Mini not responsive to controls, apparently crashed into building on RTH

[kmiller416]

In late November I received a Mini as a gift from a friend who was upgrading. I had no drone flying experience, so tried to get familiar with the drone with a few indoor flights and a number of outdoor test flights in open areas. After feeling at least somewhat confident in basic operation, I decided I would do a simple test flight from my 23rd floor balcony on a day with light winds, just planning to take it out no more than 20 m, move it back and forth a few times, and then bring it back in. While I knew the RTH feature was highly inadvisable with a balcony launch, I set the RTH altitude to be able to clear the building just in case.

I was able to take off and move the drone away from the building a few meters without difficulty, but then the drone seemed to be wandering quite a bit in the absence of input, and was frequently extremely sluggish to my inputs. About 5 minutes into the flight when I was planning on bringing the drone back, it started moving away from me to a distance of 25 m in the general direction of a busy intersection and seemed non-responsive to my inputs. My inexperience as a drone pilot made me think it was farther out than it actually was, and I panicked and initiated RTH, which the drone responded to, moving to very high altitude and then returning in my direction. I knew it would not automatically land back on my balcony, so when it reached my approximate location I cancelled RTH, but unfortunately it appeared to already be over the roof area and I couldn't see it. The log shows a couple of "GPS signal weak. Hovering unstable." errors just after RTH cancellation and not being able to see it, I refrained from any control inputs. About a second later, I started getting "Not enough force/ESC errors" and the drone appeared to be wandering somewhere around roof level. About 10 seconds later I got a "Max power load reached. Fly with caution. Not Enough Force/ESC Error" and auto-landing mode initiated a second later. Later, I got a "Propeller(s) rotating too fast" error and finally "Motor is Blocked. Motor unable to rotate." just before I lost contact, which I suspect involved a crash somewhere on the building, and unfortunately it appeared the drone settled on a ledge somewhere on the building. I've asked the building maintenance staff to try to locate it from the roof, but they tell me there's no sign of it.

I've attached the flight log. If someone is inclined to look at it and let me know if there's evidence of malfunction when the drone seemed to be wandering on its own and was sluggishly responsive to control inputs at best, or if the sequence of errors at the end is indicative of a malfunction, I would greatly appreciate it. While the drone is out of warranty, it was originally purchased with a credit card that provides manufacturer's warranty doubling, so if I can provide evidence of malfunction, there's a reasonable chance they'll cover the cost of a new one.

 

[Kilrah]

Not seeing anything wrong other than EXTREMELY erratic banging of the sticks during most of the flight and probably quite some wind/turbulence (close to tall buildings is really not a good place to fly unless you know for sure there's very little or no wind). The impression of no control was likely a combination of both, but mostly your erratic orders, I'd expect the aircraft was flying just fine otherwise.

RTH was guaranteed to be a disaster since home point was in the building, but it seems it operated just fine until it hit something while descending, all the errors are a result of that.

 

[icecoldgin]

Hmmm, "motor unable to rotate" was what caused my Mini to crash ?

 

[Tony McDrone]

The error message “motor unable to rotate” is sent when the drone has crashed and is upside down and the motors are still trying to rotate. There may be other instances when you’ll receive this message.

 

[MM2]

In late November I received a Mini as a gift from a friend who was upgrading. I had no drone flying experience, so tried to get familiar with the drone with a few indoor flights and a number of outdoor test flights in open areas. After feeling at least somewhat confident in basic operation, I decided I would do a simple test flight from my 23rd floor balcony on a day with light winds, just planning to take it out no more than 20 m, move it back and forth a few times, and then bring it back in. While I knew the RTH feature was highly inadvisable with a balcony launch, I set the RTH altitude to be able to clear the building just in case.

I was able to take off and move the drone away from the building a few meters without difficulty, but then the drone seemed to be wandering quite a bit in the absence of input, and was frequently extremely sluggish to my inputs. About 5 minutes into the flight when I was planning on bringing the drone back, it started moving away from me to a distance of 25 m in the general direction of a busy intersection and seemed non-responsive to my inputs. My inexperience as a drone pilot made me think it was farther out than it actually was, and I panicked and initiated RTH, which the drone responded to, moving to very high altitude and then returning in my direction. I knew it would not automatically land back on my balcony, so when it reached my approximate location I cancelled RTH, but unfortunately it appeared to already be over the roof area and I couldn't see it. The log shows a couple of "GPS signal weak. Hovering unstable." errors just after RTH cancellation and not being able to see it, I refrained from any control inputs. About a second later, I started getting "Not enough force/ESC errors" and the drone appeared to be wandering somewhere around roof level. About 10 seconds later I got a "Max power load reached. Fly with caution. Not Enough Force/ESC Error" and auto-landing mode initiated a second later. Later, I got a "Propeller(s) rotating too fast" error and finally "Motor is Blocked. Motor unable to rotate." just before I lost contact, which I suspect involved a crash somewhere on the building, and unfortunately it appeared the drone settled on a ledge somewhere on the building. I've asked the building maintenance staff to try to locate it from the roof, but they tell me there's no sign of it.

I've attached the flight log. If someone is inclined to look at it and let me know if there's evidence of malfunction when the drone seemed to be wandering on its own and was sluggishly responsive to control inputs at best, or if the sequence of errors at the end is indicative of a malfunction, I would greatly appreciate it. While the drone is out of warranty, it was originally purchased with a credit card that provides manufacturer's warranty doubling, so if I can provide evidence of malfunction, there's a reasonable chance they'll cover the cost of a new one.

The reason your Mini crashed was basically because of you. You should not have been flying where you were for starters and your aggressive stick inputs caused the Mini's ultimate demise. Good luck with your warranty claim, you'll need it.

 

[Meta4]

I've attached the flight log. If someone is inclined to look at it and let me know if there's evidence of malfunction when the drone seemed to be wandering on its own and was sluggishly responsive to control inputs at best, or if the sequence of errors at the end is indicative of a malfunction, I would greatly appreciate it. While the drone is out of warranty, it was originally purchased with a credit card that provides manufacturer's warranty doubling, so if I can provide evidence of malfunction, there's a reasonable chance they'll cover the cost of a new one.

What I see in the flight data is different from what's been suggested but still doesn't show any malfunction.

You initiated RTH at 5:40.2 when the drone was only 55 metres away and at almost the same height as your launch point.
RTH climbed to 43 metres and started to bring your drone back
At 6:02.9 you tried pulling the left stick down while in RTH And brought the drone down to 35 metres.
At 6:16.8 RTH had brought the drone back to 35 metres directly above the launch point and a collision happened.
The drone started falling and tumbling until 6:19.8 when the drone recovered from the collision and stopped falling.
The drone made an unexplained climb of 13 metres up to 46 metres and you initiated autolanding.
From 6:31 the VPS sensors are showing that the drone is 6 metres above something and descending.
At 6:32.3 the pitch and roll data shows another collision and the drone falling and tumbling again.

Data stops with the drone 24 metres higher than the launch spot but VPS is indicating that it has run into a surface up there.

If this data was sent to DJI they would interpret it as showing the operator was responsible for choosing a poor location and not flying safely.

 

[boblui]

The way you flew the drone is very interesting. You kept pumping the elevator and aileron sticks to the limits and you did it at a rather high frequency. In the 13-second time interval below, you have pumped the aileron 25 times, on average 2 times per second. More interestingly you only pumped to the right. Is there any reason for that ?



The same is seen for the elevator control.

Doing so is like quickly alternating between the pressing of the throttle and brake pedals of your car. You can't expect the car to move much. That's why the drone responded to control inputs sluggishly.

 

[kmiller416]

Thank you for your analysis and advice. I can provide some context. The original objective of this flight, was a short, play around with the controls for a few minutes training flight. Most of the flight was on the same level as my balcony, and winds as experienced from the balcony were extremely light. I had done a ground level, open-area outdoor flight a few days earlier where winds were significantly stronger, and the drone handled it well and was quite stable. So it really surprised me that on this flight, the drone could not seem to hover in a stable position, and seemed biased to the left, as you can especially see from about 50 s to 64s when the drone moves about 12 m to the left despite a few right inputs. After that, it generally seemed like I had to apply frequent right input pulses just to keep the drone in front of me instead of drifting left. Nonetheless, I was mentally keeping it together until 325 seconds when the drone moved around 25 m further away from me over 15 s despite no input in that direction. My impression was this was much farther than it actually was, and thought it may be a flyaway with the drone moving in the direction of a busy intersection, so, as I mentioned in the OP, I panicked and engaged RTH. I had the RTH altitude set higher than the height of the building, neglecting that my takeoff point was already about 80 m up. So, the drone was probably attempting to go to about 230m actual altitude (the building is a little over 100 m in height), and I'm guessing the winds at higher altitude were too much for it, and hence the "Not Enough Force/ESC Error" messages, and so the drone couldn't clear the building and contact ensued, ending in the crash landing of the flight somewhere on the building above me.

Thanks as well to everyone else for your honest, if harsh, appraisals, although there seems to be the notion by some that I was attempting to blame this on a drone malfunction. I think I was clear in the OP that I know I made some serious errors both in operation and judgment, but I was also wondering, based on the drone's inability to hover stably earlier in the flight, and the Not Enough Force/ESC errors later, whether drone malfunction was a contributing factor. It appears the answer is, probably not, there was just much stronger wind gusts than I realized, and I overreacted to the unexpected (but normal) behavior of the drone under the circumstances. If I get a new Mini, which is likely, I'll know I need a ton of practice in an open area, both to improve my competence and confidence, and avoid being ham-fisted with the controls.

 

[Kilrah]

and winds as experienced from the balcony were extremely light.

That's the problem with tall buildings, if the wind comes from the opposite side of the building from where you are you can feel absolutely nothing from your balcony, but 5-10m away you get strong wind and turbulence from the air rotors rushing back in towards the building...

 

[boblui]

.... the drone could not seem to hover in a stable position, and seemed biased to the left, as you can especially see from about 50 s to 64s when the drone moves about 12 m to the left despite a few right inputs. After that, it generally seemed like I had to apply frequent right input pulses just to keep the drone in front of me instead of drifting left. ....

OK, I finally saw it among the sea of noisy stick inputs and spagetti-like flight path ....

From 192 to 200 sec, the craft slide to the left by itself with no aileron input :



Same for 214 - 226 sec. In this period, the aileron stick has been pumped to the right for a few times but the craft kept sliding to the left :



Now I agree that there might be something wrong with the drone .....

 

[boblui]

Regarding the RTH problem, you have set the RTH height to 153 m which is way more than enough to clear the building roof. However, you applied fully-down throttle when the craft has ascended to 43 m so it stopped ascending further and flew to the home point at that altitude. My M2P behaves in the same way.