3 ms·

These drones are used for light shows. I'm guessing they tried to run the show too long. It looks like they all start attempting to land before their batteries

by winrid 5y ago

These drones are used for light shows. I'm guessing they tried to run the show too long. It looks like they all start attempting to land before their batteries run out, around the same time...

anotheryou 5y ago
Alternatively a wrong ground height set. They do move at different speeds between their way points. The one at 0:25 is seriously out of juice or didn't expect the ground where it is. edit: looks like they really just run out of juice: https://twitter.com/pitdesi/status/1445119043199913990 https://twitter.com/pitdesi/status/1445119043199913990
chrismorgan 5y ago
> It looks like they all start attempting to land before their batteries run out, around the same time... That doesn’t immediately sound very plausible to me (as a layman who’s never dealt with drones but has dealt with battery-powered devices): I can’t imagine auto-power-down being based on flight time: surely it would be driven by battery voltage measurements, since batteries vary somewhat to begin with and vary a lot more as they age; and I then wouldn’t expect such homogeneity of battery lives that they would all drop out of the sky in such a short range of time. You could say that their not all behaving in the same way at the same time supports the idea of low-battery emergency landing, but I think that could be adequately explained by network congestion effects too, which could line up with the claim of an attack if such an attack were some sort of frequency jamming or network saturation thing (I don’t know what wireless systems such drones use). (The second video, “another view”, is much more interesting than the first, showing the constellation going from full strength to about two dozen left 30 seconds later, with another dozen subsequently failing very close to simultaneously in the last second of the video—I wish the video had gone for even two seconds more.)
- michaelt 5y ago
  > I then wouldn’t expect such homogeneity of battery lives that they would all drop out of the sky in such a short range of time. Well, they have all been doing a synchronised routine. Presumably they were purchased as a batch, and operated together. And if these were mini-drones that only had a 10 minute battery life to begin with? I could believe they'd all run flat within 60 seconds of one another.
  - chrismorgan 5y ago
    They’re not small drones—see the footage as they reach the ground. The flat battery hypothesis also feels unlikely because it’s too obvious a failure mode. On the other hand, I suppose it’d be possible to have them send the drones up, then get told to delay by a few minutes, and so exceed the planned flight time. But the drones aren’t that high up (much less than 30 seconds of controlled descent), so that doesn’t feel so likely.
- IshKebab 5y ago
  > surely it would be driven by battery voltage measurements No, it's driven by a battery management system that measures battery voltage, temperature and input/output current in order to get a more accurate estimate.
  - chrismorgan 5y ago
    Yeah, I didn’t mean to convey that battery voltage measurements would be the only factor, but I didn’t end up actually expressing that. Thanks for the clarification. Certainly the concept still applies: it’s driven by physical factors, not just time.
    - IshKebab 5y ago
      Oh sorry I totally misread your comment; thought you said the opposite!
bigiain 5y ago
I can just imagine some guy telling his boss "I know how to make the show better, if we replace all our 2,500mAHr batteries with these 10,000mAHr ones I've found, we'll get 4 times the flying time! We could fly for 40mins instead of only 10mins!" Then all the drones drop out of the sky after 2 minutes because the "10,000mAHr" cells only store about 500mAHr: http://danyk.cz/test18650_en.html http://danyk.cz/test18650_en.html
jsjohnst 5y ago
Yep, I agree, seems like battery failure. They do a controlled descent for most of the altitude, and then drop out of the sky and bounce off the ground/cars/trees (better seen in other videos). I’ve never done this large a coordinated display before, but have ten years UAV flying experience (quads, hexes, octo, and double stacked octo) and have done smaller scale test flights with a dozen choreographed drones.

by winrid 5y ago

These drones are used for light shows. I'm guessing they tried to run the show too long. It looks like they all start attempting to land before their batteries run out, around the same time...

anotheryou 5y ago

Alternatively a wrong ground height set. They do move at different speeds between their way points. The one at 0:25 is seriously out of juice or didn't expect the ground where it is. edit: looks like they really just run out of juice: https://twitter.com/pitdesi/status/1445119043199913990 https://twitter.com/pitdesi/status/1445119043199913990

chrismorgan 5y ago

> It looks like they all start attempting to land before their batteries run out, around the same time... That doesn’t immediately sound very plausible to me (as a layman who’s never dealt with drones but has dealt with battery-powered devices): I can’t imagine auto-power-down being based on flight time: surely it would be driven by battery voltage measurements, since batteries vary somewhat to begin with and vary a lot more as they age; and I then wouldn’t expect such homogeneity of battery lives that they would all drop out of the sky in such a short range of time. You could say that their not all behaving in the same way at the same time supports the idea of low-battery emergency landing, but I think that could be adequately explained by network congestion effects too, which could line up with the claim of an attack if such an attack were some sort of frequency jamming or network saturation thing (I don’t know what wireless systems such drones use). (The second video, “another view”, is much more interesting than the first, showing the constellation going from full strength to about two dozen left 30 seconds later, with another dozen subsequently failing very close to simultaneously in the last second of the video—I wish the video had gone for even two seconds more.)

michaelt 5y ago
> I then wouldn’t expect such homogeneity of battery lives that they would all drop out of the sky in such a short range of time. Well, they have all been doing a synchronised routine. Presumably they were purchased as a batch, and operated together. And if these were mini-drones that only had a 10 minute battery life to begin with? I could believe they'd all run flat within 60 seconds of one another.
- chrismorgan 5y ago
  They’re not small drones—see the footage as they reach the ground. The flat battery hypothesis also feels unlikely because it’s too obvious a failure mode. On the other hand, I suppose it’d be possible to have them send the drones up, then get told to delay by a few minutes, and so exceed the planned flight time. But the drones aren’t that high up (much less than 30 seconds of controlled descent), so that doesn’t feel so likely.
IshKebab 5y ago
> surely it would be driven by battery voltage measurements No, it's driven by a battery management system that measures battery voltage, temperature and input/output current in order to get a more accurate estimate.
- chrismorgan 5y ago
  Yeah, I didn’t mean to convey that battery voltage measurements would be the only factor, but I didn’t end up actually expressing that. Thanks for the clarification. Certainly the concept still applies: it’s driven by physical factors, not just time.
  - IshKebab 5y ago
    Oh sorry I totally misread your comment; thought you said the opposite!

bigiain 5y ago

I can just imagine some guy telling his boss "I know how to make the show better, if we replace all our 2,500mAHr batteries with these 10,000mAHr ones I've found, we'll get 4 times the flying time! We could fly for 40mins instead of only 10mins!" Then all the drones drop out of the sky after 2 minutes because the "10,000mAHr" cells only store about 500mAHr: http://danyk.cz/test18650_en.html http://danyk.cz/test18650_en.html

jsjohnst 5y ago

Yep, I agree, seems like battery failure. They do a controlled descent for most of the altitude, and then drop out of the sky and bounce off the ground/cars/trees (better seen in other videos). I’ve never done this large a coordinated display before, but have ten years UAV flying experience (quads, hexes, octo, and double stacked octo) and have done smaller scale test flights with a dozen choreographed drones.