The Republic of Kenya set out in 2008 to make significant reforms to its electoral processes. In 2009 the government created the Interim Independent Electoral Commission (IIEC to develop and implement an entirely new process for voter registration and management of elections. The IIEC worked with system integrator CODE to implement a multi-biometric voter registration system based on secure and reliable fingerprint, facial and multi-biometric technologies from Neurotechnology.
The Republic of Kenya set out in 2008 to make significant reforms to its electoral processes. In 2009 the government created the Interim Independent Electoral Commission (IIEC to develop and implement an entirely new process for voter registration and management of elections. The IIEC worked with system integrator CODE to implement a multi-biometric voter registration system based on secure and reliable fingerprint, facial and multi-biometric technologies from Neurotechnology.
Civil unrest in the aftermath of Kenya's 2007 elections prompted the government of Kenya to completely overhaul its system of managing elections. The IIEC was formed to replace the previous Electoral Commission of Kenya and charged with the mission to institutionalize sustainable electoral processes that would guarantee fair elections. The mandate of the IIEC covered all aspects of implementing elections including reform of the electoral process; conducting a fresh registration of all Kenyan voters to create a completely new voter register; developing a modern system for collection, collation, transmission and tallying of electoral data and promoting voter education.
To undertake such a large and important mandate, the IIEC required a voter registration system that would ensure highly accurate identification of Kenya's 13 million voters and a reliable process for flagging any duplicate registrations in the system. They turned to system integrator CODE who developed and implemented a multi-biometric fingerprint and facial identification voter registration system based on Neurotechnology's MegaMatcher biometric technology.
Because Kenya's population is so widespread and diverse, ranging from bustling urban centers like Nairobi to remote rural communities and semi-nomadic groups such as the Maasai, implementing an entirely new voter registration effort required flexible software and sturdy hardware that could be used in the field to collect data and issue identification cards even in areas that are off the power grid.
Also, because Kenya's citizens speak a variety of languages and not all voters can read and write, the IIEC wanted a solution that could provide unambiguous identification to ensure accuracy of the registry and provide a fast, reliable way to check data for any duplicate registrations. For the initial phases of the project, both fingerprint (thumbprints) and facial biometrics were chosen for voter identification, and future project plans call for the use of iris scanning. Using multiple biometrics helps to ensure a higher degree of identification accuracy.
CODE recognized that Kenya's solution would require a very high degree of biometric accuracy and reliability across a spectrum of variables – from fingerprint quality to wide variations in lighting conditions between urban settings and field registration sites. The biometric solution would need to be resource-efficient, since in some cases it would be running on solar powered equipment when in remote rural areas. The solution also would need to be economical and easy to implement to allow widespread deployment while keeping the project within budget.
After studying a number of alternatives, CODE selected the MegaMatcher biometric software development kit (SDK) from Neurotechnology for the development of their CODE Digital Solution (CDS) for Electronic Voter Registration (EVR). MegaMatcher fingerprint and facial technology is used for biometric enrollment and verification in the field and then the powerful MegaMatcher multi-biometric fingerprint and face identification engine is used for matching and flagging of duplicate registrations in the national voter database. For CODE, the combination of strong NIST test results and the proven reliability of Neurotechnology's algorithms in a number of challenging field implementations were key elements in their decision.
The CODE Inc. EVR solution was developed to provide a flexible, durable hardware and software solution that would enable Kenya's electoral authorities to travel to remote areas and set up temporary voter registration sites. CODE worked closely with Kenyan authorities to implement the solution, providing tools and support to train and help electoral workers gather information and biometric data to register voters, issue ID cards and compile and process data to generate an accurate, updated registry of voters. Having strong Kenyan Electoral teams in the field helped ensure the success of the project.
The EVS solution implemented by CODE includes both hardware and software for field and office use. For each voter registered, one facial image and two fingerprints (thumbprints) are enrolled in the system using the MegaMatcher fingerprint and facial algorithms to verify the image quality and process the biometric data. The voter provides basic biographical information that is then linked with his or her digital face and fingerprint images to provide a unique multi-biometric record for each voter.
The voter is issued an ID card that he or she must present in order to vote in upcoming elections. The entire process, from application to registration ofthe biometric data to issuance of the voter ID card takes approximately four minutes.
Once the biometric data is registered in the field, it is saved onto an encrypted USB thumb drive for transport back to the national data center for processing. There the data goes to the Adjudication Module, which uses MegaMatcher technology to rapidly search the database for any duplicate registrations, comparing and cross-referencing facial and fingerprint data with all existing records in the system. If any discrepancies or duplications are found, they are examined on-screen by a trained IIEC operator and a ruling is made. Changes are recorded in the database and the duplicate cards are cancelled. Once all discrepancies have been resolved, the IIEC updates and prepares a clean, accurate voter registry that meets Kenya's stringent new legal and administrative requirements for management of elections.
To date, the CDS has been used to register and issue photo ID cards to nearly 1.5 million voters all across Kenya. These voters were part of a pilot program implemented across 1600 registration sites in 18 of Kenya's 210 voting constituencies.
Voters who participated in the pilot program used their new IDs for the first time to vote on an August 4, 2010 referendum. Registered voters went to one of Kenya's 2000 local polling stations and presented proof of their eligibility to vote. For those with the new IDs, the right hand thumbprint was then used to “check in” each voter and strike him or her from the list of eligible voters, ensuring that each individual was only able to vote once.
The Kenyan government has been pleased with the results of the pilot program so far and the August 2010 voting process was widely hailed as a success. The polling was efficient and produced accurate, credible results. In so doing, the new system has helped to reinforce confidence in the election process.
Future plans call for rolling out the system across the remaining voting constituencies, and the program is expected to be completed before the 2012 elections. When complete, approximately 13 million Kenyan voters will be registered in the new system.
The next phase of the CDS solution, which is currently in development, will incorporate iris biometrics and will take advantage of even faster multi-biometric matching capabilities offered by the MegaMatcher Accelerator hardware/software solution from Neurotechnology.