US Navy chooses Mycelial to synchronize AI/ML data between submersibles

Case Study
Delivering AI-powered insights in the cloud is challenging enough; doing so on a drone submarine beneath the sea is even more difficult.

Modern Navy vessels host a wide array of sensors and systems to enhance combat effectiveness across a multitude of vectors - system control, platform reliability, and situational awareness, among others.

Among the most advanced sensing platforms are the Navy’s Unmanned Undersea Vessels (UUVs) and Unmanned Surface Vessels (USVs). These unmanned vessels enhance Navy’s real-time decisioning capabilities by collecting data for extended periods in remote locations and relaying it back to command.

As part of the Department of Defense’s Joint All Domain Command and Control (JADC2) program and Navy’s internal initiatives to bring AI/ML capabilities to bear, next generation unmanned vessels will include the ability not only to collect sensor data, but also apply AI and machine learning to that data in real-time, at sea.

Ultimately, the success of Navy’s project relied on sharing data seamlessly between the fleet of unmanned vessels, manned vessels, and Cloud environments. Leveraging Mycelial’s Kafka connector, Navy vessels using Kafka or Red Panda as the onboard system of record synchronize with each other in real-time when connected but remain fully functional when disconnected under the sea.

Moreover, Mycelial demonstrated data delivery throughput matching heavier, JVM-based tools like Kafka and Apache Nifi with a single binary using ~7mb of memory - critical in SWaP (space, weight, and power)-constrained environments.