Case Study: U.S. Navy improves underwater mine detection with Domino Data Lab

U.S. Navy makes mine detection intelligence more trustworthy with Domino

The U.S. Navy needed a better way to monitor, retrain, and redeploy machine learning models for underwater mine and target detection as conditions changed and existing systems became unreliable after deployment. Working with Domino Data Lab and its Domino Enterprise AI Platform, the Navy sought a commercial MLOps solution to improve the performance and trustworthiness of its underwater intelligence.

Domino Data Lab helped the U.S. Navy build a modern MLOps pipeline for Project AMMO, enabling faster deployment, retraining, and governance of automatic target recognition models in AWS GovCloud. The result was a major speedup: edge model deployment time fell from six months to two weeks, and retraining dropped from 12 months to two weeks, improving the accuracy and confidence of sonar- and imagery-based intelligence.

U.S. Navy

Bobby Kinney

Brigadier General