Oh, yes, some Mars probe. Had the height of Mons Olympus coded in miles rather than metres. If you believe that ...
There was also the Venus one that's supposed to have failed because of a "." rather than a "," in a FORTRAN DO statement, but I'm even less convinced that I believe that one.
(DO 10 I = 1,10 and DO 10 I = 1.10 are both perfectly valid FORTRAN statements - the language preceded modern conventions on things like lexical analysis and keyword usage - but they behave rather differently.)
