A 72h flight sounds impressive at first. But since this is done with external power, once the motors and electronics reach a temperature equilibrium (notice there is no payload) there is not much difference between a 30 minutes or a 72 hours flight.
Well, don't forget the effect of vibrations on the electronics. So they have proved that it won't vibrate apart in 72 hours, which is good. No guarantees of the same thing with some of the APM builds I've seen....
But besides not testing to failure, they also don't tell us how many attempts it took to reach this goal. ;) So this makes for interesting reading in web forums amongst DJI's target audience, but not much use for real engineering.
One of the things on my ‘must try’ list is an umbilical powered copter for long term loiter, such as sporting events. I think 240vac (UK) is the way to go with an light onboard switch mode power unit. Of course it would have a small backup onboard battery. A thin twin core cable should be light enough for perhaps 50mts height.
Mainstream aviation carries people. The FAA exists to protect people when they travel through the air, and protect people on the ground from things which are supposed to fly through the air but sometimes crash. Also, to protect the market for air travel by maintaining the public perception of safety.
There is precedent for the FAA allowing small objects to fly (and fall from the sky) with minimal regulation because this presents only a small risk to life and property. Think hobby rockets, weather balloons, etc. As the risk goes up, so does the level of regulation. The current rules for UAVs basically say: don't fly them over people's heads, and please keep out of airspace used by manned aircraft. I don't expect this to change much.
Now, if you're flying a heavier unmanned vehicle, or want to fly up high with the passenger planes, the FAA might come up with some rules for reliability, testing, certification, etc to allow this. But they need to feel that this is worth the effort. It likely is not worth the effort to put such rules in place.
John, I agree. If the copter had a failure after only 72 hours without payload or batteries it is more of a concern than something to be impressed with.
Can I smell the approach of a bureaucratic cash cow that regulates UAV components in the same way as mainstream aviation, or at least those operated for ‘hire and reward’ by qualified operators.
Heaven forbid the day will come when we need our copters ‘signed off’ after changing a motor.
The problem is that we all know that build and component quality can vary greatly from batch to batch. That's part of the reason why finding a realistic MTBF can be is a huge endeavor. Especially with mixed component systems like this.
It's a great test though of the top to bottom reliability. I've actually wanted to do this myself for quite a while although I was also thinking to do it outside.
It's great for RTF copter manufacturers to do this because it'll very likely highlight where things go wrong whether it's poor quality ESCs, mistakes in the build, etc. As the software gets more reliable (whether we're talking DJI, ArduCopter or whatever) the percentages of crashes due to hardware failures is rising.
A 72h flight sounds impressive at first. But since this is done with external power, once the motors and electronics reach a temperature equilibrium (notice there is no payload) there is not much difference between a 30 minutes or a 72 hours flight. In real flight time this is the same as 200-300 flights. Something you would expect as a bare minimum from a flagship >$4000 professional solution.
The one thing of worth this test shows, is that there are no problems with autopilot internals (variable overflow, counter wraps etc.) during long operations.
For this test to have real value in the industrial sense, it would have to be extended until failure and repeated over and over again to calculate a somewhat realistic MTBF.
Comments
I'd like to see 72hrs of constantly switching between flight modes. Maybe throw full yaw in there as well.
Well, don't forget the effect of vibrations on the electronics. So they have proved that it won't vibrate apart in 72 hours, which is good. No guarantees of the same thing with some of the APM builds I've seen....
But besides not testing to failure, they also don't tell us how many attempts it took to reach this goal. ;) So this makes for interesting reading in web forums amongst DJI's target audience, but not much use for real engineering.
I managed to lift my bicycle off the ground on the end of a long rope.The APM handled the swinging and extra weight just fine.
One of the things on my ‘must try’ list is an umbilical powered copter for long term loiter, such as sporting events. I think 240vac (UK) is the way to go with an light onboard switch mode power unit. Of course it would have a small backup onboard battery. A thin twin core cable should be light enough for perhaps 50mts height.
Wait...two pages in, and this is the first mention of a DJI flyaway? Standards are obviously slipping here are DIYD...:-D
Mainstream aviation carries people. The FAA exists to protect people when they travel through the air, and protect people on the ground from things which are supposed to fly through the air but sometimes crash. Also, to protect the market for air travel by maintaining the public perception of safety.
There is precedent for the FAA allowing small objects to fly (and fall from the sky) with minimal regulation because this presents only a small risk to life and property. Think hobby rockets, weather balloons, etc. As the risk goes up, so does the level of regulation. The current rules for UAVs basically say: don't fly them over people's heads, and please keep out of airspace used by manned aircraft. I don't expect this to change much.
Now, if you're flying a heavier unmanned vehicle, or want to fly up high with the passenger planes, the FAA might come up with some rules for reliability, testing, certification, etc to allow this. But they need to feel that this is worth the effort. It likely is not worth the effort to put such rules in place.
John, I agree. If the copter had a failure after only 72 hours without payload or batteries it is more of a concern than something to be impressed with.
Can I smell the approach of a bureaucratic cash cow that regulates UAV components in the same way as mainstream aviation, or at least those operated for ‘hire and reward’ by qualified operators.
Heaven forbid the day will come when we need our copters ‘signed off’ after changing a motor.
The problem is that we all know that build and component quality can vary greatly from batch to batch. That's part of the reason why finding a realistic MTBF can be is a huge endeavor. Especially with mixed component systems like this.
It's a great test though of the top to bottom reliability. I've actually wanted to do this myself for quite a while although I was also thinking to do it outside.
It's great for RTF copter manufacturers to do this because it'll very likely highlight where things go wrong whether it's poor quality ESCs, mistakes in the build, etc. As the software gets more reliable (whether we're talking DJI, ArduCopter or whatever) the percentages of crashes due to hardware failures is rising.
A 72h flight sounds impressive at first. But since this is done with external power, once the motors and electronics reach a temperature equilibrium (notice there is no payload) there is not much difference between a 30 minutes or a 72 hours flight. In real flight time this is the same as 200-300 flights. Something you would expect as a bare minimum from a flagship >$4000 professional solution.
The one thing of worth this test shows, is that there are no problems with autopilot internals (variable overflow, counter wraps etc.) during long operations.
For this test to have real value in the industrial sense, it would have to be extended until failure and repeated over and over again to calculate a somewhat realistic MTBF.
-
1
-
2
of 2 Next