mermaid

A maritime radio can carry a message across miles of open water in seconds. It can also leave the wrong person unaware that the message was ever sent. That vulnerability made headlines in August, when Mark Zuckerberg's yacht Launchpad was reported to have missed a Coast Guard assistance broadcast to a nearby skiff off Alaska because its crew was monitoring another channel. The Coast Guard later determined that the skiff was not in distress, and another vessel provided assistance.

The incident exposed an awkward reality in maritime communications. A system can be functioning exactly as designed while a critical exchange still disappears from the awareness of the person who needs it. U.S. Coast Guard guidance requires vessels to maintain a watch on VHF Channel 16, while working traffic can move onto other channels, creating a communications environment in which attention remains divided across separate conversations.

The wider safety picture makes that problem harder to dismiss. Allianz Commercial recorded 2,818 shipping incidents involving vessels over 100 gross tons in 2025, even as the long-term rate of total losses continued to fall. Research published in Safety Science in 2025 also found that speech acts and lexical factors accounted for a 37.77% probability of accident risk within its analysis of 429 maritime accidents and near misses recorded between 2018 and 2023.

Language itself remains a recognized maritime safety issue. The International Maritime Organization developed its Standard Marine Communication Phrases specifically to reduce misunderstandings caused by language barriers, while recent IMO work has continued to examine the modernization of maritime communications. The basic infrastructure has evolved considerably, yet much of the information travelling through it still arrives as transient voice that must be heard, interpreted, and remembered by a human being in real time.

Dean Mancini, Chairman and CEO of TDI Technologies, Inc. (Technology Driven Innovation) and MermAId, Inc., sees that gap as an opportunity for artificial intelligence to strengthen the communications layer already used at sea. "Communications can be missed for many reasons," Mancini notes. "People get distracted, radio signals are static, transmissions aren't always clear, accents can create confusion, and sometimes you're simply on the wrong channel."

The proposition is less about putting AI between bridge teams and their radios than giving the traffic they already generate a persistent digital layer. According to Mancini, MermAId's patented technology processes marine VHF traffic at the edge, converting live voice into searchable records while supporting transcription, translation, recording, playback, and automated alerts. The system is designed as a standalone unit connected to a vessel's VHF antenna, allowing existing radio equipment to remain in place.

Mancini finds this pivotal, as radio traffic contains information that can become valuable seconds, hours, or days after it is transmitted. An emergency call can be flagged, a multilingual exchange can be translated, a day's communications can be reviewed without relying on one watchstander's memory of what passed across a particular channel.

The commercial logic emerged from a much more demanding environment. TDI Technologies' federally funded CONSCIOUS (Contextual Observations for Natural Speech Characterization On Unmanned Ships) program was developed for the U.S. Navy to help unmanned surface vessels understand and respond to bridge-to-bridge VHF communications.

The program received a Phase I SBIR award in 2020 and a Phase II award worth $1.699 million in 2022. Federal records state that the system used speech-to-text, intent classification, information extraction, and text-to-speech to help unmanned vessels interpret human maritime communications.

Mancini argues that the same underlying capability has a more comprehensive role on conventional vessels. "From a defense perspective, it's not only miscommunication, but it's also situational awareness," he explains. "They want to know what's happening in harbors, what's happening at sea, and when they have a sailor listening to one channel, they're not getting the whole picture."

That shift turns radio from a fleeting conversation into an operational data source. For commercial operators, Mancini believes that the same record can also provide an evidentiary trail after an incident, giving investigators access to what was actually communicated instead of relying solely on competing recollections.

The larger question is how far maritime technology should go in capturing information that crews already depend on. AI, Mancini argues, will not eliminate fatigue, distraction, or imperfect radio conditions. It can, however, give maritime operators another layer of awareness around a communications system that has long depended on someone being in the right place, on the right channel, at precisely the right moment. At sea, that extra layer could matter most when every transmission carries operational significance.