[r/robotics] How would you approach joint multi-camera + robot calibration when CharUco gives cross-cam disagreement?

Part of

Iterative Machine Learning And Optimization