When deploying a custom brain, the new brain shows in the menu but throws a "brain not found" error after bot reset.
Current workaround: tap power button or use sudo reboot 0.

Use a tmux info file instead:

https://forums.fast.ai/t/local-linux-setup-with-tmux-p-and-anaconda/27566

Repro, this fails:

% python -c "import flask"

Reason: dependency on jinja2 removed
https://stackoverflow.com/questions/71718167/importerror-cannot-import-name-escape-from-jinja2

Currently triaging.

Whenever the ball falls off the plate and the ball is no longer detected, the calculation for the velocity becomes inaccurate and needs to be reset.

Great suggestion from @mikeestee

1. Train an ONYX/TF model to turn "ball in video frame" into a "normalized x/y" coord.
2. Update the simulator to render the ball as a video frame, instead of synthesizing it directly from the physics.
3. Import that and then combine it with one or more concepts for controlling the servos
4. Export to moab.
5. ???
6. Profit

We use openCV computer vision algorithms to find the dominate hue of the center pixel space during calibration. Then use that color for the Hough circle detection algorithm. It does mean you have to recalibrate for different color balls, and the detection is sensitive to changing ambient lights (dark room vs sunny).

Would require the synthesis of simulated data and/or lots of real frames of data. Since we have labeled data now from ideal conditions (i.e., we can easily save videos with meta-data of circles where the ball is), we can get a good start.

Due to humask using a char for the hue argument (center), we can only have hues up to 255.

Found by @journeyman-msft.

Running menu.py -h causes crash

I can see using journalctl -f an error thrown when calling oled 0 0 for permissions (needs sudo).

Some ideas on making it easier to upgrade Moabian:

• Automatic firmware version detect and flash
• flash will default to latest firmware (and remove old firmware from main)
• Faster setup of MOABIAN variable in /etc/environment doesn't require system or firmware update.

From docker/compose#9805

"Docker Compose v1 has been EOL'd".

Solution is to curl v2 executable directly.

Capture is 256x256 which is pretty small. Let's double it to 512x512 in the view (not the capture).

On Version 3.1.1, when testing Kiosk mode, I noticed the built in Brain is no longer in the menu. Even when I scroll up to refresh the menu the brain does not show up in the menu.

While the robots three arms are calibrated during assembly, it's possible that the servos drift as the tolerance is a rough 0.5º. When the plate is in manual mode, it should be level to your desk. If not, we are building a small command-line tool to find the right +/- degree offsets to fine tune the bot.

The raw camera view plus the annotation showing how the ball is being tracked can be streamed to a web browser.

The plate level offsets as shown in the level_plate.py do not match the order of the offsets shown on the menu display under HUE INFO.

Add a text file to add serial number during setup. we can use for iot and other reporting

Thought this was done already. Minimizes ephemeral log writes to a medium that's not really meant for it (SD cards):

https://forums.raspberrypi.com/viewtopic.php?t=226425

Connect, configure, train and deploy brain docs.

There is support in the existing Moab firmware to display IP address and FW version. This issue is to send that information to the bot.

This makes it easier for headless Moab units to show their DHCP-assigned IP address when connected via cat-5 / ethernet.

After a ball falls off the plate (terminal condition), then if you pick it up and put it back on the plate, then we should reset the echo state memory.

If you don't have an ethernet cable and you don't have an HDMI monitor + keyboard, there's no way to configure Moab (easily) to put it on your wifi network.

An idea came up where one could "show" Moab's camera a QR code with wifi details.

https://www.raspberrypi.com/news/raspberry-pi-bullseye-update-april-2022/#comments

Python to Hat communication over the SPI bus is unnecessarily complex due to the need for root access.

Report from Muntazir Mehdi reporting:

ERROR: The Compose file './docker-compose.yml' is invalid because:

Invalid top-level property "iversion". Valid top-level sections for this Compose file are: version, services, networks, volumes, secrets, configs, and extensions starting with "x-".


You might be seeing this error because you're using the wrong Compose file version. Either specify a supported version (e.g "2.2" or "3.3") and place your service definitions under the `services` key, or omit the `version` key and place your service definitions at the root of the file to use version 1.

For more on the Compose file format versions, see https://docs.docker.com/compose/compose-file/

To track this down, I'll be asking for a screenshot of the following commands:

splash
docker-compose --version
docker --version
head -5 ~/moab/docker-compose.yml

My moab output:

Suggestion: put all CLI commands under bot and remove the down confusion. Does it take docker down? Or the menu as well?