The More the Merrier?

Hello from team 8816! I’m a Cocaptain on the team and I am currently working on developing our vision setup. So our current vision setup is a limelight 3 which I plan on using for april tags due to the fact that its extremely easy to set up and additionally we have an raspberry pi which I plan on using for custom object detection which we have members working on building right now. Now my question is that i’ve seen teams using up to 3 co processors before and I was wondering what is the point of using more? Is it just more optimal to use multiple coprocessors since you can consistantly get data from both rather than having to swap between pipelines for one co processor or is there another reason why?


The reason for more coprocessors is mainly to get more computing power. For example, an orange pi can run 2 cameras really well, but if you wanted your robot to have 4 cameras, you would want to add a second orange pi to handle the extra camera streams.

Echoing what @leo-les said. It’s also to get better location data. The more often your robot can see April tags the better the estimate of the robot’s current location will be.

Probably a slightly dumber response as im only just learning this branch of programming, but from my mechanically trained brain, I think of it as “if one breaks, then at least you have a backup.”


Ok I thinkt he computing power makes more sense to me. I think I’ll end up setting up three co processors just to have multiple facets of computer vision so I constantly get data from different coprocessors