Last night we tried to install Google Coral on our Limelight. The Coral was able to do it’s default detection OK, so we moved to install the pre-trained neural net pathways on our Limelight. This resulted in the LL getting into a ‘hard running’ state (fan running at full speed) where it flashed its LEDs once about every 10 sec and seemed to be unresponsive. After about 10 minutes, I powered off, unplugged the Coral and tried booting with just the LL, but it was still in that state. I opted to re-flash the LL to make a second attempt later. Any advice on what we should do differently? It seemed pretty straight forward, so not sure if we should have just been more patient with it. The directions don’t give any indication of what to expect when configuring this, or any troubleshooting. Thanks!
We’re currently running into the same problem? Have you been able to fix it? We created a model using teachablemachine.withgoogle.com and when we upload the TFLITE file, the pipeline crashes. Did reflashing your limelight work for you?
I have to ask the question…LL3 and coral…are they really needed this year? We have them…just need the justification of putting both on the bot.
Automatic intake alignment to notes on the other side of the field if picking up off the ground.
Are you using something else for the shooter?
April Tag detection on a separate Limelight for the shooter, yes.
We only have a LL3 and a webcam…no Pi. just using the webcam for lining up the intake.
I had the same issue! I uploaded a different file, and it worked. I’d be happy to share the link to the one I used!
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.