Tuesday, May 14, 2013

On PCOS and Automated Elections

I am no expert on the PCOS (Precinct Count Optical Scan) machine and Automated Elections in general. But I know how to distinguish a technical issue on hardware and a knowledge issue on the part of the user, as well as a process issue with the overall project.

The mid-term elections we just had in the Philippines is supposed to be a continuing process improvement in the part of automated elections. But it seems certain groups and people want progress to be aligned with their own definition of the word along with the benefits and twist it in mainstream media.

In the case of the PCOS, its function is very simple by itself, but expecting all 77,000 units deployed on field to work perfectly is impossible. Like any piece of electronic equipment that comes out of the factory, there is always a failure threshold. So I wouldn't be surprised if 100-200 PCOS would fail on field as long as their failures are heterogenous. I am sure most of the failures are process related or a  component blew off. If the failures are the one and the same, then we could safely assume there is a product defect.

A potential failure can also mean the operators didn't understand well how to troubleshoot the device. If so, this is a problem of training and not the hardware itself. If so, this is COMELEC and the technology partner's fault.

Then there is also the issue of delayed transmissions of results to the central servers of the COMELEC. If the PCOS had a faulty modem, then its the hardware's fault. If the connection signal is weak (coming from a telco), then why would it be the PCOS' fault? I know for a fact that the telcos have worked with COMELEC for months in preparation, lots of people were allocated to the project to ensure that things work out smoothly. Unfortunately, things happen, but it doesn't mean the project managers involved have no idea what to do next.

And lastly the process. I read and saw a lot of people and parties complained of long lines. I myself waited for half an hour under the blazing sun to vote in my hometown. But when I got my turn to vote, I saw the PCOS functioning properly - the problem was with the precinct venue (it was a single classroom for 4 clusters) and the time it took for voters to finish their ballots. Feeding the ballot to the PCOS takes less than 5 seconds. I finished shading my ballot in less than 5 minutes because I already have my list (and it also helped that I didn't even try to complete the number of needed votes). I noticed the other voters still thinking who to vote.

I am not defending the PCOS or the current process of elections. There sure are a lot of things to be improved. One thing for sure we shouldn't use the same PCOS again for the 2016 elections. But if have to, then we have to optimize the process - not just on the use of the PCOS, canvassing, etc., but on the actual voting process.