Automation ‘only 25%’ chance of success (3)

Here is the last part of IT expert Roberto Verzola’s rating of the automated election system (AES), based on Smartmatic’s own claims in full-page ads on Mar. 8, in five subsystems: (1) Hardware; (2) Software; (3) Logistics, support, preparations; (4) Telecommunications and transmission; and (5) Ballot printing, infrastructure. It was published in Jarius Bondoc’s column, “Gotcha”  (April 9, 2010 edition of the Philippine Star, Opinion Section)

Subsystem 3: Logistics, support, preparations

Claim: “36,000 voting centers surveyed [for] network signal, power.”

Given 48,000 voting centers, that’s 75 percent covered as of Mar. 8.

Claim: “904 testing … employees working two shifts.”

Even three shifts are not enough, knowing that 1,990 machines took Comelec three months to test in 2008. Smartmatic should have tested the machines in China, before shipping to RP. What we want is due diligence testing by Comelec. It is our elections; our money will pay for the machines.

Claim: “Recruitment, training of 48,000 field technicians started.”

Since the ad came out with only 60 days to Election Day, we can only gape in disbelief: “Started”?

Claim: “438 Comelec training personnel certified.”

Yet they are to train in 60 days at least 230,000 elections officials.

Claim: “Contracts with logistics providers, forwarders signed.”

From news reports, Smartmatic hired three forwarders: Germalin Enterprises (P2.3m 2006 net income), Agro Intl Forwarders (P3.7m 2008 retained earnings), ACF Logistics Worldwide (P1.1m 2008 cash balance). Given their Herculean task, their financial capabilities do not inspire much confidence.

Comelec needs to publicize results of Smartmatic’s survey for signal and power. “Problems” in power availability or signal transmission can be used to selectively affect voter turnout by regions or provinces, in a way that can bias the election outcome.

Subsystem-3 probability of success: 80 percent.

Subsystem 4: Telecommunications and transmission

Claim: “48,000 modems for transmission manufactured, delivered.”

Again the missing word here is “tested”. If the China-made modems encountered transmission kinks right in Metro Manila, something could be wrong with quality. If Smartmatic delivers a mix of good and bad units, these can be selectively assigned by region or province to ruin transmission where election cheats want to strike.

Claim: “46,000 SIM cards secured.”

Only 46,000 SIM cards for 48,000 modems?

Claim: “5,500 BGAN transmitters purchased; 680 VSAT transmitters leased.”

With the 48,000 modems, these add up to 54,180 transmitting equipment, enough for 71.8 percent of the PCOS. A serious problem haunts transmission security: instead of separate units, Smartmatic controls the entire system of passwords and digital signatures — from generation to certification. In a business setting, this is like merging in one person the duties of vendor, operator, accountant, cashier and auditor — an open invitation to fraud.

Claim: “Contract with major telcos secured.”

Yet Smartmatic’s own survey says telcos can cover at most 70 percent of precincts.

Subsystem-4 probability of success: 70 percent.

Subsystem 5: Ballot printing, infrastructure

Claim: “Over 10 million ballots with invisible ultraviolet mark and unique barcode printed.”

Printing of ballots started Feb. 8. Comelec internal memo leaked to media: as of Mar. 1 (20 days after Feb. 8), 7.9 million ballots had been printed. The arithmetic: 7.9 million ballots in 20 days is 394,000 ballots per day. At this rate, in 60 more days, Mar. 2 to Apr. 30, some 21.3 million more ballots can be printed, for a total of 39.2 million, not quite the 50 million needed for a 1:1 ballot-to-voter ratio.

This is not a single print job, but some 1,600 jobs, since each city and municipality has its set of candidates. Comelec expects to do 20 different jobs a day over 80 days: practically one job every hour. Even scheduling of print jobs, if biased, matters: ballots from later schedules run greater risk of late delivery.

Subsystem-5 probability of success: 80 percent.

In sum, our assessment of Smartmatic’s five AES subsystems: (1) Hardware, 80%; (2) Software, 70%; (3) Logistics, 80%; (4) Telecoms, transmission, 70%; (5) Ballot printing, 80%.

A fundamental principle in project management: to get the overall probability of success of a project that relies on a series of sub-projects, each of which is essential to the project, the sub-projects’ probabilities of success must be multiplied together. Essentially the same principle is followed in product design and reliability engineering.

Thus a system with five subsystems, each with 99% probability of success, will have overall success odds of .99 x .99 x .99 x .99 x .99, or .95 (i.e., 95%). If each of the five subsystems has 95% probability of success, the overall success chance is 77% (try it on a calculator). Five subsystems with success probability of 90% each will have an overall success luck of 59%.

Given its gaps, we are not ready to assign optimistic probabilities to the AES. Those willing to give Smartmatic the benefit of the doubt, and so assign 80% success rate to each of the subprojects, will yield a 33% overall chance of success — three-to-one odds of AES failure.

Our evaluation of AES subprojects: .8 x .7 x .8 x .7 x .8 — or 25% overall probability of success. Conclusion: risk of AES failure on May 10 is unacceptably high, especially for an election at this crucial juncture of political history, when failure should not be an option.

(Go to Part 1 and Part 2.)

No comments yet.