Advancements in AI Autonomy: Royal Navy’s Uncrewed Submarine Trials

VIZGARD has successfully utilized its AI-driven autonomy platform, FortifAI, to conduct sea trials with the Royal Navy’s uncrewed submarine.

Deployment of FortifAI on the XLUUV

The FortifAI system was integrated into a nine-meter eXtra-Large Unmanned Underwater Vehicle (XLUUV), with its performance being monitored remotely from a support vessel nearby.

Innovative Navigation and Detection Techniques

During the mission, FortifAI employed the submarine’s inertial navigation system, along with periscope optics and an onboard RF receiver, to passively confirm the identity and location of other vessels encountered during the journey.

real-Time Data logging and Analysis

The operation was conducted in real-time using VIZGARD’s NVIDIA-powered VizEdge device, which recorded submarine INS data, visual detections, MP4 videos, and AIS tracks for subsequent analysis. This capability allows VIZGARD to enhance and expand future functionalities.

Enhancing Data Readiness with Semi-Unsupervised Learning

This trial also provided a platform to evaluate FortifAI’s new semi-unsupervised self-training feature. The “Dataset Pruning” technique acts like an initial review by a team member, ensuring the quality of detections before they undergo final human validation.

This approach significantly accelerated the data preparation process and improved accuracy, notably in identifying distant objects that might be missed by human operators.

Tracking Dark Vessels and enhancing Situational Awareness

In addition to bolstering real-time situational awareness, FortifAI generated a list of the last known locations of unidentified “dark vessels,” which were identified using satellite imagery taken just before the mission commenced.

The system recorded the relative distance and direction of these unknown vessels, along with their estimated lengths derived from aerial views, and displayed this data as augmented reality markers on the live video feed.

Future Prospects and Collaborations

This initiative was commissioned by the Ministry of Defense to explore innovative sensor technologies aimed at enhancing platform safety and extending autonomous operations potentially over thousands of nautical miles.

VIZGARD is eager to further its research in semi-unsupervised learning and advanced computer vision for maritime intelligence, collaborating with a leading unmanned surface vehicle manufacturer in 2022.