Loading video...
Video Failed to Load
Fixed a bug in the AHRS fusion code that resulted in weird behaviour on inversions; it's much more stable now. Need to write a CAN comms layer and then it's finally back to actually doing stuff on the robot!
46,490 views • 1 year ago •via X (Twitter)
11 Comments

I'll probably write up a quick thing on how the orientation estimation math works later, it's conceptually quite straightforward and fun to see working.

Heightened volatility got you on edge? In my latest free Substack post, discover how a Hidden Markov Model (HMM) can help you navigate market corrections and safeguard your investments.

Very smooth! How stable is the orientation over time?

Thanks! I'm correcting based on the accelerometer so only yaw drifts. The gyro readings are off by about a degree per minute but seem stable, so calibration should help. There's a magnetometer on there, but I probably won't integrate for a bit.

why waste time reinventing the wheel

love it

What is the HW of this? Any (magnetic) compass included? Thinking about marine applications (boats, ...) with NMEA2K/CAN.

What's the drift on this and how do you correct for it?

That looks super cool! I want one!🫡😄🇺🇸

Nice! Hows the state estimation work, madgwick or Kalman based ?

Reminds me of the UM7-LT. Very impressive.
