I caught ROSbot Dave drooling this morning over the ease with which Turtlebot4 robots can map their environment …
Asynchronous SLAM will update the map as fast as the processor running it can handle. This may cause it to drop some laser scans or odometry data. Maps created with asynchronous SLAM may have reduced accuracy and detail, but this method requires significantly less proccessing [sic] power. This approach is ideal for use on the TurtleBot 4’s Raspberry Pi. The default parameters for asynchronous SLAM use a reduced map resolution to further improve performance on the Pi.
Launching asynchronous SLAM:
ros2 launch turtlebot4_navigation slam_async.launch.py
ROSbot Dave also noticed that I am playing with a Pi4 on my desk and hoping for a processor upgrade to come his way. It seems unfathomable to me today that I didn’t realize that the PiMoroni “heat sink case” does not have threaded holes on top to allow mounting the GoPiGo3 red board. I was thinking I just put a 40pin extra tall stacking header on the GPIO pins and mate the GoPiGo3 board. “The light bulb clicked” yesterday and I’m devastated. I was thinking I had solved the Pi4 on GoPiGo3 cooling problem, but was actually “not thinking”.