IE 11 Not Supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

How AI Translation Solved a Small Town’s Language Barrier

Pocketalk, an AI-powered handheld, real-time translation device, has helped law enforcement in a small Oregon town bridge communication gaps with non-English speakers during critical incidents and daily interactions.

Haystack Rock in Cannon Beach, Ore., on a cloudy day.
Haystack Rock in Cannon Beach. Photo by Nikki Davidson.
Cannon Beach, Ore., a town of 1,500, swells each summer with an influx of 200,000 visitors drawn to Haystack Rock, a National Geographic-recommended marine sanctuary towering 235 feet above the beach.

This presented a unique challenge for local law enforcement, where small-town logistics met a constant stream of non-English speakers, complicating everything from traffic stops to basic visitor inquiries. But AI-powered handheld translation devices are now helping the city tackle the challenge.

Cannon Beach Police Officer Cody Dietel told Government Technology in an interview that growing challenges with language barriers came to a climax when a resident who did not speak English began calling 911 on a frequent basis for non-emergencies.

“He was intoxicated quite often and was misusing and calling 911 because he needed help. People would get there and nobody knew what he needed help with,” said Dietel, adding that the problem persisted for months until one day the man got physically violent and was arrested for safety reasons.

He was taken to the nearby Clatsop County Jail, which had a device called Pocketalk. It’s a handheld electronic rectangle about the size of a cellphone that provides real-time, two-way translation. Speakers push a button and talk into it, and it delivers a verbal and written translation.
Photo of a hand holding up a Pocketalk device showing a language translation on the screen. White background.
The handheld Pocketalk device uses AI to provide real-time translation of 92 different languages.
POCKETALK
“It was an amazing tool,” said Dietel. “I was able to take that to somebody I’ve dealt with so many times and actually speak to him and tell him not only what’s going on and what he’s doing wrong, but also make sure he gets the help that he needs. When you have somebody that’s frustrated because they’re not understood, and you can finally understand them, it brings all of our energy levels down.”

Dietel was so impressed he chose to purchase a device with his own money, which later inspired his department to invest in more devices after seeing it in action.

“It’s nice to see people kind of light up because they realize that you have a way to communicate with them, which is great,” said Dietel. He added that the frequent 911 caller also stopped calling 911 for non-emergencies. Dietel now uses the device's transcript history to revisit past conversations and remember details such as the man's son's name to strengthen their connection.

While Cannon Beach's experience highlights the value of such technology for daily interactions, the imperative for comprehensive language access in law enforcement is highlighted by a tragic 2023 incident in Wisconsin. A language barrier between a sheriff's deputy and a Spanish-speaking dairy farm worker resulted in a flawed investigation into his son's death, ultimately leading the Dane County Sheriff’s Office to collaborate with the Department of Justice on improving language access for individuals with limited English proficiency.

“Our language access policies and procedures are a testament to our commitment to equality and inclusion,” said Sheriff Kalvin Barrett. “By proactively addressing language barriers, we are fostering a more connected community where everyone can fully participate.”

U.S. census data further illustrates the geographically diverse distribution of non-English speaking populations, bringing to light the need for adaptable, community-specific language access strategies nationwide.
Since 2017, Pocketalk has sold about 2 million devices and executed nearly a billion translations. Historically, they worked primarily with schools, health-care institutions and governments, but recently they’ve expanded to law enforcement.

While they feel bilingual personnel and on-demand human translation services continue to fill critical needs in specific situations, Pocketalk has heard success stories with public safety agencies who need regular and accurate real-time solutions in situations where delays can potentially put lives at risk.

Officer Dietel detailed that when the department relied on tools such as Google Translate in the past, they would often experience frustrating connection issues that could make tense situations worse. Meanwhile, using on-demand translation services would often come with uncomfortable waits that would further create a barrier between them and the people they were trying to serve.

Pocketalk technology leverages multiple translation engines, in what they call a “translation solution in the cloud.” AI is also used to provide real-time refinement of the translation itself, smoothing the output to be a clean and accurate representation of what the person said.

The translations performed through their devices are not used to build broader AI language models. Pocketalk adheres to HIPAA, GDPR, ISO 27001, FERPA, and COPPA security standards to protect data exchanged during translations.

“Security is paramount to what we’re doing, and really that’s different from some of the other companies in this space,” said Doug Bieter, head of global partnerships for Pocketalk.

Additionally, Pocketalk is positioning the technology as a solution for areas where law enforcement agencies struggle to recruit qualified bilingual talent.

“Now you can speak 92 languages immediately, it provides the ability to bring the best talent and then leverage technology to communicate,” said Bieter.

While the tool is becoming popular for law enforcement, it’s also expanding into courtrooms. In Illinois, judges are using it to officiate weddings and communicate in traffic court. In Colorado, it’s used by the Department of Motor Vehicles in state driver's license offices.

“Other solutions have been embedded into these places for decades, and now there’s an AI-based machine translation solution that really opens up translation way broader,” said Bieter. “The good thing about what we do is translation where there’s no bias, there’s no interpretation since we’re just translating word for word.”
Nikki Davidson is a data reporter for Government Technology. She’s covered government and technology news as a video, newspaper, magazine and digital journalist for media outlets across the country. She’s based in Monterey, Calif.
Sign up for GovTech Today

Delivered daily to your inbox to stay on top of the latest state & local government technology trends.