The Department of Science and Technology (DOST) said yesterday that it has tested a total of 456 Automated Counting Machines (ACMs) to be used by the Commission on Elections (COMELEC) in the May 10, 2004 elections and found out that all have an accuracy rating of 100 percent, except for one machine which has an accuracy rating of 99.998 percent.

The testing was conducted by the Metals Industry Reserach and Development Center (MIRDC) of the DOST.

MIRDC Executive Director Rolando T. Viloria, chairman of the DOST Technical Evaluation Committee, wrote COMELEC Commissioner Resurreccion Z. Borra, in-charge of electoral automation, otherwise known as Phase 2 of the Comprehensive Electoral Modernization Program of the poll body, that "the verification test would show that the machine is 100 percent accurate in reading and counting the votes that are properly shaded, provided that the ballots are fed with the right information. Hence, we safely conclude that the machine has 100 percent accuracy rating on all environmental conditions."

Viloria stressed in his letter that "the security feature of the existing test software presently installed in the machine already addresses his concern which rejects the ballots if fed improperly (inverted) and allows the machine operator to re-feed the ballots (not more than three times) in accordance with the guidelines of the COMELEC."

The number of machines tested as of Oct. 20, 2003 included the nine retested units which were earlier found with small defects, but are now in tiptop condition and ready for use.

The ACMs are part of a total of 1,991 automated counting machines ordered by COMELEC from Mega Pacific eSolutions, Inc., a consortium that includes the South Korean manufacturer of the machines, which won the bidding for the Phase 2 of the poll modernization program.

The total cost of the package is P1.3 billion for the 1,991 automated counting machines.

Executive Director Viloria said in his letter that during the tests, "the shading of some votes (in six precincts) were not done properly, i.e. marks not within the box or barely within the box. If the improperly shaded votes are excluded from the COMELEC count, there will be no errors for the precincts represented by said ballots. This means that the machine can count all properly shaded boxes/votes with 100 percent accuracy."