At Ubotica, we’re tackling one of the most pressing challenges in maritime security today—the detection of vessels that deliberately operate “off the grid.” Our breakthrough approach fuses live Earth observations from satellites with AIS data from ships to transform how we monitor and protect our critical maritime infrastructure and ensure compliance with international regulations.
The Dark Vessel Challenge
The Automatic Identification System (AIS) forms the digital backbone of modern maritime operations. AIS-equipped ships broadcast their identity, position, course, speed and navigational status in real time via VHF radio technology. This creates a vital traffic awareness network that functions in all weather conditions, enhancing safety and enabling efficient vessel management. However, this system has a critical vulnerability that threatens maritime security globally.
“Dark vessels” are ships that avoid tracking by either completely disabling their AIS transponders or engaging in “spoofing”—falsifying their data to misrepresent their identity, location, or activities. While non-broadcasting simply makes a vessel invisible to tracking systems, spoofing creates deceptive digital footprints showing vessels in false locations or with false identities. These practices are frequently associated with illegal activities including unauthorised fishing, sanctions evasion, smuggling, human trafficking, and potential threats to critical infrastructure.
Dark Vessels present a significant blind spot for protecting vital maritime assets. Infrastructure companies typically invest in AIS vessel tracking services to monitor traffic near valuable underwater assets such as subsea telecommunications cables, offshore wind farms, and oil and gas infrastructure. Yet these systems are all blind to Dark Vessels.
The security implications are substantial and far-reaching. With approximately 95% of international data traffic transmitted through subsea cables that are largely unprotected against vessels operating without proper identification, the risks to global communications networks cannot be overstated. Intelligence agencies and security experts have documented increased instances of vessels conducting seafloor mapping operations near critical cable routes connecting Europe, North America and other strategic regions—activities potentially aimed at identifying vulnerabilities in vital communications networks.
SPACE:AI: An Open Integrated Platform for Live Earth Intelligence
At Ubotica, we’re revolutionising maritime surveillance through our SPACE:AI platform—an open integrated system that unlocks true Live Earth Intelligence by seamlessly addressing these challenges through processing both in orbit and on Earth:
Onboard Intelligence
The first critical step occurs directly onboard satellites in orbit, where our SPACE:AI deploys custom AI models to identify and classify vessels from optical imagery in real-time. Ubotica’s flight-proven vessel detection AI model, operating with over 90% accuracy, analyzes imagery immediately after acquisition. Within just 90 seconds, it processes a typical 800km² image, generating a compact detection packet that includes precise vessel locations, absolute sizes, and headings. This rapid onboard processing ensures timely and efficient maritime surveillance, minimizing data transmission needs while delivering actionable intelligence in near real-time.
This edge-computing approach represents a fundamental shift from traditional Earth observation methods. Instead of downloading enormous volumes of raw image data for ground processing—a method that introduces significant delays and bandwidth constraints—our system performs the computationally intensive analysis directly in space.
The result is immediate vessel detection and classification without the need to transmit massive image files. This dramatically reduces latency from days to moments while simultaneously improving efficiency by eliminating the transmission of extraneous data.
Intelligent Data Fusion
Onboard real-time insights are instantly relayed to ground where they are fused with AIS broadcast data to build a comprehensive maritime picture. This integration reveals three distinct vessel categories:
- Legitimate vessels: Where visual detection matches corresponding AIS data, confirming proper compliance with maritime regulations
- Dark vessels: Ships that are visually detected by our satellites but are not broadcasting any AIS signal, indicating potential illegal or covert operations
- Spoofing vessels: Cases where AIS signals exist with no corresponding vessel in the satellite imagery, or where the signals appear at incorrect locations, revealing deliberate attempts to deceive monitoring systems
Our integrated platform seamlessly handles computationally intensive image analysis in orbit while performing sophisticated data fusion on the ground, delivering near real-time maritime domain awareness with complete vessel accountability—even when vessels attempt to hide their presence or activities near areas of critical infrastructure, for example, over sub-sea cables.
Real-World Results: Singapore Case Study
Our CogniSat-6 satellite has demonstrated this revolutionary capability in some of the world’s busiest shipping lanes, including Singapore’s congested maritime zones. In a single orbital pass, our satellite identified and processed dozens of vessels simultaneously, distinguishing between those properly broadcasting AIS signals and those operating dark.
The system successfully identified legitimate vessels such as the “INDIAN HARMONY” bulk carrier and “LEO ASPHALT I” tanker—both properly broadcasting their AIS signals—while simultaneously detecting multiple dark vessels attempting to operate without identification in the same region. This real-world demonstration showcased our SPACE:AI technology performing onboard vessel identification from optical data, transmitting this intelligence to ground stations for AIS data fusion, and immediately revealing vessels attempting to operate covertly.
What’s particularly impressive is the scalability of our solution. In high-traffic environments like Singapore, where managing maritime security requires continuous monitoring of hundreds of vessels, our edge-computing approach efficiently processes large volumes of maritime traffic, providing critical awareness of all maritime activity—including ships that would remain invisible to traditional monitoring systems relying solely on AIS data.
In another case, we processed Synthetic Aperture Radar (SAR) observations from the Sentinel-1 satellite in the Irish Sea. When fused with AIS data (the ship’s AIS was primarily inactive during transit, although it was intermittently activated—for example, switched on for approximately six minutes before being turned off again. This intermittent AIS operation makes it difficult to track the vessel without aerial reconnaissance), the imagery clearly revealed the Russian vessel Yantar—known for its seafloor mapping capabilities—being closely monitored by an Irish Navy patrol vessel. SPACE:AI can also extract metadata such as estimated vessel dimensions, direction of travel, and speed. This observation is particularly significant as it demonstrates our system’s ability to track vessels of special interest in European waters, especially those potentially engaged in subsea infrastructure reconnaissance. The integration of SAR data further showcases our platform’s all-weather and day or night monitoring capabilities.
The Future: Comprehensive Maritime Awareness
Our vision extends beyond satellites alone to a comprehensive maritime surveillance ecosystem spanning from seabed to orbit. SPACE:AI is built with an open architecture that incorporates data from a wide range of third-party sensors—from underwater acoustic arrays and surface buoys to coastal radars and space-based systems.
This approach allows maritime security operations to integrate our satellite intelligence with existing sensor networks rather than replacing current investments. Our platform unifies these disparate data sources into a coherent operational picture, harmonising diverse data streams through advanced AI processing to detect and track vessels across all maritime environments, even those attempting to evade traditional monitoring. This open architecture ensures the system evolves alongside emerging sensor technologies, future-proofing against evolving threats.