Lur

Lur is a passive sonar analytics package built by Helsing , the Munich-based defence software firm, for use aboard submarines and other naval platforms. Announced in 2024, it represents the company’s first move out of the land and air domains it had previously focused on and into undersea warfare, a field where AI-driven signal processing has been a long-standing aspiration but a slow-moving practical reality. The system is built around a partnership with ThyssenKrupp Marine Systems (TKMS), Germany’s principal submarine builder, and is positioned for integration into TKMS-built boats in service with the Deutsche Marine and export customers.

Technically, Lur runs as a software layer that ingests raw acoustic data from a host platform’s existing sonar arrays and applies machine-learning classifiers to detect, identify and track underwater contacts in real time. The pitch is that contemporary submarines generate vastly more acoustic data than human sonar operators can attend to in the watchroom, and that vessel-class fingerprints, propeller cavitation patterns and other distinctive features can be learned from large libraries of recorded signatures. By doing this classification on board rather than relying on shore-based post-mission analysis, Lur is meant to compress the time between a sound being picked up and a contact being called. The system is designed to run on standard naval computing hardware rather than dedicated sensor-vendor stacks, which is the more general Helsing pattern of treating the sensor as a commodity and the analytics layer as the differentiator.

Germany is the launch customer through the TKMS relationship. The most likely platform path is the Type 212CD, the next-generation common design being built jointly for the German and Norwegian navies, with delivery of the first boats expected later in the decade. Beyond Germany, TKMS exports its submarines widely — to Norway, Italy, Israel, Singapore, South Korea and others — and Lur is positioned to follow that footprint where customers want it. No deployed counts or contract values have been disclosed publicly, and the system has not been combat-tested.

Lur sits at the front of a broader shift in undersea warfare toward software-defined sensing, where firms outside the traditional sonar establishment are competing to own the classification layer. It complements Helsing’s existing work on the Altra electronic-warfare suite for fighters and the HX-2 strike drone, extending the same idea — that the value lives in the algorithms, not the box — into the maritime domain. Whether the system delivers on its promise will depend on access to enough real-world acoustic training data, a resource that submarine operators have historically guarded closely.