It might come as expected to a lot of us by now that the Comelec’s automated election system (AES) encounters one problem after another. Recently, news came out that the markings on the ballots for the May polls are unreadable by the Precinct Count Optical Scan (PCOS) machines that will be used in the upcoming national and local elections. According to the Comelec itself, errors in the ultraviolet (UV) markings printed on the ballots are due to the high-speed printing process used by Smartmatic.

Just like how rushed the Comelec handled the whole automation process, the rush to print ballots in time for the May 10 elections made the markings overlap with other printed characters forcing the Comelec to disable the PCOS machines’ UV reading capability that is supposed to automatically verify the authenticity of the ballots inserted in the machine.

Ultraviolet markings are often used in paper money and high-security documents like passports to make it difficult to generate a fake or counterfeit document. Cashiers would usually use a UV lamp that will make light blue or violet markings visible in the document. These lamps, sometimes called blacklights, are needed to make the markings glow. Without it, a casual observer would see only the ordinary printed characters in the document. A set of UV lamps (together with backups) should be included in the election paraphernalia.

Together with the barcodes and other security features in the ballot, the UV markings would have given a tight control on the authenticity of the ballot itself. Take that assurance away and we add another reason that erodes our trust in the whole May 2010 exercise itself.

To make matters worse for an already overloaded set of tasks for the board of election inspectors (BEI), the Comelec said that the members of the BEI would just use hand-held UV lamps to make sure that the ballots are genuine. This will be done just before the voter inserts the ballot to the PCOS machine. Adding another step in an already tight schedule would increase the possibility of a queue buildup. Instead of making things faster, the Comelec’s adjustment makes the process bottleneck at the PCOS machine tighter.

How will Smartmatic and Comelec disable the UV function? Will it recompile the source code? Or will it enable some undocumented configuration file or option in the PCOS machine? If so, how will this affect the other functions of the PCOS machine? Is it only the UV reading function that will be deactivated?
Have Smartmatic and Comelec verified and tested the whole scanning and counting program in the PCOS if one disables the UV function? Without the source code open for verification, the general public will just have to cross fingers that there are no other routines that will not work when this reconfiguration is done.

How will this “deactivation” of the UV function be carried out? Via the compact flash card? Via the keypad? Remotely? This UV fiasco teaches us that there other unknown and undocumented options in the PCOS machine that can be activated or not by a specialist. We just hope that he or she is truly a Comelec authorized personnel by the Comelec and not some saboteur.

With their every misstep in their rush to automate the elections, it is the Comelec and Smartmatic themselves that give us reason to doubt the veracity of any election result from the AES. We have not even tackled the procedural problems such as the new rules on pre-proclamation protests or the problems in the canvassing side. There are other vulnerabilities in the AES such as transmission issues and security features that we will tackle in future columns.

The vulnerability (or lack thereof) of any critical machine to do its job correctly is a function of how well its interlocking subsystems interact correctly. These subsystems can fail if not properly tested and any possible failure is not planned for. The robustness of the whole system is not something that an engineer wants to prove during the actual operation of the machine itself. One would have even forgiven the Comelec and Smartmatic if this were just a machine of inconsequential repercussions. However, the whole AES is not just a matter of getting a simple counting machine to work.

What will be tested on May 2010 are not just issues of misplaced ultraviolet markings or configuration problems but also the misplaced faith on untested machines by an institution that suffers credibility problems. More importantly, what will be tested on May 2010 is not just the Comelec and Smartmatic’s automated election system but also our readiness as a people to assert our democratic will if all systems fail.