Election results transmission


Data on the performance of the automated election system employed shows that only about 92 percent of election returns were transmitted during the 2010 elections and only about 73% in the 2013 elections. The AES used in both elections relied on the mobile phone infrastructure of the two telecommunications giants.

The mobile phone infrastructure was augmented with satellite communications through the use of broadband global access network terminals (BGAN) to cover areas where the mobile phone infrastructure did not have coverage.

Transmission failures can occur for various reasons. For one, the modem used could have failed. Weak or absent signal is another reason. The terrain and the location of the voting center are also among the reasons why transmission could fail. Even the positioning of the PCOS machine inside the voting precinct is a factor to consider. In highly urbanized areas, voting centers may be surrounded by tall buildings making it difficult to transmit. Not all was rosy with BGAN terminals. If the voting center is located in a valley, the satellite may actually be below the mountain ridge.

A teacher-friend who served as a member of the Board of Election Inspectors at a school in the Quiapo district narrated her experience, thus: “We were all happy that the counting of votes and the printing of election returns were completed in such a short time. By 9:00 pm, we were all ready to transmit. The problem was there was no signal in our precinct and there was only one area in the school grounds with respectable signal. We shared a modem. We had problems and had to call a technician for assistance. We had to take turns and some of us ended up waiting until about 3:00 am the day after election day.”

This is one area that the Commission on Elections (Comelec) should have already addressed as it prepares for the 2016 elections. But it is grappling with the development of the terms of reference for the transmission service.

Perhaps the Comelec can engage consultants with the requisite telecommunications expertise.

In an email to the Comelec Advisory Council, I suggested tapping the assistance of PETEF whose membership includes, among others, telecommunications experts who can help in the design and implementation of an infrastructure that will aggregate the services of the telecommunications duopoly. They can assist in identifying the necessary equipment and support services which the Comelec can then bid out.

Among the first things that the Comelec can also do is to conduct a nationwide survey of telecommunications coverage. It can provide the two telecommunications giants a list of the voting centers and ask them to confirm signal coverage in the areas where the voting centers are. The Comelec can also ask its Election Officers in the various cities and municipalities to conduct an on-ground survey to determine if there is actual cell phone coverage in the areas where the voting centers are located. The survey should also include the type of building structures (wooden or concrete), terrain (flat or mountainous, etc.), the presence of buildings in the immediate vicinity, and the presence of other types of telecommunications services that can be used.

The survey suggestion is nothing new. Part of the preparations for Namfrel’s quick count operations is conducting an on the ground survey of all means of communication available nationwide—including some remote areas. The survey was done by Namfrel volunteers.

The survey results may be used for planning purposes. For one, it will enable the Comelec to identify areas where there is no cell phone coverage and from there plan for alternatives.

Talking of alternative technologies for areas outside the covered areas by the two telecommunications giants, satellite communications may be used.

Portable cell sites can be deployed and integrated with the duopoly’s cell phone infrastructure. Portable cell sites can actually be assembled locally using parts available in electronic shops. It can be operated using open source software.

Presently, the Department of Science and Technology and the Information and Communications Technology Office are testing an infrastructure called TV Whitespace in Bohol. The test is yielding promising results and deployment is now being planned for the provinces of Bulacan and Cavite.

IP radios also provide another alternative.

Even the short message service, a functionality available in mobile phones, can be used to transmit election results. Of course, this will require the use of smart phones and it has to be properly programmed to carry the election results in a secure manner.

Let’s face IT. A whole range of technology options for election results transmission is available. What needs to be done is (1) conduct a detailed study of the requirements, (2) plan and design what the election results transmission infrastructure will look like, and (3) identify and implement the appropriate technologies to be used.


Please follow our commenting guidelines.


  1. Renato B. Garcia on

    The past 15 years have created a nationwide telecom coverage down to the polling centers. Transmission today is of good quality and cheap. The cost of transporting Election Returns from polling centers to municipalities is not only high, but full of security problems and risks. Ballot boxes and ERs get lost or stolen during the transport, despite being accompanied by tens or even a hundred of poll watchers for whom transport and meal allowances have to be provided as well.
    Alternative electronic transmission schemes to the cheaper and reliable telco mobile data service are needed only in the remote islands and localities.
    REPLY: but, sir, even if the ballot baoxes have been stolen, THE TRUE ELECTION results IN THE PRECINCT are already known by ALL–the Board of Election Inspectors who sign the certificate, the authorized poll watchdogs who received copies of the results, the rival political parties whose representatives have also witnessed the counting ballot by ballot. But now, sir, the counting, or NOT COUNTING and WRONG COUNTING, happens secretly without anyone to see INSIDE THE PCOS MACHINES.

    • Renato B. Garcia on

      In manual counting, only the BEI Chairman can see the ballot he reads and picks up from his stack. At times some do use the same ballot. With a hundred people in the small classroom, and the long several sheets of TARA for more than 300 candidates incl senators, party list, congressmen and local officials, it is a mess. Even a CCTV camera pointing to the ballot being read and the several Taras around the room will not make it transparent.
      In OMR, the voters’ paper ballots are scanned to provide a backup and stored in a 2nd cfcard, copied later in CDs at COMELEC central whse. W DRE, paper ballots are printed and verified by the voter and placed in small ballot boxes. All computer programs are audited and certified by an international certification entity (ICE), one among 4 accredited by the U.S. Electoral Accreditation Commission.
      But here again, we are talking of ER transmission, not automation vs manual elections.

  2. Virgilio V. Hernandez on

    The recommendations of Lito Averia are critical and must be immediately applied by COMELEC. We’re these recommendations with the Commisioners already?

  3. jesus nazario on

    Why are we insisting on a solution architecture that is “totally” dependent on 100% good quality datacomm link presence with the requisite reliability when the country still has deficient datacomm penetration of problematic quality and reliability ?

    Why don’t we just use something like (or an improved/expanded version thereof) the circa year 2001 solution architecture which was planned to use just 1,991 satellite discs planted in all cities and municipalities. These so-called city/municipal central counting sites are where voted batched OMR-readable ballots are brought physically after polls have closed and counted via CCOS machines (central count optical mark reading machines in contrast with PCOS) and from which points ER transmissions will be mad up the consolidation hierarchy (provincial and central).

    This solution in year 2001 was planned to use only some 2,000 high speed CCOS machines (150 ballots per minute counting speed per unit which is equiv to the 2,000 units being able to count some 18,000,000 ballots per hour nationwide) compared to 2010’s 76,000+ or 2013’s 78,000+ PCOS units or or the planned 2016’s 120,000+ PCOS units or some 40 to 60 times less counting machines (and correspondingly less modems and datacomm links) and therefore also 40 to 60 times less number ER transmission points. A subset of this solution was successfully pilot tested during the 2008 ARMM elections in 4 out of the 5 ARMM provinces and bested hands-down the DRE which was piloted in just one province. The year 2001 solution was 6 or more times cheaper than the horribly expensive PCOS solution. With the growth of voters from some 36M in 2001 to some 50M in 2016, the same type of solution will need some 2,800 CCOS units and the same number of satellite discs/modems or just about PHP 2.8 billion.

  4. Renato B. Garcia on

    The satellite service has to be separately bidded out from the local telecom service. The cost of the HW for 5,500 BGANs, at $1,100@, is half of the ABC for the ITB on Election Results Transmission. Portable satellite, BGANs, were used in 2010 n 2013 in less than 25% of the polling precincts were there was no GPRS mobile cellular coverage.
    I did submit my comments on 5 major issues on the specs and requirements during the Prebid Conference last February 2015 on the ITB for Election Results Transmission.

    Past President, PETEF
    Consultant COMELEC EnBanc for AES 2010
    COMELEC AES2010 PMO Adviser
    Lead, COMELEC AES 2010 Transmission TWG