controller sends the location of the next waypoint and gimbal position (when required) to the H
    2. H is now programmed to fly to the next waypoint, it begins turning and going there while moving the gimbal to its next position at the same time. At this point the H is flying itself until it reaches the waypoint.
    3. The H reaches the waypoint and waits for instructions. If there are none then it stops.
    4. Goto 1
    Does this seem right?
     
  17. Tuna

    Joined:
    Jun 9, 2016
    Messages:
    1,068
    Likes Received:
    933
    Location:
    UK
    Nope, because the Typhoon actually needs three waypoints (the 'current' one that it's just reached, plus the next two) to be able to calculate the curved cable cam route. For those who've worked with CAD software, it calculates a spline through the three waypoints. If it just sent a waypoint at a time, we'd see kinks and jumps as it went from one section to the next.

    So it's definitely not sending a single waypoint at a time, and it doesn't make any sense to send just three when you could send the whole mission. Remember also that the Typhoon is flying at 11m/s for a mission. If the signal is delayed when it's going from one waypoint to the next, the Typhoon would be flying 'blind' until the next instruction was successfully sent.

    You can check out the PX4 architecture (Architectural Overview ยท PX4 Developer Guide) for an idea of how it all hangs together for an equivalent flight controller, but in essence there is a 'Flight mode' (e.g. Flying manually, flying a mission, return to home) that decides what it is doing. That flight mode runs a given loop/program unless or until it's told to do something else. That 'instruction to do something else' might be a command from the ST-16, or a signal from the flight controller itself. One example of a signal from the flight controller is loss of R/C connection for more than a certain period. When that happens, the current flight mode is stopped, and the return to home flight mode is triggered.

    So the flow is more like:

    1. The mission is uploaded to the H. It enters CCC mode, flies to the first waypoint and waits
    2. The Typhoon continues with the mission (initially just waiting)
    3. If the ST-16 sends a speed command (ie start mission/reverse) the mission is started, slowed down etc.
    4. If signal is lost for more than a certain period, RTH is triggered.
    5. If the mission is complete, the Typhoon waits at last waypoint
    6. Goto 2
     
  18. Texy

    Joined:
    Apr 25, 2017
    Messages:
    68
    Likes Received:
    13
    I see what you mean about planning all the curves in advance now. I have a complex flight to plan and it involves the H going around behind a tower that is some distance away. I see now that in order to accurately guess what will happen is go out and fly my H behind the tower to see if comes out on the other side or if it stops! I'm actually glad that it comes to this because it means more time flying, and that is my favorite thing to do.
     
    Tuna likes this.
  19. Tuna

    Joined:
    Jun 9, 2016
    Messages:
    1,068
    Likes Received:
    933
    Location:
    UK
    Just remember to set RTH height to higher than the tower - so if it does loose signal and heads back to you it will fly over it and not into it! :)
     
    Texy likes this.