In the recent ARMM election, the COMELEC tested two different technologies to determine which will be suitable to be used in the upcoming national elections in 2010. According to the COMELEC reports, about 50% to 60% of the registered voters trooped to the 1,901 precincts to participate. The electronic voting system has consolidated the results two days after the election which was a "smashing success" according to Senator Richard Gordon.

It is known in the computing industry that an electronic voting system is difficult to develop. There are many challenges in the process: the translation into computer countable data the votes in an election. There is also the need for verification, tallying and consolidation of the counts that are produced. The US, for example has had a number of horrible experiences of fraud due to flawed electronic voting machines and systems.

In the recent ARMM elections, direct recording electronic (DRE) machines and optical mark reader (OMR) technologies were deployed. More than 3,000 DREs were deployed in Maguindanao. Each DRE has a touchpad with the pictures of the candidates and voters will have to press the picture of the candidate to register a vote. After selecting all candidates for different positions, a print-out will be generated. The print out should match the selections on the screen. DRE is simple to use but is more expensive compared to the OMR. According to some critics, P20B of government funds will be needed for the 2010 elections if the DREs are to used.

In Shariff Kabunsuan, Lanao del Sur, Basilan, Sulu and Tawi-Tawi, OMRs devices were used. Voters using this technology will be shading the oval shape next to the name of the choice. This is similar to the national entrance tests that many of our high school students undergo each year. Other groups are also proposing web-based solutions.

An electronic voting system has to be accurate. It should accurately translate voter's intent to the appropriately counted votes. It should also be secure so that it is impossible to change someone else's vote once cast. It should be able to protect the voter's anonymity. Proofs that can be used to sell votes should not exist. On the DRE machine, the generated report is placed into a ballot box (which can be stolen). This will also serve as a basis for a recount if ever the results of the machines are questioned.

An electronic voting system should also be scalable. In the last national elections, we had 45M registered voters. If we are to automate national elections, the deployment of such systems has to be customized per congressional district, province, and town. The ARMM test is just the tip of the iceberg of what we might encounter in the next national elections. Both the OMR and DRE technologies should be tested for all possible voter behaviors.

The software that runs in the electronic voting machines will always have bugs. An honest to goodness code audit and review should be done in a manner that is as transparent as possible. The companies can open the source code of the machines to allow developers to review and comment on the code.

I attended a public demonstration of the DREs and OMRs for the ARMM elections last month. After a few questions on handling errors of both machines, I asked about the security of the data transmission and consolidation of results. What encryption is used in storing data? Is it encrypted during transmissions? At first, I got the impression that the vendor does not know the product to be used. Then later, he confessed that for security purposes, my questions cannot be answered.

Security by obscurity is a false sense of security. Banks and credit card companies follow standards-- well-known standards-- to secure their data. They even have certifying bodies for these. Unfortunately, the security of electronic voting machines rests on secrecy. Secrecy cannot point out flaws in the security mechanism until it is too late. Automation will indeed make it a bit difficult for outsiders to change election results. But it also becomes easier to crack if security mechanisms are not up to the standards of critical systems.

These technologies will also only speed up the counting of votes that were cast, regardless of how a citizen is able to arrive at that decision. The power of guns, goons and gold (the old 3G) that has historically determined the winning candidates in our local and national elections in our country will still be there. Will the advent of automated counting machines only serve to make the votes delivered by this old 3G network counted faster?