Going on 3 years without a crash. What's my secret?

[Donnie Frank]

Hey doods;

First off, I'm going to qualify my initial assertion by stating that I DID have 2 crashes last year due to prop failures. Both crashes occurred during automated flights. So these crashes were not caused by unpredictable software behavior, run-aways or pilot error. I made a video about those crashes:

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

Outside those failures, I haven't had a crash in years. I attribute my flight success to a couple things:

1) Practice, practice, practice. Get a small drone with prop guards and fly around the house. If you can fly a small, cheap drone around the house, then your skill level is probably better than 90% of the drone pilots out there. This skill directly translates to more expensive drones, as the controls and behaviors are exactly the same.

2) Don't count on technology to save your ***. While Object Avoidance looks good on paper, in all reality, it's is NOT reliable. A very large percentage of crashes I read about are either due to flat-out failure of the OA, or unreasonable expectations for the OA system. I have TURNED OFF OA on all my vehicles, and don't use it (and therefore, don't COUNT on it) at all.

3) Probably the most important protocol to follow: Find a firmware and software version that work, and STICK WITH THEM. Another large portion of crashes seem to be due to random, erroneous drone behavior. Every time DJI releases a new firmware/software version to be more and more compliant with ever-growing FAA regulation, make no mistake; YOU ARE THE BETA TESTER. I run legacy (read: YEARS OLD) firmware and software on ALL THREE of my birds (Inspire 1, v2, Mavic Pro and P4P), and I could not be happier. My drones are rock-solid reliable, but more importantly, BEHAVE PREDICTABLY.

Here's a list of the FW/Software I use:

Inspire 1:

Go App: v3.1.1
App database: v00.00.01.04
Aircraft: v1.8.1.00
Remote: v1.6
X3 Camera: v1.8.1
X5 Camera: v1.11.1.50


Mavic Pro:

Go App: v3.1.1
Go4 App: 4.0.8
Aircraft: 01.03.0700
Remote Controller: 01.03.0700
Flight Database: 00.00.01.04
Basic App Fly Safe Database: 01.00.01.05
Precise App Fly Safe Database: 01.00.01.05


Phantom 4 Pro:

App: v4.0.8
Aircraft: v1.3.509
Remote: v01.04.01.00
Precise Fly Safe database: v00.00.01.04
Basic Fly Safe database: v00.00.01.04


4) When flying automated flights, an elevated level of diligence is required. Most of my jobs utilize Intelligent Flight Modes or Waypoint automation. Because I don't rely on Object Avoidance, I HAVE to understand my environment, the path the drone is going to take, and more importantly, be on the ready to stop the mission should something go awry. The "ATTI" or "Sport Mode" buttons are your panic buttons. If you're flying an automated flight that looks like it might land you in a tree or smash you into the side of the building, toggle these switches. You can bring the drone back to GPS mode.

5) Test, test, test. Never once have I ever gone into the field with an untested platform. EVERY aspect of that job has been tested with a near-duplicate of that mission under controlled circumstances like a park or open field.

My partner and I just got a job doing extended LIDAR missions with his M600. You can absolutely bet that we are going to test the bejesus out of this BEFORE loading up the vehicle and traveling 5 hours to the job site. I have planned dummy payload flights to simulate the LIDAR unit's payload and current draw. Once those tests have flown successfully, we'll move on to tests with the actual LIDAR unit, which includes data acquisition. You can bet that we will reproduce every nuance of our planned mission, including battery changes mid mission and exit strategies, but in a controlled environment. Which brings me to #6.

6) ALWAYS have an exit strategy. Practice abandoning and recovering from rogue missions. Put yourself in pretend, emergency scenarios under controlled conditions. For example, what if you're flying a remote mapping mission and a low-flying helicopter comes into your mission space? You have to:

A) Pause the mission.
B) Bring the drone down to a safe altitude.
C) Continue the mission.

Without practicing this, you might panic and/or not be able to continue your mission, which means starting over. I HATE starting over.

Anyway, I hope all this helps. I posted this because I see crashes and runaways almost daily in the 3 forums I belong to. I don't fly every day, but I DO fly every week. So my experience is valid.

Good luck!

D

 

[dannybgoode]

I have to disagree with what you say about. firmware and beta testing etc. Old firmware doesn’t somehow become more stable and it’s not updated.

If there is a problem with it, it gets fixed by a new firmware release. It doesn’t magically fix itself over time.

Further any new firmware will have been through a full beta test regime. Sure it might have bugs but then so might the old one but the old one will never get fixed.

I’ve mentioned this elsewhere but as well as the headline features most firmware releases for any device contain stability improvements and minor bug fixes which over time can prove to be quite significant to the performance of the device.

I’ve been flying for 2 years not and have never crashed either. I always update all software and firmware at the earliest opportunity.

 

[ZX14-R]

The amount of crashes you have proportionally relates to the following:

Amount of time flown
Amount of risk taken
Amount of skill/training you have
The equipment you use Controller/Device
.....

 

[Donnie Frank]

I have to disagree with what you say about. firmware and beta testing etc. Old firmware doesn’t somehow become more stable and it’s not updated.

Correct. The assertion that it "becomes more stable" is erroneous. However, the assertion that it REMAINS STABLE is absolutely true. For ANY devices in life, find a STABLE firmware and STICK with it.

Another erroneous statement might be that firmware BECOMES unstable. For those reasons, sticking to STABLE firmware is absolutely much, much better than taking chances on NEW firmware. There are literally dozens of crash reports in this forum alone that can be attributed to a firmware update.

If there is a problem with it, it gets fixed by a new firmware release.

Oh from the mouths of babes.... With all due respect, that is a VERY naive statement.

Let me ask you this. What if there are NO problems with the firmware? Then what would be the wisest approach?

It doesn’t magically fix itself over time.

Correct. But more importantly, it doesn't BREAK itself over time.

Further any new firmware will have been through a full beta test regime.

Again...from the mouths of babes... I don't even know where to begin. Here are some bullet points:

* Released firmware is often alpha tested by hired testers, but rarely beta tested in the true sense of the word, and often not THOROUGHLY beta tested.
* Even the term "beta testing" covers a broad band of testing regimens.
* Software / Firmware are often released as "beta versions." However, not all companies are as forthcoming regarding the testing of their software. DJI notoriously releases "not fully tested" versions of their software / firmware. Again, I refer you to the contents of this forum as proof of that.
* Regardless of a company's intentions, NO software / firmware beta tests will ever equal real-world, over time, in-the-field testing. This is where the user comes in. Companies release "tested" software / firmware, but ultimately it takes field testing to flush out bugs. There are literally hundreds of examples throughout history to corroborate this.

Sure it might have bugs but then so might the old one but the old one will never get fixed.

I specifically stated to use "stable firmware." I came up with the above versions because they were not only fully tested for months or years by the drone community, but also proved themselves to me in the field.

I’ve mentioned this elsewhere but as well as the headline features most firmware releases for any device contain stability improvements and minor bug fixes which over time can prove to be quite significant to the performance of the device.

Perhaps in some cases. But ask anybody who works with firmware / software on a regular basis. It's not uncommon to fix one bug and induce another. You'll find that once a firmware is stable, a company will release a new version with more features (or in the case of DJI, more restrictions). These "updates" often add NEW bugs to the firmware. My firmware choices are a balance of feature needs vs. stability. Yes, I give up some features. But my rich reward is not only stability, but then I don't have to endure all the FAA-test-safety-EULA-Nazi "features" that get deployed and are FORCED onto the user.

I’ve been flying for 2 years not and have never crashed either. I always update all software and firmware at the earliest opportunity.

With all due respect, if you've never been caught in the field miles from Internet and Cell service with a locked up drone because you had to "sign in" and agree to a new EULA or take a test, then you're probably not flying on the same professional level I am. Drone income accounts for 1/2 my revenue stream. So I not only fly regularly, I fly professionally, which means I don't have the luxury of staying "connected" or being locked up in the field. My work takes me miles from civilization quite often. Does yours?

I have been bitten by the "must sign in or we're locking up your drone" bug ONCE. Never gain.

D

 

[Donnie Frank]

The amount of crashes you have proportionally relates to the following:

Amount of time flown
Amount of risk taken
Amount of skill/training you have
The equipment you use Controller/Device
.....

Agreed. But hopefully you can agree that having a firm understanding of what goes on UNDER THE HOOD of your drone is ALSO part of that recipe for success. And that includes a firm understanding the drone's software and firmware. And I'm not talking from a user point of view as in "knowing where all the buttons are and what they do." I'm talking from a power-user point of view of fully understanding how to manipulate firmware and software to make your drone powerful, durable and reliable.

D

 

[DanMan32]

No firmware or app is bug free. Even from day one, we find issues. However I would agree sticking with a version of firmware/app that you are most comfortable with where you know what the quirks are and how to deal with them.
As for app, that means you also have to stick with a mobile device and OS and never change those.

 

[WithTheBirds]

I have to disagree with what you say about. firmware and beta testing etc. Old firmware doesn’t somehow become more stable and it’s not updated.

If there is a problem with it, it gets fixed by a new firmware release. It doesn’t magically fix itself over time.

Further any new firmware will have been through a full beta test regime. Sure it might have bugs but then so might the old one but the old one will never get fixed.

I’ve mentioned this elsewhere but as well as the headline features most firmware releases for any device contain stability improvements and minor bug fixes which over time can prove to be quite significant to the performance of the device.

I’ve been flying for 2 years not and have never crashed either. I always update all software and firmware at the earliest opportunity.

Whether you might agree is largely irrelevant. The fact you have installed all updates and had no issues tells us nothing about the reliability and stability of an earlier firmware version. You make many assertions here I suspect you have no evidence for and you competently ignored an important observation being the potential issues encountered in being a beta tester for GEO fencing implementations. .

 

[sar104]

I've yet to see a crash caused by a firmware bug. Geo-fencing has certainly caused a few issues, but I can only recall one incident where a drone was lost due to autoland in a TFR - and that is still clearly pilot error.

 

[Donnie Frank]

No firmware or app is bug free. Even from day one, we find issues.

If you can find bugs in the aforementioned software / firmware, lemme know. I haven't been able to find any. And I've been using them for a couple years. But perhaps you know something I don't.

However I would agree sticking with a version of firmware/app that you are most comfortable with where you know what the quirks are and how to deal with them.

I don't use the firmware for subjective reasons like "comfort." I use it for OBjective reasons like "no bugs." But again, I defer to your expertise. If you were able to find some bugs in the aforementioned versions, please share your findings.

As for app, that means you also have to stick with a mobile device and OS and never change those.

Well, to a degree...

I HAVE allowed my iPads to update to iOS 12.2 (if memory serves). But I always have an exit strategy to roll them back if I have to (even though, technically, Apple only allows roll backs for 30 days). And my NEW Gen 3 iPad running iOS 12.5 works well with the legacy apps, but does NOT play well with the Inspire 1. According to the forums, this is a Gen 3 chipset issue, not a firmware or software issues.

Besides, talk about ROI! What's wrong with using "old iPads." You say that like that's a bad thing. I use my iPads for my sound company and drone company exclusively. All other apps are stripped. For me, iPads are tools, not toys. I'll do email on my PC. The end result is they're rock solid and will work until the batteries explode. HA!

D

 

[dannybgoode]

@Donnie Frank - I agree with parts of what you say about firmware but your op is about how you’ve managed to avoid a crash in 3 years, not how you’ve avoided any problems in the field with EULA’s etc.


Updating or not the firmware has little or nothing to do with crashing. As for avoiding issues after updates then I find the best course of action is to update and give it a quick test to see if there’s anything that needs agreeing to or unlocking.

Sure if I relied on my drone for income I might not update the night before a gig but I’d still update at the earliest opportunity.

Of course people are free to choose whether to update or not but it is naive advice to tell people never to do it as there are many valid reasons for staying current.

 

[brettbrandon]

@Donnie Frank OT Just wanted to say hey, havent seen ya since the SAC days. If you are using iPads for sound Im guessing Behringer (cant be sac)?

Fly safe...

 

[Donnie Frank]

@Donnie Frank - I agree with parts of what you say about firmware but your op is about how you’ve managed to avoid a crash in 3 years, not how you’ve avoided any problems in the field with EULA’s etc.

True. The EULA/sign in avoidance is a peripheral benefit.

Updating or not the firmware has little or nothing to do with crashing.

I disagree. Between Mavic Pilots forum, Phantom Pilots forum and Inspire Pilots forum, there are literally dozens of crashes directly linked to firmware updates.

As for avoiding issues after updates then I find the best course of action is to update and give it a quick test to see if there’s anything that needs agreeing to or unlocking.

That doesn't work. I fully tested my Inspire 1 rig the night before a shoot (as I always do). The next day, on the job site, I got the dreaded "Sign in or don't fly" screen. Somehow, between the time I had shut down my drone and flew the next day, DJI deployed something while I wasn't looking. When I looked, my trusty Go App v3.1.1 was now v3.1.1.54. It turns out that Apple was the culprit. Apparently an iOS update reset the "Don't allow apps to update" setting to "Oh go ahead and screw me any time you wish" setting. The answer and the cure is a hyper level of diligence that goes far beyond "testing your gear the night before." Lesson learned. Everything is locked down. That's why I use the broad terms "firmware / software." This covers everything that touches our drones, which includes iOS and more importantly, iOS settings.

Sure if I relied on my drone for income I might not update the night before a gig but I’d still update at the earliest opportunity.

I respect that you do what makes sense to you. But understand that all three of my drones are rock solid. How do you get any better than that?

My flow chart is simple. Buy the drone and fly it. Does it fly without problems? Am I happy with the features? If yes to both, lock it down. If not, go to the next version and re-ask those questions until both answers are yes. Then lock it down.

Of course people are free to choose whether to update or not but it is naive advice to tell people never to do it as there are many valid reasons for staying current.

Name one reason that trumps "rock solid reliability."

The funny part is sometimes DJI or <insert drone software company here> doesn't even offer new features. Sometimes the update is the result of some new FAA ruling or compatibility fix ("Now works with iOS x.x.x").

And don't count on DJI's release notes to be forthcoming. Their release notes are s***. Couple all this to the fact that many drone crashes are attributed to firmware updates, and you get severely diminished returns on your update investment. And honestly, to assert that firmware or software updates have zero risk is naive.

The other thing I touched on was having an exit strategy. You'll find that most companies are not willing to give away legacy .ipa files. In fact, some will even insist that legacy app acquisition is impossible, which is preposterous. When I DO allow software to update (and remember, I use a couple third-party apps for waypoint flying), I always have a way of rolling back. For whatever reason, what is easy in the Windows/Linux environments is very difficult (and impossible for most) in the Apple environment.

And to be clear, I'm no Apple fan boy by any stretch. I use their products out of pure necessity. When I started my business, I needed the most reliable platform. And there were simply too many issues with the Android offerings. But I digress....

D

 

[Donnie Frank]

@Donnie Frank OT Just wanted to say hey, havent seen ya since the SAC days. If you are using iPads for sound Im guessing Behringer (cant be sac)?

You are correct. I still assert that SAC is a far more powerful mixing environment than any of the more popular "budget conscious" digital offerings. It's just too clumsy for practical use. Besides, any more these days my clients provide their own sound system and FOH. I just bring my iPads and headphones for the same money I was charging for a full sound system. Gaining a positive reputation has its perks. SAC would be great on tour in controlled environments, but my situations are too dynamic.

Funny...I was actually thinking of revisiting the SAC forum to see what surface compatibility SAC has. The version I have is years old. I'd almost be afraid to update it.

The peripheral benefit of the iPads is that I can use them for my drone company AND my sound company. And I have to say that they work very well for both.

Fly safe...

Will do. You, too. Good hearing from ya.

D

 

[Harro]

From what you are saying, Donnie Frank, I would assume that you are running the original firmware that came on your drones when you bought them, otherwise you must have updated your firmware at some point in time. If so, how did you decide on the version to use?

 

[Donnie Frank]

From what you are saying, Donnie Frank, I would assume that you are running the original firmware that came on your drones when you bought them...

On the Inspire 1, yes. Not so for the Mavic Pro or P4P.

...otherwise you must have updated your firmware at some point in time.

Correct.

If so, how did you decide on the version to use?

I wanted the versions that gave me the features I needed pre-Naziware. There's a clear line where DJI starts "legislating" their drones. So only FW below that line is considered. From there, I start with the latest qualifying version and work my way backwards. Most of my birds have had 3-5 different FW versions installed. The versions I chose have worked flawlessly for well over a year with no Naziware.

D

 

[Arce101]

Hey doods;

First off, I'm going to qualify my initial assertion by stating that I DID have 2 crashes last year due to prop failures. Both crashes occurred during automated flights. So these crashes were not caused by unpredictable software behavior, run-aways or pilot error. I made a video about those crashes:

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

Outside those failures, I haven't had a crash in years. I attribute my flight success to a couple things:

1) Practice, practice, practice. Get a small drone with prop guards and fly around the house. If you can fly a small, cheap drone around the house, then your skill level is probably better than 90% of the drone pilots out there. This skill directly translates to more expensive drones, as the controls and behaviors are exactly the same.

2) Don't count on technology to save your a**. While Object Avoidance looks good on paper, in all reality, it's is NOT reliable. A very large percentage of crashes I read about are either due to flat-out failure of the OA, or unreasonable expectations for the OA system. I have TURNED OFF OA on all my vehicles, and don't use it (and therefore, don't COUNT on it) at all.

3) Probably the most important protocol to follow: Find a firmware and software version that work, and STICK WITH THEM. Another large portion of crashes seem to be due to random, erroneous drone behavior. Every time DJI releases a new firmware/software version to be more and more compliant with ever-growing FAA regulation, make no mistake; YOU ARE THE BETA TESTER. I run legacy (read: YEARS OLD) firmware and software on ALL THREE of my birds (Inspire 1, v2, Mavic Pro and P4P), and I could not be happier. My drones are rock-solid reliable, but more importantly, BEHAVE PREDICTABLY.

Here's a list of the FW/Software I use:

Inspire 1:

Go App: v3.1.1
App database: v00.00.01.04
Aircraft: v1.8.1.00
Remote: v1.6
X3 Camera: v1.8.1
X5 Camera: v1.11.1.50


Mavic Pro:

Go App: v3.1.1
Go4 App: 4.0.8
Aircraft: 01.03.0700
Remote Controller: 01.03.0700
Flight Database: 00.00.01.04
Basic App Fly Safe Database: 01.00.01.05
Precise App Fly Safe Database: 01.00.01.05


Phantom 4 Pro:

App: v4.0.8
Aircraft: v1.3.509
Remote: v01.04.01.00
Precise Fly Safe database: v00.00.01.04
Basic Fly Safe database: v00.00.01.04


4) When flying automated flights, an elevated level of diligence is required. Most of my jobs utilize Intelligent Flight Modes or Waypoint automation. Because I don't rely on Object Avoidance, I HAVE to understand my environment, the path the drone is going to take, and more importantly, be on the ready to stop the mission should something go awry. The "ATTI" or "Sport Mode" buttons are your panic buttons. If you're flying an automated flight that looks like it might land you in a tree or smash you into the side of the building, toggle these switches. You can bring the drone back to GPS mode.

5) Test, test, test. Never once have I ever gone into the field with an untested platform. EVERY aspect of that job has been tested with a near-duplicate of that mission under controlled circumstances like a park or open field.

My partner and I just got a job doing extended LIDAR missions with his M600. You can absolutely bet that we are going to test the bejesus out of this BEFORE loading up the vehicle and traveling 5 hours to the job site. I have planned dummy payload flights to simulate the LIDAR unit's payload and current draw. Once those tests have flown successfully, we'll move on to tests with the actual LIDAR unit, which includes data acquisition. You can bet that we will reproduce every nuance of our planned mission, including battery changes mid mission and exit strategies, but in a controlled environment. Which brings me to #6.

6) ALWAYS have an exit strategy. Practice abandoning and recovering from rogue missions. Put yourself in pretend, emergency scenarios under controlled conditions. For example, what if you're flying a remote mapping mission and a low-flying helicopter comes into your mission space? You have to:

A) Pause the mission.
B) Bring the drone down to a safe altitude.
C) Continue the mission.

Without practicing this, you might panic and/or not be able to continue your mission, which means starting over. I HATE starting over.

Anyway, I hope all this helps. I posted this because I see crashes and runaways almost daily in the 3 forums I belong to. I don't fly every day, but I DO fly every week. So my experience is valid.

Good luck!

D

What other forums do u belong to?

 

[Donnie Frank]

What other forums do u belong to?

The Inspire Pilots forum and the Phantom Pilots forum.

D

 

[Arce101]

Ok I have mavic zoom I agree with you fully no issues I’m running nld mod on my mavic lowest firmware and had no problems...(deleted by moderator...) check this out all the firmware updates I think in my opinion is a way for dji to crash our drones by implementing a supposed fix to software issues that don’t exist in all actuality to get us to fix our drones after the crash but I could be wrong that’s just my take on it.

 

[GadgetGuy]

No firmware or app is bug free. Even from day one, we find issues. However I would agree sticking with a version of firmware/app that you are most comfortable with where you know what the quirks are and how to deal with them.
As for app, that means you also have to stick with a mobile device and OS and never change those.

Sandbox everything!

 

[FASTFJR]

About 7 years of flying DJI products, somewhere around 3000 flights and never crash.......Until this past June when I was in Bermuda flying off a beach with cliffs on one side. After a 5 minutes of flying I decided I wanted to sit down in my beach chair. Brought the drone to a hover about 50ft away from me and about 20ft from a large rock. As I was sitting down I bumped the controller against my leg and it took off towards the rock. Almost saved it. Luckily the drone was fine, cracked a prop and scuffed up the body.