Overview
- An August study led by Meng Hao at the China Helicopter Research and Development Institute describes an AI-driven anti-submarine system that fuses sonar, radar, magnetic readings and oceanographic data to detect and track targets.
- In the team’s simulations, the system achieved about 95 percent success and cut the modeled chance of a submarine escaping to roughly 5 percent.
- The researchers say the AI functions like an autonomous commander, reconfiguring sensors and adapting to zigzags, silence runs, decoys and drone distractions in real time.
- The project reports large‑language‑model interfaces for operators and envisions integration with aerial drones, surface ships and unmanned underwater vehicles to form a three‑dimensional hunting network.
- A separate December paper from Wang Honglei’s group models persistent magnetic “Kelvin wakes” from submarines, claiming signals near 10^-12 tesla that airborne magnetometers could detect.