Grabbing a huge bucket-full of guesses, and using inline files for this test case.

#!/usr/bin/perl

# http://perlmonks.org/?node_id=1207825

use strict;
use warnings;

open my $file1, '<', \<<END;  # replace with your real open
10000000 10000000 10000000 10000000 10000000
00000011 00000011 00000011 00000011 00000011
10000001 10000001 10000001 10000001 10000001
11000000 11000000 11000000 11000000 11000000
01100000 01100000 01100000 01100000 01100000
END

open my $file2, '<', \<<END;  # replace with your real open

CHAIN_TEST =

    pattern = 0;
    apply "grp1_load" 0 =
        chain "chain1" = "010";
    end;
    force   "PI" "XX0XXXXXX" 1;
    measure "PO" "XX" 2;
    apply "grp1_unload" 3 =
        chain "chain1" = "010";
    end;

    pattern = 1;
    apply "grp1_load" 0 =
        chain "chain1" = "101";
    end;
    force   "PI" "XX0XXXXXX" 1;
    measure "PO" "XX" 2;
    apply "grp1_unload" 3 =
        chain "chain1" = "101";
    end;

end;

SCAN_TEST =

    pattern = 0;
    apply "grp1_load" 0 =
        chain "chain1" = "010";
    end;
    force   "PI" "100000000" 1;
    measure "PO" "10" 2;
    pulse "/CK" 3;
    apply "grp1_unload" 4 =
        chain "chain1" = "010";
    end;

    pattern = 1;
    apply "grp1_load" 0 =
        chain "chain1" = "110";
    end;
    force   "PI" "110000101" 1;
    measure "PO" "11" 2;
    pulse "/CK" 3;
    apply "grp1_unload" 4 =
        chain "chain1" = "100";
    end;

    pattern = 2;
    apply "grp1_load" 0 =
        chain "chain1" = "001";
    end;
    force   "PI" "010000110" 1;
    measure "PO" "10" 2;
    pulse "/CK" 3;
    apply "grp1_unload" 4 =
        chain "chain1" = "010";
    end;
END

my $contents2 = do { local $/; <$file2> };

while( <$file1> )
  {
  my @replacements = split;
  print $contents2 =~ s/"PI" "\K\w+/ shift @replacements /ger;
  }
[download]

Does the output match your desired output (which you didn't show us) ?