DB<130> $a='ABCBBBCBADBCCACDDDAC'
 => "ABCBBBCBADBCCACDDDAC"

  DB<131> grep { /(.)\1\1/ } unpack '(A5)4', $a 
 => "DDDAC"
[download]

edit

unpack pattern shortened

update

IMHO its also possible to include the same principle into one regex (matching groups of 5 characters and investigating each match with embedded Perl) but I doubt it's more readable.