The first ever automated national elections with more than 50.7 million registered voters is still ongoing as of writing time since the transmission of the data to the national canvassing server is slow mainly due to problems of transmission in several areas of the country. As clear winners emerge in both national and local polls, it seems that although Comelec claims that they have successfully conducted the May 2010 elections, they cannot clearly say that the automated election system (AES) was a resounding success.

All of the vulnerabilities of the AES that we have earlier mentioned about have cropped up during election day. A week before the elections, we saw the SysTest Labs source code review and found that they encountered many problems with regard to the Precinct Count Optical Scan (PCOS) machine and the municpal canvassing servers. SysTest cautioned Comelec to address these problems with procedural checks but we have not seen these checks in place during election day. The source code review itself also warned about the non-certification of the national and backup servers that is now being used in the national canvassing at the PICC and Congress.

We were faced Tuesday before the elections with the problems with regard to the misconfigura-tion of the compact flash cards. Comelec and Smartmatic had to go through a logistical nightmare of recon-figuring and delivering the CF cards for the machines within a few days of the opening of the polls. It thus comes as no surprise that the main pre-election problems that were reported in our monitoring activities in KontraDaya prior to the opening of the polls would be related to the Final Testing and Sealing (FTS) of the PCOS machine. These reports continued even up to a election day with some precincts still noting delayed FTS activities at 7:00 am.

As voters from more than 329,000 precincts trooped to the 76,000 clustered precincts (CP) during election day, this delay in the opening of the polls resulted to long lines early in the morning. Voters faced their initial frustration in the difficulty of finding their name from the voter’s list, figuring out their precinct and getting their sequence number. Some people reported being unable to find their names even if they actually registered last year. Most of those who were affected were the youth who beat the deadline in the extended registration last December.

Surprisingly, despite the field test data that they have, our simulations and a host of calculations from other concerned citizens, the Comelec, the PPCRV and the BEI instituted (albeit not uniformly) a ticket system that introduced an extra bottleneck that worsened the queue. In some instances, only around ten people at one time were allowed to come in the voting area. The Comelec was forced to arrest the problem at midday by ordering the BEI’s to simplify the procedure and introduce parallel queues for voters identification. They also extended the voting hours to 7:00 pm.

The remaining bottlenecks were the BEI chairperson (now with a simplified task, to which some BEI chairs took to the extreme) and the PCOS machine. Things would have been manageable at this point had not a host of PCOS machines failed. Some machines had exploding batteries. Some machines did not boot at all and had to be replaced. A lot of frustration comes from these failures which aggravated the fact that many voters stayed on the line despite the heat and long queues. A major adjustment that removed the queue in some places was when the BEI skipped the PCOS machine altogether and asked voters to simply leave their ballots in the ballot box with the assurance that the BEI would be the one inserting these to a working PCOS machine at the end of the day.

Come evening, the reports shifted to talking about failures in transmission of the results. There were precincts that have no modem at all. There were others that cannot connect to the system. Some precincts were forced to go manual or to bring the compact flash cards to the next working PCOS machine at the nearest clustered precinct.

The Comelec was able to coax the automated election system to work with hodge-podge and ad hoc solutions on election day itself. It simply did not work as originally planned. It is only due to the fact that tally shows a clear lead for the national positions that lessened the doubt on the veracity of the automated election system. This will not always be the case. Who should answer for this fiasco and what do we do about our flawed AES system?