We had similar issues with drones not flying correct distances, or just not flying at all. You are correct, if the drone is told to fly too far it does not do well. We had similar problems with it not flying far enough. To get it to fly further I just had my kids code in multiples of the same distance to get our desired final desired travel length. It was a good way to use Functions or whatever it was. Honestly, we started having so many issues with DroneBlocks and inconsistencies (even when the drones were properly coded) that I spent more time trying to troubleshoot than I did teaching the kids. We axed the coding part of the Drones course. In reality, coding isn’t really used for a whole lot with how drones are sold today. Waypoints will do everything most people are ever going need.
@nrichhart thanks for your feedback and we’re truly sorry to hear of all the trouble you’ve been having. We’ve run into issues over the years with:
- Environmental factors
- Drone firmware inconsistencies
- Firmware limitations (can only fly 500cm max, cannot flip with battery less than 50%)
- Wifi interference when lots of drones are flying
- Hardware failures
- DroneBlocks features blocked by district IT
The list goes on and on. But one thing I want to focus on is your comment about using waypoint navigation. If the end goal is to navigate a drone from point A to point B I would agree with your statement with waypoints being the way to go. We learned a tough lesson when we decided to bring drones indoors. Indoor positioning/navigation is much more challenging than outdoor navigation (FAA regulations aside). GPS-denied navigation has introduced so many challenges related to being able to maintain position (Tello does a pretty amazing job) and then navigate based on optical flow. Somehow trying to judge distance by using a relatively inexpensive sensor on the drone.
We have been prototyping with some new hardware that should improve performance, but it’s not perfect. If we are able to leverage external sensors such as UWB beacons (like with our drone light show kits) you can get much better accuracy/reliability. Just know that we are constantly working to support new hardware and provide new content to further engage students.
Let me also make one last statement as I started developing several apps for DJI drones before beginning work on DroneBlocks. In a lot of cases it’s 3rd party developers that enable a new capability for a platform. If you look at many of DJI’s apps (Litchi, DroneLink, DroneDeploy, etc) they have waypoint navigation that was created by a non-DJI employed developer. Businesses have been built on the hardware platform just like we’ve witnessed with mobile phones. My hope is that we foster an interest in coding so that students then go off and create cool apps/services for drones or whatever cool hardware comes next.
agree. we spend more time troubleshooting (open/closing the software, reconnecting, restarting drones, etc.) that we are just about ready to drop it. we fly 8 drones at a time. 2 will usually not listen to instructions. it’s become a hindrance to the program.
thanks for sharing!
it seems like the Tello is fantastic at internal navigation without GPS when not being flown by drone blocks.
the inconsistencies come when the code and the drone are not in sync. for example the code will skip blocks, the drone will just start floating off into space, etc.
proof: when we think the accelerometers are broken because the drones are drifting, we test the Tello by flying it manually using the tello app, and all of a sudden the drone is amazing.
we believe, after much testing, at this point that Wi-Fi interference is our last unresolved issue. has droneblocks seen any hardware solutions? external Wi-Fi modules for each computer?