Closed osrf-migration closed 11 years ago
Original comment by Maurice Fallon (Bitbucket: mauricefallon).
I've been playing with the IMU in the current model. Obviously the orientation in the message is perfect - as its read from the world pose.
However the linear accelorations and angular rates seem terribly off. For example when absolutely stationary I would have expected -9.81 but I get hugely different values.
Do you think that these can be reliably extracted from ODE? Or has anyone used them in the past? (I just don't want to do IMU+Vision fusion of this wasn't possible.
Also, I understand plans are to move to Bullet so addressing this effort now for ODE is probably a waste of time
Original comment by Steve Peters (Bitbucket: Steven Peters, GitHub: scpeters).
Original comment by Maurice Fallon (Bitbucket: mauricefallon).
My comment above was about the quality of the IMU data but there is currently NO IMU in the head at all.
Could one please be added before the next release of DRCSIM?
I talked with David at CRL and he confirmed it has one - its also in the specs here: http://www.theroboticschallenge.org/local/documents/MultiSense_SL.pdf
This is really important for sensor perception - esp for orientation estimation.
Original comment by John Hsu (Bitbucket: hsu, GitHub: hsu).
drcsim 2.0 will publish head link states inside a sensor_msgs/Imu message under /multisense_sl/imu
topic, albeit it now needs an additional patch correctly situate the imu according to spec.
To the best of our knowledge, according to CRL,
Gyro and Accelerometer Mounting Locations:
Here are links to datasheets on the IMU components we’re using:
Expected Future API:
Original comment by Maurice Fallon (Bitbucket: mauricefallon).
Thanks John, I didn't see one in the urdf but i do see it in the plugin code now.
Original comment by Maurice Fallon (Bitbucket: mauricefallon).
Theres no gravity in the IMU readings! This is a plot of the IMU data. Top right plot is the acceloration reading: z should be -9.81:
Here is an example for real data from a $1800 MicroStrain IMU:
And the same sensor in real motion:
Original comment by Maurice Fallon (Bitbucket: mauricefallon).
ps: the units are seconds and the gazebo plot was the atlas swaying back and forward slightly
Original comment by John Hsu (Bitbucket: hsu, GitHub: hsu).
added gravity vector to imu sensor in pull request
Original report (archived issue) by John Hsu (Bitbucket: hsu, GitHub: hsu).