Electronic Poll Books – Cuyahoga County

Electronic Poll Books – Cuyahoga County

Anticipated Approval Dates for Electronic Pollbooks

Memorandum of Understanding
03/14/17—County Council First Reading
03/28/17—County Council Second Reading
04/12/17—County Council Third Reading and Approval

Electronic Pollbooks Contract
03/16/17—Technical Advisory Committee Approval
04/25/17—County Council First Reading
05/09/17—County Council Second Reading
05/23/17—County Council Third Reading and Approval

With the most recent state budget allocation of $12.75M for the procurement of electronic poll books, researching the possible purchase and implementation for Cuyahoga County has become one of this agency’s top priorities.  Our primary objective is to provide the voters of Cuyahoga County with the most efficient means of checking in at the polls on Election Day.  Electronic poll books provide a method to streamline this process while simplifying poll workers’ duties, enhancing internal BOE processes and procedures, and providing the most accurate and reliable records available with regard to voter participation.

Electronic poll books will affect the way elections are administered in Cuyahoga County. Before making such a significant investment, the Board of Elections must determine which model of electronic poll book – combined with software requirements – can best meet the needs of our unique voting population.  Currently there are seven (7) e-poll book vendors certified in the State of Ohio.   


Product Demonstrations (August 2015)

All seven vendors were invited to Cuyahoga County in August to demonstrate the functionality of their products.  (A link to the video of each vendor’s presentation can be found below.) The audience was made up of senior management staff, board members, election activists, media and election officials.  Each person was asked to complete a vendor survey, drawing specific attention to key components of each vendor demonstration.

Our goal, after careful review and consideration, is to narrow the decision to two or three viable options for Cuyahoga County.  The products/vendors selected will be further vetted during the next stages of pilot projects in the November 2015 General Election and mock elections shortly after.  


Presentation Videos (August 2015)


Individual Vendor Meetings (September 2015)

Based on feedback from the product demonstrations, the CCBOE invited three vendors, KNOWiNK, Tenex, and VOTEC back to the Cuyahoga County Board of Elections for a more in depth demonstration of their electronic poll book systems.  These meetings took place Wednesday, September 16th through Friday, September 18th and allowed the senior management team the opportunity to ask additional questions, address concerns, and review specific functionalities more extensively than the initial demonstrations allowed for.

After examining and evaluating these three electronic poll book systems, we determined they could all be viable options for Cuyahoga County. All three systems have pros and cons and would require some customization to adapt to our specific needs. The vendors from all three companies have expressed their willingness to work with us in order to create an electronic poll book system that fully addresses our needs.

In order to make the most informed decision about which product to select, we are including all three vendors in our pilot project, which will take place during the November 3, 2015 General Election. Ultimately the success of these units depends on how they perform on Election Day at our polling locations. We will work closely with the poll workers during our pilot project and survey their interactions with the units. Additionally, we will solicit feedback from the voters and their experiences with the electronic poll books. Furthermore, after Election Day we will also conduct an extensive mock election with the electronic poll books in order to test more complex scenarios that did not arrive during testing in the field during the election.

Individual Vendor Meeting Videos (September 2015) 

Public Security Presentations – Thursday, October 22, 2015

Security of election technology has always been a top priority of the Cuyahoga County Board of Elections.  Based on our experiences with selection and implementation of our tabulation system, we were certain that assistance from outside the agency was necessary. Director McDonald enlisted Jeremy Mio, Security and Research Manager in the Department of Information Technology of Cuyahoga County. We relied heavily on his expertise to assist with the overall security analysis of each system.   The initial phase of the security analysis was a public security presentation in which the three vendors, KNOWiNK, Tenex and VOTEC, were required to have their own security experts, in both the hardware and software, outline the strengths of their system.   
In addition to being an active participant in the question and answer portion of the security presentations, Mr. Mio also reviewed the final lab reports from the testing required during state certification.  During this review it was determined that although all of the electronic pollbook systems underwent required testing during the certification process, the main focus of that testing was the protection of the voter registration data.  
Listed below are the sections of the Ohio Certification Testing Matrix relating to security.  

  • All information contained on the electronic pollbook to be encrypted.

  • Electronic pollbooks have the capability to store a local version of the voter registrationdatabase to serve as a backup; and produce a list of audit records that reflect all actions of the system including in-process audit records that display all transactions.

  • Electronic pollbooks be capable of providing secure reliable transmission of voter and election information to Board of Elections.

In a report presented at the January 14, 2016 Board Meeting, Jeremy Mio recommended additional security testing be required of any vendor seeking to provide electronic pollbooks in Cuyahoga County.  The testing was conducted by NTS Labs and Mr. Mio’s final security analysis is detailed on page 24 of this report.

PILOT PROJECT (NOVEMBER 2015) – KNOWiNK AND TENEX
During the November 3, 2015 General Election, the Cuyahoga County Board of Elections conducted an electronic pollbook pilot project in which we tested two vendors’ products on a small scale during a live election. The two vendors selected to participate in the pilot were KNOWiNK and Tenex. Each vendor was assigned two four-precinct polling locations in inner-ring suburbs of Cuyahoga County to demonstrate the effectiveness of their electronic pollbook during a live general election. 

  • Four Polling Locations, 16 Precincts, 4,241 Ballots Cast

    • Euclid – Tenex
    • Euclid Early Learning Center
    • Sts. Robert and William Catholic School
    • Lakewood – KNOWiNK
    • Emerson Elementary School
    • Horace Mann Elementary School

Euclid and Lakewood were selected by the Core Group due to their physical location which provided easy freeway access in case of any unforeseen issues, the number of registered voters assigned to ensure robust voter participation, and the diverse demographics of the communities.
Representatives from KNOWiNK and Tenex were responsible for conducting training sessions for the Precinct Election Officials hired to work at electronic pollbook locations on Election Day. Both vendors worked closely with the Board of Elections’ staff and Election Day workers to ensure everything ran smoothly. 
The pilot project gave the Board of Elections an opportunity to observe the training, feasibility, connectivity, time, and any adverse factors that could arise on Election Days. Additionally, we were able to solicit feedback from voters and Election Officials regarding the functionality of the products tested at their polling locations. The feedback received was overwhelmingly positive and optimistic. Voters and Election Officials alike were excited about the implementation of new technology in the near future.
   
PILOT PROJECT (DECEMBER 2015) – KNOWiNK ONLY
An unanticipated small recall election in Ward 3 of East Cleveland in early December provided us an additional opportunity to test electronic pollbooks.  Unlike the previous pilot project, the focus was mainly on internal processes and procedures relating to the logistics of electronic pollbook deployment. Staff from the Election Support Department delivered the units and Election Officials took primary responsibility of setting up the units in the polling locations based on newly enhanced location diagrams (example below) to streamline the voter process.  KNOWiNK provided us with eight units to deploy in the three polling locations in Ward 3 and provided minor assistance throughout Election Day.  The feedback from Election Officials and voters in all five precincts was again overwhelmingly positive.

MOCK ELECTION (MAY 2016)
The Cuyahoga County Board of Elections invited KNOWiNK and Tenex to be part of a mock election on May 10, 2016.  The Board of Elections decided to host a mock election to gain a better understanding of how each electronic poll book would perform under diverse variables on Election Day, which was identified as a best practice during our research and conversations with other election jurisdictions.  After our initial exposure to the products during the pilot project, our staff identified a few product enhancements we felt were necessary for use in Cuyahoga County.  The team worked closely with both vendors on customizations such as simultaneous same day election setup, automatic generation of electronic versions of required Secretary of State (SOS)OS forms, ballot inventory monitoring, end of the night signature reconciliation, and election official payroll.  Since the mock election would be the final demonstration opportunity prior to our final selection, the intention was to utilize as many of these enhancements in this process as possible.

The Pink Room on the first floor of the Board of Elections transformed into two separate polling locations, each housing two precinct tables and one voter assistance table on either side of the room. Board of Elections staff was assigned to serve as Precinct Election Officials (PEOs) and attended the mandatory training sessions for both vendors.  We invited about 3,000 veteran PEOs, who normally serve in the field on Election Day, to act as voters to assist in the process. Turnout for the mock election was much higher than anticipated as 189 individuals attended to aid in our selection process.  Turnout for the mock election was much higher than anticipated as 189 individuals attended to aid in our selection process. 
Prior to the event, 325 voter check-in packets were created for each vendor which included specific “role-playing” scripts (e.g., regular voters, provisional voters, voters who had recently moved, etc.) and corresponding forms of identification based on the scenarios being simulated. Upon arrival, each “voter” was provided with a brief overview of the process, a voter check-in packet (assigned at random) and instructions on how to proceed. The “voter” was required to check-in on both units using the voter information/identification provided and then complete a brief survey about their experience.  

An overview of the survey responses is detailed in the next section.
Check-in time and accuracy was monitored by Board of Elections staff.  Employees did their due diligence throughout the day to make the mock election as close to a real election as possible from start to finish. We tested each unit’s functionality with opening procedures such as PEO check-in and the time it takes to set up the unit. Furthermore, when it came time to close the polls we tested ballot reconciliation reporting (Certificate #1) and printing 11:00 a.m./4:00 p.m. voter lists.

Since the August 22nd meeting, we have identified forty polling locations, containing a total of 146 precincts, with the highest number of provisional ballots cast in both the 2008 and 2012 presidential elections.  These polling locations were evenly distributed amongst the two vendors based on their geographical location.

The e-poll books will be used in a limited capacity to look up where voters live and where they should vote, nothing else. Voters will not sign-in or be issued ballots based on e-poll book information.  Taking these steps should help cut down on the number of provisional ballots cast in this Election. In the 2008 and 2012 elections there were close to 30,000 provisional ballots cast.
  The voter look-up units will be stationed at the Voter Assistance Table and used to assist the poll workers with quickly and accurately determining voters’ assigned precincts and polling locations based on their current addresses.  Voters that are in the incorrect polling location will be provided with a print out of step by step driving directions to their current polling locations which will also be able to be sent via email or text message if desired.

Our e-poll book team has been working with both vendors to iron out the details of the applications on each system and will oversee the vendor led training sessions for the election officials assigned to the locations involved in this project.  Ms. Soeder will report back on the success of the voter look-up units after the election.

Pilot Project

Voter Lookup | November 2016

  • We identified 40 polling locations, containing a total of 146 precincts, with the highest number of provisional ballots cast in both the 2008 and 2012 presidential elections. These polling locations were evenly distributed amongst the two vendors based on their geographical location.

  • The electronic books were used in a limited capacity to look up where voters live and where they should vote. Voters did not sign-in or issued ballots based on electronic poll book information.

  • The voter look-up units were stationed at the Voter Assistance Table and used to assist the election day officials with quickly and accurately determining voters’ assigned precincts and polling locations based on their current addresses. Voters that came to the incorrect polling location were provided with a print out of step by step driving directions to their current polling locations which was also able to be sent via email or text message if desired.

  • The goal of using the electronic pollbooks in this capacity was to cut down on the number of provisional ballots cast in this Election. In the 2008 and 2012 elections there were close to 30,000 provisional ballots cast.

Funding/Cost Analysis

Fortunately, we are able to bypass the independent bidding process as the Ohio Department of Administrative Services is handling the procurement aspect of the purchase.   DAS worked closely with the Secretary of State’s Office to determine the procurement process and negotiated standard terms and conditions on behalf of us and all other counties in Ohio. They have also prepared and disseminated a price schedule of all products and services offered for each certified vendor in the State of Ohio.  All of the contract details and pricing information can be found on the State of Ohio’s website at the link below. 

http://procure.ohio.gov/proc/viewContractsAwards.asp?contractID=33616

News Articles:
Electronic poll books will be at voting locations across the state by November 2016 (News article by Karen Farkas, Northeast Ohio Media Group)

Electronic Pollbooks Memo to Dale Miller

Follow up regarding electronic pollbooks