Possible LAND Detection Issue Under Low Simulated Motor Vibration

I found a possible issue where small vibration interference may affect landing detection / disarming.

Under small vibration interference (VIBE below 20m/s²), after the vehicle touches the ground in LAND mode, it remains armed, does not enter ON_GROUND, and does not automatically disarm. In Mission Planner, the vehicle can be observed continuously hitting/bouncing near the ground instead of completing landing and stopping. The logs also show that motor outputs continue after touchdown, followed by another SIM Hit ground, EKF lane switch, GPS Glitch or Compass error, Vibration compensation ON, and other related messages.

This phenomenon may indicate a robustness issue in LAND mode’s landed detection or disarm logic under small simulated motor-vibration interference, or that the protection provided by Vibration Failsafe may be insufficient in this scenario.
It may be worth checking the robustness of LAND detector / disarm logic and Vibration Failsafe protection in this scenario.

This can be reproduced in SITL by doing the following:
param set SIM_ACC1_RND 0.35
param set SIM_ACC2_RND 0.35
param set SIM_ACC3_RND 0.35
param set SIM_VIB_MOT_MASK 15
param set SIM_VIB_MOT_MULT 20

guided
arm throttle
takeoff 6
loiter
wait until the vehicle is stable
land

Post a .bin log demonstrating the issue.

Hi @Claire-190,

This sounds very much like an AI generated report and is simply based on simulation, not a real world test.

Yes, it is possible that high vibrations can stop the landing detector from working. High vibrations cause all kinds of problems on real vehicle so it’s an important factor to address when building the hardware. It’s not a software issue though, the software works within specified limitations (<30m/s/s of vibration normally, small peaks up to 60m/s/s may be OK). If the hardware can’t do that, then it’s a hardware problem.

Some relevant wiki pages:

1 Like

Thanks. I uploaded the .bin log here:
Release LAND low-vibe SITL log · Claire-190/SITL-LOG

To clarify, this is only a SITL observation.

What made me think it may be worth checking is that the logged VIBE stayed below 20m/s/s and never exceeded 25/30m/s/s. In the no-vibration control run, LAND disarmed normally after touchdown. In this low simulated motor-vibration run, LAND touched down but stayed armed, LDET.Count did not build up, and RCOU outputs continued after touchdown.

I agree that excessive vibration is normally a hardware issue. My question is whether this low-reported-VIBE SITL behavior is expected.

Hi @Claire-190,

This is all expected and these constant AI queries are just eating up developer time with no real benefit to users.

We appreciate the use of AI for real contributions but please understand the cost to our team

Thanks for the clarification, and sorry for the noise.

I was trying to check whether this SITL behavior was expected before filing anything more formal. I understand your ponit and will be more careful about what I post here.

1 Like

None of the flags in LDET that indicate landing rise in the log.