Speech Primed for Future Elections

Article Featured Image

After months of recounts, court battles, and controversy in Minnesota’s election of Al Franken to the U.S. Senate, Americans across the country head to the polls this November to decide a host of local and state contests. And once again, voters won’t have access to Prime III, Juan Gilbert’s revolutionary multimodal speech and touchscreen voting technology. Instead most voters will have to rely on unreliable paper ballots and outdated voting machines.

Nonetheless, Gilbert, a computer science and software engineering professor and chair of the Human Centered Computing Lab at Clemson University, remains optimistic about Prime III and its chances to drastically improve the accessibility, accuracy, and security of elections.

According to Gilbert, Prime III is now better than ever: He and his team recently tweaked the system, adding new capabilities that make it even more accessible.

After learning about a voting system that enables visually impaired people to cast ballots via a keypad, Gilbert conferred with the system’s inventor and added similar technology to Prime III. The system now allows voters to use keypad navigation in addition to the system’s speech and touchscreen functionality.

“Now you can navigate via speech only, keypad, and/or touch,” Gilbert says. “And you can do it in any combination. It speeds up the interface. It speeds up interactions for blind users. It makes it even more universally accessible.” 

Gilbert and his team also added speech-enabled anonymous spelling capabilities to Prime III. Now if a visually impaired voter wants to cast a ballot for a write-in candidate, he can do so—privately and secretly—with the power of voice. “We developed a speech interface interaction to accommodate anonymous spelling via voice,” Gilbert says. 

The anonymous spelling feature divides the alphabet into groups of five letters and uses voice prompts—Is the first letter of the candidate’s last name between the letters A and E?—to help voters spell out candidates’ names. By answering the prompts with yes or no responses, users anonymously spell out a portion of the write-in candidate’s name. At that point, the system speculates as to the full name of the write-in candidate and prompts the user for verification. 

“It’s not even a question any longer—our system is the most accessible voting technology ever created,” Gilbert says. “No other system can accommodate more people than we can.”

Additionally, Gilbert is pursuing a grant from the U.S. Election Assistance Commission and the National Institute of Standards in Technology, which received $5 million in federal funding for an accessible voting program.

“Obviously we’re going to compete for that,” Gilbert says, noting the outcome of the program will reform how the government addresses the issue of accessible voting.

But as to when Prime III might be available to the American public, Gilbert admits the makers of voting machines will likely not adopt his technology unless mandated to do so by the federal government. “Unless somebody tells them this is what you need to do, they’re just not going to do it,” he says. “They’re doing what they’ve always done, and they’re content with that. So that’s a problem, but I think with this grant, that’s going to change things.”

The expense is likely another reason for sticking with the status quo. “I understand their perspective…that’s why this is probably going to take longer than we would have hoped,” Gilbert concedes.

SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues