Embarrassing confession from a noob

That's not the worst that could have happened. You'll heal (for free). You saved your P3, that's the good news here.
In the future, maybe you manually take off and land. You'll naturally be in control and instinct might not be to freak out about what's happening if you're the one making it happen.
And perhaps you could consider rolling back the firmware. I know now everyone has trouble with the latest firmware but clearly you are. It might help with the confidence anyhow.
 
  • Like
Reactions: speedytwenty
Maybe you have a reading comprehension problem, so I'll do you a favor and ask one more time: what does his quad vertically ascending without throttle input have to do with his skillset?

Skillset comes into play with the rescue but not this cause.

Sent from my Nexus 6 using Tapatalk

Someone with a proper skill set would have known to not do auto takeoff with an 8 foot ceiling (or indoors at all). A skill is not only how well you know how to do something but also how much you know not to do.

OP knows where he went wrong.
 
  • Like
Reactions: N017RW
Someone with a proper skill set would have known to not do auto takeoff with an 8 foot ceiling (or indoors at all). A skill is not only how well you know how to do something but also how much you know not to do.

OP knows where he went wrong.
App says the aircraft will ascend to 3.9 feet and then hover using auto-takeoff.

He had double the space necessary. What's wrong with that? There are lots of scenarios outdoors where there is a vertical "ceiling" you need to stay under before running into trouble.

What are you suggesting? Never take off with <250% of the vertical airspace you need? 500%? 1000%?

Sent from my Nexus 6 using Tapatalk
 
  • Like
Reactions: speedytwenty
It's foolish to put so much trust in an autopilot. ESPECIALLY with all the reports of similar issues and many disabling the VPS all together.
Then you ask "What's wrong with that?"
This story explains it.
 
It's foolish to put so much trust in an autopilot. ESPECIALLY with all the reports of similar issues and many disabling the VPS all together.
Then you ask "What's wrong with that?"
This story explains it.
Given the issues with VPS on the latest FW I definitely agree.

I just don't think this is pilot error - it's a system error that could have happened to anyone. Even a seasoned pilot who didn't read forums and know there are FW issues w/ the VPS system.

Sent from my Nexus 6 using Tapatalk
 
  • Like
Reactions: speedytwenty
App says the aircraft will ascend to 3.9 feet and then hover using auto-takeoff.

He had double the space necessary. What's wrong with that? There are lots of scenarios outdoors where there is a vertical "ceiling" you need to stay under before running into trouble.

What are you suggesting? Never take off with <250% of the vertical airspace you need? 500%? 1000%?

Sent from my Nexus 6 using Tapatalk
If the margin of error is 4 ft, I want to be in control the whole time, indoors or out.
 
From DJI specs:
Hover Accuracy: Vertical: +/- 0.1 m (when Vision Positioning is active)

According to how this product is designed and intended and advertised to use, OP should have been fine. Period.

I'm pretty sure VPS was not advertised to work that great over carpet (as you said yourself).
 
Agreed. Because carpet absorbs ultrasonic waves. If the systen were functioning as designed, carpet should have the opposite effect: the VPS system should think it is much further away from the ground than it is resulting in the aircraft crashing into a floor, not a cieling.

Sent from my Nexus 6 using Tapatalk
 
  • Like
Reactions: speedytwenty
Please post a copy of the f/w algorithm from which you state: "If the system were functioning as designed, carpet should have the opposite effect" otherwise your just making claims you have no knowledge of and may be skewing your overall judgement.
 
Please post a copy of the f/w algorithm from which you state: "If the system were functioning as designed, carpet should have the opposite effect" otherwise your just making claims you have no knowledge of and may be skewing your overall judgement.

I'm assuming here, but i would be willing to bet that VPS doesnt have logic that tells it to crash the p3 into the ground if it gets garbled ultrasonic data back!
 
Please post a copy of the f/w algorithm from which you state: "If the system were functioning as designed, carpet should have the opposite effect" otherwise your just making claims you have no knowledge of and may be skewing your overall judgement.
I don't code for DJI so I'm not stating as fact.

I do have experience working as an engineer with ultrasonic distance sensors.

Take it for what it's worth.

Sent from my Nexus 6 using Tapatalk
 
I'm assuming here, but i would be willing to bet that VPS doesnt have logic that tells it to crash the p3 into the ground if it gets garbled ultrasonic data back!
You would hope there is some inbuilt safety measures but how would it know it the signal is garbled? The barometer information? Indoors that might be garbled too.

I'd be curious to know.

Sent from my Nexus 6 using Tapatalk
 
You would hope there is some inbuilt safety measures but how would it know it the signal is garbled? The barometer information? Indoors that might be garbled too.

I'd be curious to know.

Sent from my Nexus 6 using Tapatalk

Well I was waiting for you to tell us based on your knowledge of design and intended function.
 
I know very little about how the VPS on the P3 works and I don't think it would cause it crash into the ground during takeoff. But it seems to me that if the VPS is used to determine the altitude during auto-takeoff, that when the signal is absorbed, it either a) over calculates the altitude which wouldn't cause the quad to continue ascending or b) doesn't complete the calculation and continues to ascend while waiting.
 
  • Like
Reactions: AlmostTan
b) would be an interesting thought. I don't know the acoustic reflectivity of carpet, but I my guess would be that it still does reflect somewhat. I would think that for b (incomplete calculation) to occur the signal would have to be completely attenuated but again I don't know anything about carpet. If it's somewhere in the range of soft tissue or fat then it's entirely possible...still leaves room to speculate on how their programming is setup. Whatever it is, is wrong/not good.
 
  • Like
Reactions: speedytwenty

Members online

Forum statistics

Threads
143,066
Messages
1,467,354
Members
104,934
Latest member
jody.paugh@fullerandsons.