Field Canvassing Tablets

When I was working on political campaigns, I would often work closely with the field program that involved volunteers and paid staff members knocking on doors of voters. The part of the campaign was extremely data intensive in both which voters to target as well as properly recording the results for later use. When working on larger campaigns that employed a significant number of paid canvassers, I found that we invariably had one or more individuals who would try to fake their block walking results. This often resulted in the person not walking their assigned area at all or just partially, then updating paper walk sheets with bogus results to make it appear as if they walked the entire neighborhood.

Project Goals

1. Reduce paper walk lists from field campaigns
2. Reduce time to enter walk results into NGPVAN's voter database
3. Reduce data entry errors entering results into voter database
4. Identify faked walk results and take appropriate action with staffers
5. Use off-the-shelf software to create this solution (no programming)

In 2012, I began working on a possible solution to this problem. NGP VAN's voter database website had a mobile app that could be used on phones and tablets. The goal of the project was to create a tablet configuration that could be given to our paid canvassers. The tablet needed to be locked down so users could only access the VAN voter database app. The tablet also needed to have GPS enabled with routine pings setup to track its physical location. The locked down tablet would prevent it from being used for purposes other than canvassing voters. The GPS pings would allow us to track the tablet in case of theft or if the results appeared to be suspect.

NOTE: These tablets were intended for paid field canvassers rather than volunteers helping a political campaign.

Development & Testing

Based on the requirements, the tablets needed to be accessible on a local carrier network. I chose to go with Verizon as I was already a customer and the carrier had good coverage over the Dallas/Fort Worth metroplex. I chose to purchase the Galaxy Tab 2 7.0 due to its relatively low cost compared to other tablets available. The processing power and graphic card requirements were extremely low to run the applications. Another reason that I went with Android was because I could get a discount on bulk purchases, which was not offered for iPad tablets.

Tablet Components

• Galaxy Tab 2 7.0
• Otterbox Defender case
• 42Gears SureLock mobile app
• 42Gears SureMDM mobile app
• NGPVAN MiniVAN mobile app

I started with two initial tablets for testing purposes. The Otterbox Defender cases protected the tablets from damage. The main issue was ensuring that power was full before a block walk program began. If the tablet ran out of power, the block walk for that canvasser was effectively over without a replacement device. We did run into some technical issues with some tablets just dying or randomly rebooting. However, these issues seemed to be extremely isolated.

Otterbox Defender case

The Otterbox Defender case was an essential component for the tablet program. Over the years that we operated the program, a number of tablets were dropped on various surfaces by us and field canvassers. The nicks and scratches on the Otterbox cases clearly showed the history of these incidents. However, none of our tablets suffered significant scratches or cracks on the devices due to the Otterbox cases.

42Gears.com SureLock

The SureLock application was developed by 42Gears.com. The program was designed to lock down tablets and other mobile devices for use in kiosks. This would allow companies to use a simple Android or Apple iPad tablet in a kiosk that might sit in a shopping mall or company's lobby. This application gave us the ability to lock down various settings in the tablet, discussed below, and force NGPVAN's MiniVAN app to run on screen. Users would not be able to exit the MiniVAN application. We did experiment with enabling other apps but had limited success with our field canvassers. It was better to keep things simple with just having the tablet forced to only run MiniVAN and nothing else.

42Gears.com SureMDM

The SureMDM application was also developed by 42Gears.com. An MDM is a Mobile Device Management application that is often used for company-issued phones to lock down certain security settings and remotely wipe devices, if stolen. We did examine a competitor's MDM application but they would not allow GPS tracking pings less than 15 minutes apart. The combination of SureLock and SureMDM gave us the ability to set our GPS tracking pings to a time that worked best for monitoring our field canvassers. We could also use SureMDM's web interface to see where devices were in near real-time and remotely wipe or reboot a device. Thankfully, we never needed to actually remotely wipe or reboot a device. No tablets were stolen during our program.

GPS Tracking Interval

The GPS Tracking Interval was an area that took some adjustment to get right. We finally settled on a three minute interval between GPS pings for the tablet's location. Any interval below three minutes appeared to cause undue stress on the tablet's battery life with little benefit. An interval beyond three minutes made it difficult to verify if a block walker was actually stopping at homes or not with large gaps between location pings.

Disable Auto-Rotation

One issue that became apparent during testing was the need to disable auto-rotation of the tablet screen. We kept all tablets locked to a vertical screen position. The processing power of the tablet was not extremely powerful, which would slow down tablet response times when trying to auto-rotate the screen. This could be problematic when the block walker is trying to speak with a voter, look at any screen data, and enter their responses.

Screen Sleep Timer

Another issue that appeared in testing was the need to set a long timeout before the screen went to sleep on the tablet. This caused a larger drain on the battery but helped offset a problem where the block walker is talking to a voter and their tablet goes to sleep while doing so. This could cause confusion and delays as the canvasser is trying to reawaken the tablet while the voter is staring at them. We settled on a 10 minute sleep timer for the tablets, which normally gave ample time before the tablet screen went blank.

Field Canvassing Results

After initial testing, we began expanding the tablet program and began using them on campaigns with paid canvassers. The results were seen immediately with the campaigns that the tablets were used on. We significantly reduced the amount of time spent entering completed block walking data into NGPVAN's Voter Database. We were also able to identify block walkers suspected of faking their results and provide more definitive evidence that the results they entered were falsified.

Paid canvassers that used the tablets tended to have a love or hate relationship with the devices. The tablets provided more information about voters and weighed significantly less than paper walk lists. Many users also found it easier to enter data into the tablets. Other paid canvassers hated the devices because they knew they were being monitored through them via GPS tracking. However, most paid canvassers who hated the tablets were also often found to be faking entire or partial walk results.

Compare Results: GPS Pings versus Entered Data

The following images were captured when producing the evidence that two paid field canvassers with two different tablets were falsifying their walk results. Their walk results were thrown out with both canvassers being fired. Their assigned walk list was then reassigned to other canvassers to get actual results. Each mark on both maps had a date and time associated with the marking. These corresponded to when the GPS ping was taken for the image on the left. The timecodes for the image on the right was when data was entered into the MiniVAN mobile app.

Houses north of Northwest Highway were not even walked. The large spaces between houses south of Northwest Highway indicated that the field canvassers walked the street but did not actually knock on any doors. At best, they simply hung a door hanger and moved on without trying to see if any voter was home to speak to.

GPS Tracking Pings	Targeted Voter Locations
"Where they actually were"	"Where they claimed they were"

Real Time Monitoring

Another benefit of the GPS Tracking was the ability to perform near real-time monitoring of where tablets were in the field. The regular GPS tracking pings allowed us to project the location, within the last 3 minutes, of every device in the field on a map. The following image is from a campaign that was utilizing twenty three tablets for their field program. On this particular day, canvassers were speaking with targeted voters living on what are called the "M Streets" of Lower Greenville in Dallas, Texas. The "M Streets" name comes from the fact that all the streets in that area start with M, like McCommas, Morningside, Mercedes, Merimac, Monticello, etc.

This gave the staff who are coordinating the field program a heads up if something looks amiss. The staff member could either call the canvasser to see how things are going or actually drive over to that location to confirm everything is ok.

Tablet Return On Investment (ROI)

The tablets proved to be a long-term investment that benefited greatly from the election cycles specific to Dallas, Texas. In Dallas County, Municipal Elections end right around the time that Primary Elections are about to begin. Primary Elections roll directly into Runoff Elections and then General Elections. When the General Election that runs on even numbered years concludes, the local Municipal Elections that run on odd numbered years begin kicking off. This cycle allowed the tablets to be in near constant use on campaigns with minimal downtime.

As the tablets were on the Verizon Wireless carrier network, deprovisioning the tablets in between every election would have been difficult or increased costs of maintaining the units. As their use gained in popularity, it was easier to keep them online and unused for the one to three months in between the next field program for an upcoming election. This was also coupled by the fact that the Otterbox cases were not designed to be repeatedly taken apart and put back together to replace SIM cards in the tablets.

In a different state that only hosts elections every two years, significant changes would have had to be made to make the tablet program cost effectively. This would have required using a month-to-month subscription for a wireless carrier service. It would have also increased the initial investment into each tablet as we would not have been able to use one or two year contracts with Verizon to reduce initial costs. Overall, a dedicated tablet program with thirty tablets or more would best be suited for a full-time political consultant who is moving from one campaign to the next throughout the year.

Verizon Wireless Technology Innovation Award

In 2013, my business sales contact with Verizon Wireless learned about the work that I was doing with the field canvassing tablets. She suggested that I enter into Verizon Wireless' Technology Innovation Award contest. The contest was designed for companies that are using Verizon Wireless services in new and innovative ways. The tablets that I was working on caught their attention. A few months after submitting the paperwork, the content organizers sent a camera crew over to make the following video for the contest. The contest finale held a lunch event where we learned that our field canvassing tablets had won the grand prize. The prize money was a great benefit as it covered all our initial investments into the tablet program and gave us additional funds to expand it further. Before my wife and I ended the program in Fall 2014, we were operating thirty tablets.

Here is the actual award: