Can't install ti-edgeai or ti-vision packages

General Discussion
1

I loaded “BBAI64 Debian 12.7 2024-09-04 XFCE Flasher” and am following along in the “Getting started” at Quick Start Guide — BeagleBoard Documentation. The step titled “Update ti-edge-ai 8.2 examples” fails with:

E:  Unable to locate package ti-edgeai-8.2-base
E:  Couldn't find any package by glob 'ti-edgeai-8.2-base'
E:  Unable to locate package ti-vision-apps-8.2
E:  Couldn't find any package by glob 'ti-vision-apps-8.2'
E:  Unable to locate package ti-vision-apps-eaik-firmware-8.2
E:  Couldn't find any package by glob 'ti-vision-apps-eaik-firmware-8.2'

My /opt directory only contains:

bb-code-server
node-red
source
ti-zephyr-firmware
u-boot

I checked /etc/apt/sources.list and only debian.org repos are defined. What are the TI repos? Why are they not in the image to begin with or why does the “Getting started” instructions not include steps to add them?

FYI, even though this is supposed to be the “XFCE” image, it boots to a console. I had to manually install XFCE to get a desktop.

I tried this with both BBAI64 Debian 12.7 and the older 12.6 but got the same result.

2

The pre-built TI EDGEAI image is available here: ARM64 - Debian 11.x (Bullseye) - Monthly Snapshots - 2023-10-07

All “Debian 12.x” images are setup to be used with bb-imager, or you need to edit ‘sysconf.txt’ in the boot partition to setup your initial user.

If you want TI’s latest EDGE AI grab this build: BBAI64 now can use TI SDK10.0 and debug R5

Regards,

1 Like
3

Is it possible to host working SD card image with Debian 12 and TI SDK10.0 somewhere to try out?
From the BBAI64 now can use TI SDK10.0 and debug R5 post I’m having a hard time figuring out how and where I should do all these steps.

4

Me too! And the funny thing is, I wote the original documentation for this board while I was at beagle.org! Yes I’m still alive and I have a real nice project requiring video camera image tracking!

2 Likes
5

Thank you very much!

We looked at using the BBAI64 for a new open source wildlife camera but we read about some issues taking the BBAI64 into and out of suspend-to-RAM (necessary to conserve energy because of the very limited solar energy available on forest floors) sowe’re looking at using the BeagleY-AI.

Will