[Dprglist] RBNV Addenda

David P. Anderson davida at smu.edu
Tue Jan 25 23:27:21 PST 2022


Howdy DPRG,

Really good RBNV this evening.   So much progress.   Makes me feel like 
a slacker :)

Chris, very cool python implementation of the navigation algorithms.   
Definitely Gold Star worthy.   In the video the robot is clearly 
steering and correcting in real time.  And the odometery location 
doesn't really drift much.  Using the IMU for heading appears spot on.   
Also gratifying to hear that it was not that hard to port and get it to 
work, and incorporate into your existing environment.

Seems like you might now be able to throw in a little rudimentary 
obstacle avoidance to bounce the robot away from detections, and the 
navigation should handle the rest.  So one might toss a box in it's way 
while it drives it's pattern...

Doug,I have a correction to make.  The navigate turn constant on Rcat is 
120 (i.e., 12.0%) not 600.  I read the wrong parameter. The navigation 
rotation rate is 8, so 15 times through the control loop to reach full 
value if starting from 0, a little more than 1/2 sec at 25 Hz.

I had a good belly laugh thinking about our newest Gold Start video rule 
(has to be made since last RBNV) and how we are slowly making up rules 
for this game.  Organically, as it were.

Our First Grade teachers would be proud.

cheers!

dpa





More information about the DPRGlist mailing list