Turtlebot3 reborn

I finally found the TurtleBot3 robot!  It was in a box, under a desk, hidden away.  I thought I'd lost it.

Turns out I bought this a LOT longer ago that I remember:

https://tonyisageek.blogspot.com/2021/10/turtlebot3-arrived-and-well-i-couldnt.html

https://tonyisageek.blogspot.com/2021/10/turtlebot3-status-update.html

It turns out that when I installed and set this all up before, the world has moved on quite significantly and the versions of everything have moved on.

So I did what anyone would do... I wiped the SD-Card and started the installation again from scratch.

What it did need was the ability to install all the graphical front-end software onto a "decent" laptop - so I repurposed the IBM ThinkPad as per the instructions and then proceeded to basically follow the instructions for installing onto the device itself.

PC install instructions : https://emanual.robotis.com/docs/en/platform/turtlebot3/quick-start/#pc-setup

Raspberry Pi installation instructions : https://emanual.robotis.com/docs/en/platform/turtlebot3/sbc_setup/

The last part of the RPi install should take about 2 hours to execute, well, I don't actually know as it got to midnight and I decided to just let it carry on & I'd come back and see the progress in the morning.

UPDATE: It's now 3am....status check.  been running for 4hrs it says & still 1/7 complete.  Hmmmm... I was reading one of the previous posts & it did say that I was using an RPi 3 (not a 4), so maybe that is why the compile/build time is a LOT longer - glad it's running over night then?!

It's doing the: colcon build command, so it is compiling & building, so it will get there, so long as the SD-Card is still okay.  Guess I'll find out in about another 4-8 hours time?!!

I left it running for 18+hours?!?!

I have also detected that the battery is dead - which, if I recall happened before?!  turns out from the website, it actually says to keep the charge at 50% otherwise it goes to low and then cannot be revived - so I've ordered a new one for £60, probably turn up in a weeks time, but that is fine as I won't be in a real position to do disconnected moving about until then.

 

Well, that was a BUST.

Now, looking back at the instructions, yep - it was right, however it did not indicate the OOM issue on the SSH connection screen, it was on the real machine monitor screen.

The laptop screen just showed that it had failed - but only after I noticed the OOM above:

I did a quick check to see the version and memory of the RPi - ah, looks like it only has 1GB RAM! and 50% of that was being used.

Okay, time to up the swapfile to 4GB and re-run.

Running the colcon command & it's running faster this time:

and the memory usage seems to be making a difference!

LOOKING GOOD! and it finished.

ah, so my theory looks to have been correct.  It looks like it used about 1.8GB RAM

Right, that took about 35-ish minutes, so back to doing the rest of the steps!

Woo!hoo! okay, got right down to the RViz running on the laptop, connecting to the TurtleBot3 device.

Right, to get that bit to work. several things are needed.

1 = a tab with SSH running on the RPi where I run the bringup command:

2 = a tab with SSH running on the RPi where I can run ros2 topic list & ros2 services list

3 = a tab for the ThinkPad laptop, where I ran the same commands as above - this proves the comms between the ThinkPad laptop and the RPi

4 = Now, on the ThinkPad laptop let's run up the RVIZ software to "see" the TurtleBot3

This shows RVIZ opening up and showing a zoomed out image of the TurtleBot3:

Zooming in, you can clearly see the TurtleBot3 and the sensor output values:

and here we see the physical TurtleBot3 device with lots of flashing lights!

Woo!hoo! now, this is where I want / need to be - I really do need the battery, but I can still do a lot of testing

It does show the difference time makes.  Going through the setup of the ROS setup, I realised just how much I now do know about how all of this works or should work, at least.  But it does make a lot more sense.  I'm glad I "invested" all that time getting the Johnny5 robot working a few months back: 

https://tonyisageek.blogspot.com/2024/12/ugv-lidar-direct-code.html