I’ve spent the last week adding two smart cameras to my robot which currently provide 6 regular camera images and produce two point clouds (the grey structures in the image below). When I first got them to work and turned on the depth computation, I hadn’t charged the battery on the robot in some time and the additional power sucked in for the depth computation caused the battery to die and the whole robot turned off.
The pair of cameras draw an additional 60 watts of power right now and I haven’t even turned on the artificial intelligence computation yet. Well, I don’t think I have. The cameras provide a lot of capability and I’ve had to rewrite the manufacturer’s sample code to get the ability to run two cameras at once on the same robot, and I’m not sure yet of what features I’ve enabled besides depth calculation and stereo vision.
The text at the very bottom of the picture shows that when I run the base robot code and enable the visualizer (rviz2) software, 3/4 of the computer’s capability is being consumed. Enough that it makes it hard to do anything else on the computer, like compose this message while the robot software is running.
Progress is good.
