Hi folks,

as someone who has never had the opportunity to participate in a calculus of probability course I don't have the faintest clue of how covariance works. But it seems that I need to in order to get navigation, SLAM and actual driving to run optimally.

Assuming a differential robot with wheel encoders (decoupled from the driveshafts to avoid slippage) and a laser scanner, why, where, how... covariance... and... well, actually... I basically don't even know where to start asking.

Is there anybody out there who can explain covariance to me like I'm five? I'd be deeply grateful and I apologize for a not that well couched question.

Cheers, Hendrik