Roman Bapst

@MaEtUgR > Ideally you could ask the geofence library: How close is the next geofence? Please keep in mind that I consider a geofence lookup an expensive operation as it...

@NAmmann > From my last experiences with the geofence avoidance i think the premature trigger of the geofence action and other actions like pushing a drone away from the geofene...

@NAmmann Interesting, thanks for the sharing the graphical illustration. I would suggest the following: 1) Draw the geofence in QGC using the two volumes you described. 2) Trigger the failsafe...

@roman-dvorak Great to see your work on auto-gyros! I briefly went through the PR and noticed quite a few places where there are some hacks in place. I think you...

@priseborough I started looking into this a bit and found that something causes the height state covariance to decrease during horizontal maneuvers. I replayed the bad log (with and without...

@priseborough Could you let me know if replaying the posted logs works for you? I can replay them but the replay log is very short (maybe one second).

@priseborough I can provide replay logs tomorrow morning. I tried to set the rows/columns corresponding to vertical velocity and height of the KHP matrix to zero but that did not...

@priseborough I did an outdoor flight today in which I zeroed the entry of the optical flow observation vector corresponding to the vertical velocity state. The logs shows that this...

@priseborough Hm, what is your conclusion from that?

@priseborough I think I found the mistake I made. I forgot that the kalman gain was computed independently of the H_LOS matrix. I zeroed the entries in the H_LOS matrix...