How New Tech Lets Military Underwater Robots Talk Securely Across the Deep

Keerthana S December 18, 2025 | 02:52 PM Technology

The Israeli startup says its SeaSphere platform allows fleets of unmanned underwater vehicles to share data securely and adapt missions in real time. As maritime security concerns grow and underwater infrastructure becomes increasingly strategic, an Israeli robotics startup claims it has overcome one of the toughest challenges beneath the sea: reliable long-distance communication without surfacing.

Autonomous underwater vehicles typically struggle to exchange information while submerged. Surfacing to transmit data exposes them to detection, a serious risk in contested or sensitive waters. The company now says it has developed a software-based solution that enables underwater robot fleets to communicate securely while remaining underwater, using artificial intelligence designed for reliability rather than spectacle.

Figure 1. Military Underwater Robot Fleets.

A long-standing communication challenge

Unmanned underwater vessels are playing a growing role in defense, surveillance, and the protection of critical infrastructure such as pipelines and subsea cables. Yet communication has remained a persistent bottleneck. Radio signals are ineffective underwater, while acoustic communication is slow, limited in range, and prone to interference.

As a result, many systems still rely on surfacing to relay information, creating moments of vulnerability. The startup says its latest update to the SeaSphere fleet management platform eliminates that need by enabling submerged, long-range data exchange. Figure 1 shows Military Underwater Robot Fleets.

SeaSphere allows groups of autonomous vessels to share sensor data, interpret it collectively, and respond in near real time. Each robot can adjust its behavior based on what the rest of the fleet detects, while remaining aligned with overall mission objectives.

Coordinated fleets without constant human oversight

According to the company, the breakthrough goes beyond communication. The system enables decentralized decision-making across an entire fleet. When one vehicle detects an obstacle, threat, or anomaly, that information is instantly shared, allowing others to reroute, slow down, or take on new tasks autonomously.

“Communication between vessels is one of the biggest challenges in multi-domain, multi-vessel operations,” said co-founder and CEO Idan Levy. “Our goal is to enable hundreds of unmanned vessels to operate together, share data, and coordinate both on the surface and underwater.”

While designed with military missions in mind, the company says the technology is also suited to civilian uses such as port security, undersea cable protection, and safeguarding maritime supply routes [1]. Reducing the need for constant human control allows fleets to respond faster in environments where real-time oversight is not feasible.

Why the company avoided cutting-edge AI trends

The AI system behind SeaSphere was developed under the guidance of Teddy Lazebnik, an AI researcher and professor at the University of Haifa. Rather than relying on modern large language models or deep neural networks, the team chose older, mathematically grounded algorithms.

“Newer algorithms can be very powerful, but they are also less predictable,” Lazebnik explained. “By sacrificing some performance or ‘wow factor,’ we gain explainability, predictability, and robustness.”For defense and safety-critical operations, the startup argues that transparency and reliability are more important than raw performance, especially when managing large autonomous fleets.

From stealth to deployment

Founded in 2024, Skana Robotics emerged from stealth earlier this year and is now pursuing government and enterprise customers across Europe. The company sees growing demand driven by geopolitical tensions and heightened concern over underwater sabotage.

Levy confirmed the startup is already in talks with a major government agency for a significant contract and hopes to finalize an agreement before the end of the year. A commercial release of the software is planned for 2026, with large-scale trials scheduled ahead of launch.

References:
  1. https://interestingengineering.com/ai-robotics/underwater-robot-fleets-gain-secure-communication-boost
Cite this article:

Keerthana S (2025), How New Tech Lets Military Underwater Robots Talk Securely Across the Deep, AnaTechMaz, pp.331

Recent Post

Blog Archive