Has your (or do you know of) RealSense or Obstacle Avoidance not worked the way you expected?

[Skeets]

Hi folks!
Just wondering how good RS and/or OA actually works. Has it ever not worked the way you expected it to? I am not sure if I am confident that it would work 100% of the time (given the proper factors of use). There have been several times that I have tested it where it seemed like it was going to hit an obstacle so I pulled away from the obstacle by changing course but I am not sure if RS/OA would have kicked in if I gave it more of a chance to.

Thanks!
Ed

 

[nechsproject]

To be honest I have been having trust issues as well. I have not tried to let it direct itself. I have been researching complete operation of the RS and feel like I am confident enough to give it a go this weekend. I will post a video if all goes well, if not I will post one of what happens .

 

[PatR]

To be honest, I've had an H since the middle of May last year and have had two of them since December. I have yet to bother messing with OA and don't have Real Sense modules. I have yet to figure out why so many people spend so much time TRYING to run into stuff just to watch automated obstacle avoidance work. There are a lot more fun things to do with an H than try to run into stuff. It also makes a heck of a lot more sense to make it a point to avoid flying into stuff than to see how close you can get to something before a safety device kicks in. It's like people taking their cars out and running into a brick wall at different speeds to find out how fast they can go and hit something before they start getting hurt. Crash test dummies indeed...

 

[Skeets]

To be honest, I've had an H since the middle of May last year and have had two of them since December. I have yet to bother messing with OA and don't have Real Sense modules. I have yet to figure out why so many people spend so much time TRYING to run into stuff just to watch automated obstacle avoidance work. There are a lot more fun things to do with an H than try to run into stuff. It also makes a heck of a lot more sense to make it a point to avoid flying into stuff than to see how close you can get to something before a safety device kicks in. It's like people taking their cars out and running into a brick wall at different speeds to find out how fast they can go and hit something before they start getting hurt. Crash test dummies indeed...

I have no intention on TRYING to flying into stuff and that isn't quite the point of my post. I would like to record my family and pets hiking, bike riding and what-not using the Wizard on Watch-Me mode and want to have enough confidence in the OA/RS that it won't fly into something. I've seen the videos of them flying towards the houses, buildings, tractors, etc., to see how good it works, but I would disagree that they WANT to fly into them, but rather testing how well the system works.

 

[PatR]

I've seen the videos too. Truthfully, the separation distance before obstacle avoidance kicks in with any of the brands is too small for my comfort. Personally, I would set up any follow or lead activity that involved people or pets to have an extremely adequate separation before employing the flight mode.. The H systems don't "crowd" the target subject unless they were initiated with minimum separation. Then again, I try to adhere to the law of unintended consequences and plan putting safe flying practices first, not "wow" factors.

If anyone is allowing an H to get so close to a pet or person that OA even MIGHT become necessary they should not be flying anything at all. None if us has the ability to prefict what a kid or pet may do, let alone a remotely piloted aircraft.

 

[Skeets]

I've seen the videos too. Truthfully, the separation distance before obstacle avoidance kicks in with any of the brands is too small for my comfort. Personally, I would set up any follow or lead activity that involved people or pets to have an extremely adequate separation before employing the flight mode.. The H systems don't "crowd" the target subject unless they were initiated with minimum separation. Then again, I try to adhere to the law of unintended consequences and plan putting safe flying practices first, not "wow" factors.

If anyone is allowing an H to get so close to a pet or person that OA even MIGHT become necessary they should not be flying anything at all. None if us has the ability to prefict what a kid or pet may do, let alone a remotely piloted aircraft.

You do know what ASSume means, right? And that is exactly what you are doing. I don't know where you got the assumption that I am setting up my H at a distance that would be unsafe. Quite the opposite in fact. I am not asking about the effectiveness of OA/RS so that it can play chicken with people and pets, but the effectiveness of OA/RS so that it can avoid trees and other obstacles while it is a safe distance away from people and pets. I suppose the way you use YOUR H is the only way?

 

[Mickeyboo]

I took mine off, in the winter it drained the battery more quickly and when shooting video it was always unsmoothly halting...it sits in a drawer at my house will probably never use Intel "nonsense" again...it's been implemented much better by another company in my experience with the two...

 

[nechsproject]

To be honest, I've had an H since the middle of May last year and have had two of them since December. I have yet to bother messing with OA and don't have Real Sense modules. I have yet to figure out why so many people spend so much time TRYING to run into stuff just to watch automated obstacle avoidance work. There are a lot more fun things to do with an H than try to run into stuff. It also makes a heck of a lot more sense to make it a point to avoid flying into stuff than to see how close you can get to something before a safety device kicks in. It's like people taking their cars out and running into a brick wall at different speeds to find out how fast they can go and hit something before they start getting hurt. Crash test dummies indeed...

Yes, but it can also be compared to taking your car to the track to see what it can truly do. And where would we be without crash test dummies?

 

[PatR]

You do know what ASSume means, right? And that is exactly what you are doing. I don't know where you got the assumption that I am setting up my H at a distance that would be unsafe. Quite the opposite in fact. I am not asking about the effectiveness of OA/RS so that it can play chicken with people and pets, but the effectiveness of OA/RS so that it can avoid trees and other obstacles while it is a safe distance away from people and pets. I suppose the way you use YOUR H is the only way?

I understand where you are coming from and I certainly don't think my way is the only way. I do firmly believe we want to put absolute trust in technology to avoid having to do some things manually though. Technology is only effective when it works exactly as designed, which is too often not the case. With zero certification standards we just don't have any assurances at all, and nothing to to go on to help us develop failure percentages. Even using forums such as this and others the only info we have comes from those that elect to participate and share info. To my way of thinking that's not very useful in statistical development.

 

[Dr Delta]

A few days ago this video appeared on this forum.
As for Real Sense etcetera look carefully at the second half, I saw some vids on RS before, but that it would work this great... I did not expect that.

Thank you Mr Edwards

 

[Skeets]

A few days ago this video appeared on this forum.
As for Real Sense etcetera look carefully at the second half, I saw some vids on RS before, but that it would work this great... I did not expect that.

Thank you Mr Edwards

Wow! The best example of RS/OA capabilities that I have seen! Thanks!

 

[Haydn]

Great video.
But I reckon he speed up the footage.
Under the tree over the fence around the car. That was my favourite butt clench moment

 

[Steve Edwards]

A few days ago this video appeared on this forum.
As for Real Sense etcetera look carefully at the second half, I saw some vids on RS before, but that it would work this great... I did not expect that.

Thank you Mr Edwards

Thanks for noticing. I did this video last week. An I did speed up the video. I thought it may be risky, but I had to know if it really worked. I appears to work just fine.

 

[Skeets]

I wonder if it did happen to fly into something and got damaged, would Yuneec fix/replace it?

 

[Dr Delta]

I wonder if it did happen to fly into something and got damaged, would Yuneec fix/replace it?

Rather silly question, nobody on this forum can answer that.
Nevertheless, in a case as nicey recorded as this one I suppose Yuneec will take their responsabilities.

 

[Dr Delta]

Thanks for noticing. I did this video last week. An I did speed up the video. I thought it may be risky, but I had to know if it really worked. I appears to work just fine.

Cool vid, Steve. Of course you speed up the vid, Comedy Capers like.... nice! But in the scenes that were really important the speed gets more to normal.

I did something like it myself, walking. Your vid is much better and emphasizes functionality of the RS in a much broader extent. Great stuff!

 

[Skeets]

Rather silly question, nobody on this forum can answer that.
Nevertheless, in a case as nicey recorded as this one I suppose Yuneec will take their responsabilities.

Well, I suppose it was my fault with ending the statement with a "?", but I am sure you did see "I wonder if"

 

[Dr Delta]

Of course.
That is why I wrote the second sentence.

 

[rumj48]

I pretend to film video of a not to stright running creek , a lot of trees at the edges, can the OB on the H will help to not crash ? Appreciate the help from experienced pilots

 

[Mrgs1]

I pretend to film video of a not to stright running creek , a lot of trees at the edges, can the OB on the H will help to not crash ? Appreciate the help from experienced pilots

Shouldn't be totally relied on, some branches or obstacles may not be detected, I would use as an aid only.