swarm.reset_estimators() fail to complete successfully

[scf-ku]

Hi,

I am currently trying flight tests using the Crazyflie 2.1 Brushless, AI-deck, Multiranger deck, and Flow deck.
I am attempting to use multiple units to form a swarm, but errors are occurring at specific points.

When attempting to run the sbs-swarm sample code with the above configuration, the process always stops at reset_estimators().

Are there any effective methods to address the above issue?

Best Regards.

[ArisMorgens]

Hello @scf-ku!

Really interesting to hear that you're testing all these different decks.

The function reset_estimators() is a safety measurement that blocks the Crazyflies from taking off if you they don't have an accurate position estimation. In your case, this could mean that the Flow deck is not attached, or that it can't position the Crazyflie well.

Let me know if you managed to figure it out or if we should further investigate it.

[scf-ku]

Thank you for your response.

This is the console screen obtained from cfclient.

IMU: BMI088: Using I2C interface.
IMU: BMI088 Gyro connection [OK].
IMU: BMI088 Accel connection [OK]
IMU: BMP390 I2C connection [OK]
ESTIMATOR: Using Kalman (2) estimator
CONTROLLER: Using PID (1) controller
MTR-DRV: Using brushless motor driver
SYS: About to run tests in system.c.
SYS: NRF51 version: 2024.10 (C21B)
EEPROM: I2C connection [OK].
STORAGE: Storage check [OK].
IMU: BMI088 gyro self-test [OK]
DECK_CORE: Deck 0 test [OK].
MR: Init front sensor [OK]
MR: Init back sensor [OK]
MR: Init up sensor [OK]
MR: Init left sensor [OK]
MR: Init right sensor [OK]
DECK_CORE: Deck 1 test [OK].
DECK_CORE: Deck 2 test [OK].
SYS: Self test passed!

It seems to be Flow deck and other two decks also working normally.
Is there anything else I should check?

I've also attached the results of running sbs_swarm.py with three different deck configurations.
The program execution does not proceed further.

Connected to Crazyflies
Light check done
Waiting for estimators to find positions...

[ArisMorgens]

Hello again!

Thank you for attaching the console output. You're right - all 3 decks seem to be working normally.

This might have to do with the optical sensor of the Flow deck and the surface underneath it. Basically, if the surface doesn't have texture, the position estimates start drifting.

Having said that, I think you can comment out the reset_estimators() command as it's mostly used for the absolute positioning systems like the Lighthouse or the Loco. Since the Flow deck works normally, you can skip checking for accurate positioning when the Crazyflie is on the ground. When it'll take off, a larger area will be visible and the Flow deck will stabilize the drone.

If you want to avoid that, you could you try powering on the Crazyflie exactly at the point that you want it to take off from, meaning not moving it after powering it on.

Let us know if this works and keep us updated on your progress.