SMAUG Innovation Update Series: Underwater Detection and Location
Underwater detection and location focus on developing and refining advanced underwater sensing and cooperative technologies for the SMAUG system. These technologies include acoustic detection, short and long-range sonar scanning, visual monitoring through cameras and UAVs, and an underwater robotic swarm for coordinated exploration. Each tool is designed to operate independently to detect, classify, and locate potential threats or objects of interest in the marine environment, while generating structured data that will feed into the SMAUG Command and Control Centre (C2) for real-time situational awareness.
Throughout this work, several technical challenges have emerged due to the demanding nature of underwater environments and the complexity of integrating heterogeneous sensing and robotic technologies.
- Acoustic detection and localisation: While acoustic classification capabilities are performing well, accurately determining the position of detected objects remains challenging. Ports are highly noisy environments with multiple reflection sources, which complicate precise localisation. To address this, we have adopted a practical approach based on estimating the distance to the sound source and refining signal processing pipelines to improve robustness in real-world conditions.
- Sonar-based automatic recognition: Automated interpretation of sonar data is more complex than initially expected, particularly for real-time object detection and classification. Although promising AI proof-of-concept models have been developed, there is still work required in data management, labelling and the implementation of reliable autonomous responses. To mitigate this, we are expanding datasets, improving annotation processes, and iteratively enhancing the architectures to increase detection performance.
- Cooperative robotics and communication: From a swarm perspective, ensuring stable underwater communication between agents remains one of the most demanding challenges. Communication links degrade rapidly underwater, making continuous coordination unreliable. A strategy has been implemented whereby individual robots periodically surface to establish a stable communication channel before diving again to continue operations. Complementary improvements in communication protocols and cooperative behaviours are being explored to further enhance swarm performance.
The most impactful results of the underwater detection and location work stem from the successful development and validation of the core underwater detection and classification technologies. This work has delivered operational prototypes across acoustic sensing, sonar scanning, aerial surveillance and autonomous swarm robotics — all of which are now ready for integration into the SMAUG system and subsequent testing in real use cases. These technological advances collectively demonstrate a major step forward in maritime security and underwater threat detection.
A key contribution is the acoustic kit, which integrates hydrophone signal acquisition with advanced ML/DL classification. Indoor and outdoor tests have validated its capability to detect and classify different vessel types, providing long-range situational awareness to SMAUG partners who are now incorporating this data into the Command-and-Control (C2) environment.
Similarly, the sonar scanning solution has achieved rapid scanning rates (up to 50 m/min) and can detect small objects (≈200 mm). Proof-of-concept AI-based perception, combined with data fusion between sonar and cameras, enhances precision and reduces false positives — outcomes that other partners are already using to define automatic threat response modules.
The aerial surveillance dronebox, brings persistent above-water observation with visual and thermal sensing capabilities. Its robust and weather-resistant design enables direct integration with maritime platforms developed elsewhere in the consortium, supporting coordinated multi-sensor operations.
Finally, our work has shown that autonomous robotic swarm intelligence is viable for underwater security missions, enabling cooperative scanning and wider area coverage. Interoperability tests confirm that swarm data can be shared and exploited by system-level components being implemented by other partners. Now, all these individual results must be integrated into the Smart Integration of Data sources using the broker developed, and the data sent there.
The Future
The work on the underwater detection and location has been successfully completed!
The final phase of the project will focus on the full integration of all developed technologies within the SMAUG system and their deployment in real use-case scenarios. The key outcomes expected include:
- Seamless interoperability between all sensing modules — acoustic, sonar, visual and swarm — through reliable data transmission to the C2.
- Validation of the complete multi-sensor detection chain in operational port environments, demonstrating real-time threat identification, classification and localisation.
- Execution of coordinated missions involving autonomous underwater and aerial assets, confirming their contribution to situational awareness and incident response.
- Achievement of performance targets under realistic maritime conditions, ensuring the system is ready for practical adoption by end-users.
Inner Dialogue with UPM, who is leading the work of the Underwater detection and location
“Our involvement in the SMAUG project has clearly demonstrated that real maritime scenarios are far more complex and unpredictable than initial laboratory assumptions. Underwater environments and specifically ports present significant operational challenges — from acoustic noise to communication constraints and limited visibility — which demand continuous innovation in both sensing technologies and autonomous systems. One of the strongest lessons learned is the crucial value of interdisciplinary collaboration. Progress in maritime security cannot rely solely on advances in a single field; it requires coordinated expertise in robotics, signal processing, AI, communications, and operational deployment. SMAUG has allowed us to build this shared knowledge and test integrated solutions that would be impossible to achieve independently. Looking ahead, we believe the future of maritime security will depend increasingly on intelligent, cooperative systems capable of adapting to evolving threats — and SMAUG is helping pave this path forward.”