M3 more "fragile"?

[vindibona1]

Below I am posting the second video I've watched of a M3 making incidental contact with a tree branch and the gimbal not (fully) surviving the minor crash. It is obvious that the M3 leads with the dual camera and the top of the drone offers virtually zero protection for the camera. In a straight nose down crash of any kind the M3's gimbal is toast. OTOH, the M2, while still exposed has some protection from the "roof" of the drone. Also, the M2 has the possibility of installing a lens hood for added protection using the gimbal guard mount. FWIW I rarely if ever fly my Mini 2 without a gimbal guard and can know that on more than one occasion it has saved the gimbal from damage.

While I think there are some great things about the M3, it appears that you have to take an extra level of caution because almost any crash is gonna be an expensive one. Below find one video. If you want to see more check out 51 Drones crash vid.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[jonatious]

Mavic 3 needs an airbag like this

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[CanadaDrone]

It's really hard to believe durability is signicantly different one way or the other - if you crash any drone (no easy feat these days, if you are flying responsibly), you should expect damage, especially on the most fragile parts like the gimbal. The overall design is the same and most hands-on reviews I have seen so far have said if anything it feels more substantial than the M2.

There are way too many variables to make blanket statements about crash durability, especially this early on in the product life cycle. Unless you could repeatedly crash the two drones in an identical way, there is no way to objectively test the theory. Just as an example, I have cracked smartphone screens from only a couple feet off the ground and I have had smartphones survive falls from my ear height onto concrete - the exact way the "crash" happens can make all the difference and is not a variable you can easily control.

 

[globetrotterdrone]

Yeah, I mean, I guess you can deduct that the bigger camera and gimbal will have more mass and therefore kinetic energy falling down, but I doubt that there is more space for errors with the M2P. Yeah, it's smaller, but as @CanadaDrone pointed out, there are way too many variables to consider.

In a free fall on concrete, none of them have a chance.
You must be lucky, that it will hit soft ground with the bottom non-rectangular first either way.

 

[vindibona1]

It's really hard to believe durability is signicantly different one way or the other - if you crash any drone (no easy feat these days, if you are flying responsibly), you should expect damage, especially on the most fragile parts like the gimbal. The overall design is the same and most hands-on reviews I have seen so far have said if anything it feels more substantial than the M2.

There are way too many variables to make blanket statements about crash durability, especially this early on in the product life cycle. Unless you could repeatedly crash the two drones in an identical way, there is no way to objectively test the theory. Just as an example, I have cracked smartphone screens from only a couple feet off the ground and I have had smartphones survive falls from my ear height onto concrete - the exact way the "crash" happens can make all the difference and is not a variable you can easily control.

I stand by my statement. If I can find it I will post a clip of my M2 taking a pretty serious crash, and surviving with zero damage other than props. I seriously doubt that the M3 would not have had to make an expensive trip to the repair shop.

I did a cursory look for the clip and couldn't find it quickly. Finally found the original in the editor and am uploading it. While flying low over a golf course my M2 clipped an unseen tree branch and fell to the ground. Somehow I managed to get it back up on the air, but in ascending again, clipped something else and sent it careening into a chain link fence, which took out chunks of the props. Amazingly the gimbal and everything else on the M2 were unscathed. Never in a million years with its configuration would that crash not have damaged the M3's gimbal.

Edit: See the video below of my M2P's "double crash". It's crazy that after all that, even with damaged props I could fly it home. While the gimbal was catywampus after the crashes and sort of stuck, whatever pushed it aside wasn't hard enough to damage it and couple pushes on the 5D button recentered it for the flight home. The M3's camera protrudes so far from the body one has to wonder why DJI didn't provide more upwards viewing capability? Which drone can do that, a Parrot Anafi?

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[CanadaDrone]

I stand by my statement. If I can find it I will post a clip of my M2 taking a pretty serious crash, and surviving with zero damage other than props. I seriously doubt that the M3 would not have had to make an expensive trip to the repair shop.

I did a cursory look for the clip and couldn't find it quickly. But while flying low over a golf course my M2 clipped a tree branch and fell to the ground. Somehow I managed to get it back up on the air, but in ascending again, clipped another branch with sent it careening into a chain link fence, which took out chunks of the props. But the gimbal and everything else on the M2 were unscathed. Never in a million years with its configuration would that crash not have damaged the M3's gimbal.

Well, the fact is that you don't know, and nor do I or anyone else except maybe DJI. The few people who actually own the drone haven't even had it for 2 weeks and incredibly small sample sizes aren't useful for drawing conclusions anyway.

Unless you are able to crash a M2 and M3 in exactly the same way, and enough times to rule out any other errors, any comments regarding durability are at best pure speculation. It's great that your M2 survived a crash, but that doesn't tell us anything about M3 durability.

 

[Deleted member 103366]

It's a big lump of a camera that is very exposed, perhaps some protection is in order as in a lens hood or something similar to those side rails I have seen advertised for the FPV.

For the record I put my m2z into the top of a 50ft+ tree. From there it fell to the ground and hit a ground level boat hull and maybe concrete blocks on the way down.
It damaged both left hand side arms, stoved the nose cone in, popped the camera's left hand pitch pivot cover off ( it may act as a counter balance) and broke one of the gimbal's damper rubber mounts. The gimbal worked perfectly immediately afterward the crash and continues to do so though obviously I have replaced the lost cover and fixed the broken damper rubber mount.
It must have been some impact to stove the nose cone in as badly as it was.

 

[vindibona1]

It's a big lump of a camera that is very exposed, perhaps some protection is in order as in a lens hood or something similar to those side rails I have seen advertised for the FPV.

For the record I put my m2z into the top of a 50ft+ tree. From there it fell to the ground and hit a ground level boat hull and maybe concrete blocks on the way down.
It damaged both left hand side arms, stoved the nose cone in, popped the camera's left hand pitch pivot cover off ( it may act as a counter balance) and broke one of the gimbal's damper rubber mounts. The gimbal worked perfectly immediately afterward the crash and continues to do so though obviously I have replaced the lost cover and fixed the broken damper rubber mount.
It must have been some impact to stove the nose cone in as badly as it was.

I've crashed my Mini 2 twice, once a very hard bounce off a hard floor and once a gentle tumble onto concrete. In both instances the hull was damaged. On the hard it the plastic window on the camera actually popped out, the the camera and gimbal, both protected by a lens hood were undamaged.

 

[mschaffer66]

I might have said it somewhere else...

But the bigger concern should be that if I can't sense those tree branches how reliable is the advanced RTH going to be?

 

[vindibona1]

I might have said it somewhere else...

But the bigger concern should be that if I can't sense those tree branches how reliable is the advanced RTH going to be?

I think that's largely irrelevant. I think the "advanced" part of the M3 is advanced NAVIGATION around obstacles more so than detecting them. I'm not sure that in most cases RTH would be affected as typically the drone is programmed to go to a certain altitude for RTH before heading home. If you're inside a clump of trees and hit the RTH button it would seem likely that any drone would run into problems. Programming the RTH altitude is on the pilot, not on the drone. And if traveling forward in RTH mode I would think it would see a barren tree and avoid it long in advance while seeing small tree branches or electrical/phone lines would be an impossibility for all drone OAS's anyway.

 

[CanadaDrone]

I might have said it somewhere else...

But the bigger concern should be that if I can't sense those tree branches how reliable is the advanced RTH going to be?

Only a single sample, but this guy tried his best to crash the M3 specifically into thin branches and he couldn't. He also showed that his Air2S did crash in a similar scenario, demonstrating an improvement:

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[Deleted member 103366]

But isn't it part of the concern with the advanced RTH, that is does not climb to the set RTH height and thus relies on the AO to successfully navigate obstacles, if it doesn't see them how can it navigate around them?
Admittedly, with the opening video it seems as if the downward looking sensors did not pick up the branches and it is questionable as to whether or not they are used in an RTH, other than, perhaps, during the descent.

 

[GFields]

Wow..

 

[hiflyer201]

So I guess you can't use it to fly through trees. Bummer

 

[mschaffer66]

seeing small tree branches or electrical/phone lines would be an impossibility for all drone OAS's anyway.

And there in lies the problem of giving control to the drone of the path it takes in an "advanced" RTH scenario. It's path is only as safe as the ability of the sensors to find it's way.

If I launch my drone straight up 100 feet and I set my "dumb" RTH altitude to 100 feet I know its got nothing but open sky above me. However, if its "advanced" RTH decided that its going to start rounding corners to make the RTH more of an "as the crow flies" navigation and its sensors don't see the power line or thin branch between it and home that my dual optical sensors saw when I sent it up you could end up with some problems.

Now I will say that if the branch is part of a larger tree, hopefully it would see the larger tree and avoid it in general. But my bigger concern would be something like a power line.

I'd like to see someone take their brand new $2,200 toy out and put the "advanced" RTH through its paces and find out.

 

[CanadaDrone]

I believe you can also set the M3 to use the "old" RTH method - someone correct me if I am wrong though please.

Personally I have never used RTH in my life, so for my usage it makes no difference.

 

[mschaffer66]

I believe you can also set the M3 to use the "old" RTH method - someone correct me if I am wrong though please.

Personally I have never used RTH in my life, so for my usage it makes no difference.

I believe you are correct. Someone had said that in another post.

I don't know as though I've ever used RTH either. I don't trust it to not do something dumb...

 

[CanadaDrone]

I believe you are correct. Someone had said that in another post.

I don't know as though I've ever used RTH either. I don't trust it to not do something dumb...

Exactly. I never fly out of VLOS and I always return home with ~25-30% battery as per best LiPo practices. I can't imagine a scenario where I would ever need the drone to automatically RTH aside from a hardware failure, and I would prefer to be in control the whole time.

 

[hiflyer201]

I don't know as though I've ever used RTH either. I don't trust it to not do something dumb...

The problem with not ever using RTH is if it's needed it becomes very scary especially if you've lost the image on the controller. Don't ask how I know but I soon after went to an open field and used it in as many situations as I could simulate. I had only simulated RTH with the screen icon before and really didn't have a clue as to how to properly use the button on the controller.

 

[vindibona1]

And there in lies the problem of giving control to the drone of the path it takes in an "advanced" RTH scenario. It's path is only as safe as the ability of the sensors to find it's way.

If I launch my drone straight up 100 feet and I set my "dumb" RTH altitude to 100 feet I know its got nothing but open sky above me. However, if its "advanced" RTH decided that its going to start rounding corners to make the RTH more of an "as the crow flies" navigation and its sensors don't see the power line or thin branch between it and home that my dual optical sensors saw when I sent it up you could end up with some problems.

Now I will say that if the branch is part of a larger tree, hopefully it would see the larger tree and avoid it in general. But my bigger concern would be something like a power line.

I'd like to see someone take their brand new $2,200 toy out and put the "advanced" RTH through its paces and find out.

There is no doubt that the M3 has some real "improvements" over the M2 under some circumstances. The thing with the AOAS is that while an improvement isn't fool-proof. I think it might be great in a "follow me" situation much like the skydio. We know that there is potential for it to shoot higher resolution images than the M2's or Air2s. The question there is will current monitors be able to see and take advantage of the extra resolution?

But getting back to the "fragility" issue, I guess the AOAS come heavily into play. I suppose it could allow someone to fly closer to solid objects than one might think of doing otherwise. IMO that protruding camera is this drone's Achilles heel. How many repairs do you get with Care/Refresh??? I think the one thing that would have set this drone apart and make it worth $4k+ would be a removable gimbal where DJI could have sold a variety of camera configurations. I think they missed an opportunity to truly take this drone to the status that they would like it to be. Over time we'll get more reports of the M3 which will shake out the performance vs cost equation a bit more.