DJI Drone Help Forum

Mavic Air 2 frontal crash into wall during RTH

Hi Pilots,
New to DJI drones and MA2, so hope I can find some insights here to understand what happened.
Last week after issuing RTH, the MA2 headed towards the home point direction but flew past it and hit a brick wall 4 meters behind. There was no response to any controller commands before it hit the wall and fell 3 meters. Was in normal mode and obstacle avoidance and APAS enabled in settings. (distance to take-off/home point about 10 meters and 4 mts altitude),
Result is a dent in the nose and four broken propellers. I don’t know if there is any internal damage after the fall but it seems to respond well after replacing the propellers, but afraid to take it out far for fear of a similar incident and causing more damage or losing it.
For the 30-odd flights the past three weeks, RTH functioned pretty reliable and accurate, except one occasion where it headed in the opposite direction, but I could manually regain control.
I’ve create a DJI support case and have been issued a repair return label without any explanation of what may have caused this behavior. Most worrying is that even if the home point was off, why was there no obstacle avoidance active. Looking at the logs (AirData), some seconds after RTH kicked in, there is an entry “Lighting environment too dark. Vision System unavailable” but the drone kept on flying uncontrollable into the wall. As I understand from the documentation, if between 5 and 20 m, it maintains current altitude and heads home.
When aircraft is flying forward:

  1. The aircraft brakes when an obstacle is sensed from in front and files backward to a safe distance. It ascends until no obstacle is sensed and continues to ascend another 5 m and then continues to fly forward. No lost connection so no failsafe RTH, batteries at 50%, enough satellites…

Hope someone can shed some light on the issue or share some good thoughts.

Can you supply the flight logs for your incident? That will provide the basis for informed answers. Look here
to see how to get those logs.

Hi Bud, appreciate the response. Excellent work with the log retrieval topic.
The flight logs (.txt) were uploaded to
(Was shooting my son doing lay-ups and dunks so it’s a wierd flight path).
It seems anyone can also access the Airdata journal with the direct link, and view the HD flight player details - let me know if that can help.
Also the latter part of the DatCon log breakdown shows:
663.811 : [L-API]go_home_check
663.811 : [L-CMD]basic_cmd_table send_id 6
663.811 : [L-API]basic_cmd_go_home
663.812 : [L-N_MIS]Init plan 0 from requester 5 for reason 18. Pre plan 5
663.822 : [L-N_MIS]Exit mis 14. New mis 12
663.822 : [L-N_MIS]send cur mission exit data, mis:0, exit_reason:0.
663.822 : [L-N_MIS]Switch new mis 12 req 5 from mis 14 req 4
663.902 : [L-SDK]app API sub_mode 7 2
663.902 : [L-MIS][NAVI MISSION] req sub_mode: AutoAvoid on
663.902 : [L-SDK][USER]send ack 0x00
663.905 : [L-MIS][SDK] new open GS(cam)
663.905 : [L-MIS]received on/off 2 from dev 2
663.905 : [L-MIS]ack on/off 0 to can
663.913 : [L-MIS][SDK]reset user_emergency_brake by ctrl dev changing
668.284 : [L-FMU/LED]type:0, normal flash action:
668.284 : [L-FMU/LED]c0:0,1;c1:0,1;c2:0,1;c3:2,3;c4:0,1;c5:0,1;c6:0,1;c7:0,1
669.167 : [L-FMU/LED]type:0, normal flash action:
669.167 : [L-FMU/LED]c0:0,15;c1:0,3;c2:0,13;c3:2,3;c4:0,10;c5:0,3;c6:0,10;c7:0,3
681.112 : [L-NS][AHRS],qg obv buffer size[215/270]
690.239 : [L-N_MIS]req abort all for App Go-Home Exit
690.245 : [L-N_MIS]send cur mission exit data, mis:0, exit_reason:0.
690.245 : [L-N_MIS]Switch new mis 3 req 2 from mis 25 req 17
690.860 : [L-MIS][SDK] new open GS(cam)
690.860 : [L-MIS]received on/off 1 from dev 2
690.860 : [L-MIS]ack on/off 0 to can
690.887 : [L-N_MIS]Exit mis 3. New mis 14
690.887 : [L-N_MIS]send cur mission exit data, mis:0, exit_reason:0.
690.887 : [L-N_MIS]Switch new mis 14 req 4 from mis 3 req 2
690.890 : [L-MIS][SDK] new open GS(cam)
690.890 : [L-MIS]received on/off 1 from dev 2
690.890 : [L-MIS]ack on/off 0 to can
690.910 : [L-MIS][SDK] new open GS(cam)
690.910 : [L-MIS]received on/off 1 from dev 2
690.910 : [L-MIS]ack on/off 0 to can
690.920 : [L-SDK]app API sub_mode 7 2
690.920 : [L-MIS][NAVI MISSION] req sub_mode: AutoAvoid on
690.921 : [L-SDK][USER]send ack 0x00
690.938 : [L-SDK]app API sub_mode 7 2
690.938 : [L-MIS][NAVI MISSION] req sub_mode: AutoAvoid on
690.938 : [L-SDK][USER]send ack 0x00
709.711 : [L-N_MIS]Init plan 0 from requester 6 for reason 12. Pre plan 5
709.721 : [L-N_MIS]Exit mis 14. New mis 12
709.721 : [L-N_MIS]send cur mission exit data, mis:0, exit_reason:0.
709.721 : [L-N_MIS]Switch new mis 12 req 6 from mis 14 req 4
709.809 : [L-SDK]app API sub_mode 7 2
709.809 : [L-MIS][NAVI MISSION] req sub_mode: AutoAvoid on
709.810 : [L-SDK][USER]send ack 0x00
709.819 : [L-MIS][SDK] new open GS(cam)
709.819 : [L-MIS]received on/off 2 from dev 2
709.819 : [L-MIS]ack on/off 0 to can
709.832 : [L-MIS][SDK]reset user_emergency_brake by ctrl dev changing
717.352 : [L-FDI][CTRL]: fault on , tilt_ctrl_fail
717.352 : [L-CFG]PM:[fc_fault_log_error_set_status] “fault_log_info.index” to 35||
717.352 : [L-CFG]PM:[fc_fault_log_error_set_status] “fault_log_info.reason” to 4||
717.352 : [L-CFG]PM:[fc_fault_log_error_set_status] “fault_log_info.flag” to 1||
717.352 : {tska}INFO:fc fault log error reason: 0x044, flylog index: 35||
717.371 : [L-RC]craft ctrl failed!!!
717.390 : [L-CFG]PM:[ack_cb_fc_fault_log] “fault_log_info.flag” to 0||
717.390 : {tska}INFO:fc fault log ack ok!||
717.493 : [L-FDI][CTRL]: fault off, tilt_ctrl_fail
717.572 : [L-N_MIS]Exit mis 12. New mis 11
717.572 : [L-N_MIS]send cur mission exit data, mis:0, exit_reason:0.
717.572 : [L-N_MIS]Switch new mis 11 req 6 from mis 12 req 6
717.573 : [L-FDI][CTRL]: fault on , tilt_ctrl_fail
717.713 : [L-FDI][CTRL]: fault off, tilt_ctrl_fail
717.733 : [L-FDI][CTRL]: fault on , tilt_ctrl_fail
717.894 : [L-FDI][CTRL]: fault off, tilt_ctrl_fail
718.054 : [L-FDI][CTRL]: fault on , impact_in_air
718.054 : [L-CFG]PM:[fc_fault_log_error_set_status] “fault_log_info.reason” to 2||
718.054 : [L-CFG]PM:[fc_fault_log_error_set_status] “fault_log_info.flag” to 1||
718.054 : {tska}INFO:fc fault log error reason: 0x042, flylog index: 35||
718.072 : [L-CFG]PM:[ack_cb_fc_fault_log] “fault_log_info.flag” to 0||
718.072 : {tska}INFO:fc fault log ack ok!||
718.094 : [L-FDI][CTRL]: fault on , tilt_ctrl_fail
718.153 : [L-FDI][CTRL]: fault off, impact_in_air
718.172 : [L-GIMBAL]reset gimbal acked len 1 buf 0, cb_type=0
718.235 : [L-FDI][CTRL]: fault off, tilt_ctrl_fail
718.333 : [L-FDI][CTRL]: fault on , tilt_ctrl_fail

Cheers, Maz

Thanks, but I didn’t write that guide. @sar104 put that guide together.

Please provide the .DAT.

I can’t tell from the GE image where the wall is but there were two home points recorded. First at 50.953263 , 3.816569 and then for the rest of the flight 50.953252, 3.81659. Would this second HP have caused an impact with the wall?

Question. What was the altitude setting for return to home?

Hi, I will need some guidance on uploading the .DAT as I get Error: You may only upload DJI GO TXT flight logs, and “Sorry, new users can not upload attachments” uploading here.

I didn’t notice the different home points, and quite sure I didn’t update in flight as I only moved out towards the garden. The wall was about 4 meters behind the launch point. The flight path overlay on the map is off by about 3 meters, so hard to specify, but it should have been roughly 50.953212, 3.816578 if I check Google Maps. The coords 50.953252 is the wall where it crashed. 50.953263 is inside my living room! Becoming more of a mystery…

Hi Starz, 20m (66 ft)
“Setting new Return-To-Home altitude to 20m (66 ft). Data Recorder File Index is 35. Setting new Maximum Flight Altitude to 20m (66 ft)”

Try using Dropbox or GoogleDrive.

Hi Bud, files posted to

GPS is only accurate to within ~20m. Use the calculations in that proximity with great care. That is a very tight window to calculate. JMO. You will most likely be taking these numbers with a grain of salt so to speak. @BudWalker…Thoughts?


HP was set during starting rotors, but 1.9 meters away from drone position.
HP was reset during takeoff,1.8 meter different again. GPS signal reflecting house/wall, so no accurate position was set.

Looks like that the HP was set ‘at’ the wall position.

During Last RTH no OA available (low light), pushing backward stick did reduce to speed but guess to late. If overshooting the landing spot best is to press the pause/cancel button once!


The HP was set twice. First at -123 secs and then at 1.766 secs.
-123.099 : [L-HOME]First. lati:50.9532625 longti: 3.8165692 alti: -0.53
1.766 : [L-HOME]Update. lati:50.9532524 longti: 3.8165895 alti: -0.54
This is normal behavior. The first setting is just to have something if launch occurs before the FC has settled on a location solution. The update occurs at launch.

Regardless, the real question is was the MA2 closing on the HP or was it moving past the HP? The FC thought that the MA2 was closing on the HP as can be seen here

The top chart shows the distance decreasing to 0.0 and the bottom shows the impact with the wall.

I agree with @Fly_Dawg and @JJB that the HP wasn’t set quite correctly.

Any particular reason for this relatively low setting for return to home. Maybe I’m missing something.

When close to home (when RTH initiated) Craft returns to home at current height.
For every DJI drone this is different, check the manual.


Probably better to ask this.
Any particular reason to hit the return to home button when the ac is only 30 feet away and 12 feet off the ground. Am I looking at this wrong?

why not… :slight_smile:

But always be prepared to pause a automatic mode, by pressing the button!
If you see that drone is overshooting the, what you think is the HP spot, button press!

Thanks everyone, appreciate the input even if I can’t understand it all at the moment. I’ve used to RTH way too often at short distances as I wanted to test out the different algorithms. Adn so far it worked out well so maybe I got too confident. I won’t forget to have a finger on the kill switch next time.
But I still cant understand why obstacle avoidance or apas didn’t work at all, or is this normal behavior. the low vision system kicked in just after RTH, so I assume full autonomous mode, or?
And could it be a compass calibration issue of the home point effset?

I’m still trying to understand the use of return to home at such a short distance and VERY low altitude.
Please enlighten me.

Hi Starz, no specific reason except trying out the different features and capabilities - still in learning mode.

HP is set by GPS data of the drone, nothing to do with the drone or mob device compass
OA need enough light to work properly, (too) low light conditions = no OA