Havelsan launches BlueVision AI maritime surveillance system for military and commercial ships.

Turkish defense company Havelsan introduced the BlueVision artificial intelligence-powered maritime surveillance and perception system to automate object tracking and improve close-range tactical situational awareness for naval and commercial vessels. The system fuses 4K ultra-high-definition daylight imagery, thermal sensors, and Automatic Identification System (AIS) transmissions to convert raw video feeds into continuous, structured track data without requiring manual target selection. This software-driven processing layer is deployed on the Turkish Navy Sancar Armed Unmanned Surface Vehicle to deliver autonomous navigation, real-time collision-risk evaluation, and passive threat detection.

The BlueVision system features a 5.1-kilogram salt-mist resistant optical unit housing a 30 Hz 4K daylight camera with a 98-degree field of view and a 25 Hz thermal imager. Integrated with Havelsan Advent Combat Management System, the platform successfully completed collision-avoidance and search-and-rescue test activities under the coordination of the Turkish Presidency of Defence Industries.

Related topic: How Havelsan solves multi-domain data overload with AI-powered command architecture

The BlueVision is an AI-powered digital lookout that automatically detects, identifies, and tracks everything around a ship to improve navigation and maritime safety. (Picture source: Havelsan)

On July 3, 2026, the Turkish defense company Havelsan introduced the BlueVision, an artificial intelligence-powered maritime surveillance and perception system designed to convert live electro-optical imagery into continuously updated object tracks for navigation, autonomous operations, search and rescue, and maritime security missions. The system combines a 4K UHD daylight camera, a thermal imager, AIS data, navigation inputs, and optional radar fusion to create a single maritime picture in which visible contacts are detected, classified, tracked, and assessed for risk without requiring manual target selection.

The BlueVision is already installed aboard the Turkish Navy’s Sancar Armed Unmanned Surface Vehicle, which entered service with the Turkish Naval Forces in 2026, and is integrated with Havelsan’s Advent Combat Management System. In that configuration, it supports autonomous navigation and mission execution by providing structured object data rather than a simple video stream, including real-time detection, target tracking, collision-risk evaluation, and search-and-rescue support. The BlueVision’s baseline sensor package is built around a 4K UHD visible camera operating at 30 Hz with a 98° field of view and a thermal camera operating at 640 × 512 resolution, 25 Hz, and a 69° field of view. This gives the system a broad daylight observation sector while preserving night and low-visibility detection through thermal imaging.

The camera unit weighs 5.1 kg, equal to 11.3 lb, and measures 314 × 233 × 124 mm, making it compact enough for USVs, patrol craft, Rafnar boats, commercial container vessels and other vessels where topside weight and installation volume matter. The optical unit is salt-mist resistant and complies with IP67, IK10, MIL-STD-810G and ONVIF S/G/T standards, which are important for maritime use because the sensor must survive water ingress, impact, vibration, temperature variation and integration into existing video networks. The system also includes an AI Box processing unit and a portable multi-touch panel display unit, while external camera integration can be added when a vessel already has usable shipboard cameras. 

The difference between the BlueVision and a conventional electro-optical camera is the processing layer behind the image. A normal EO system mainly gives operators a video feed, sometimes with manual zoom, stabilization, and a single target-tracking function. The BlueVision continuously performs object detection, classification, optical angle detection, distance estimation in nautical miles or yards, categorical distance assessment across close, mid, and far ranges, coordinate estimation, real-time two-dimensional obstacle mapping, processed video streaming and recording, and collision-risk assessment. Each visible object inside the camera field of view can become a separate track with position and risk information, instead of remaining an unstructured visual contact that depends entirely on the operator’s attention.

This matters during crowded navigation, where a bridge crew may simultaneously face merchant ships, small craft, buoys, drifting debris and uncooperative contacts, while still having to monitor radar, AIS, ECDIS, navigation plans and voice communications. The system is aimed at contacts that radar and AIS do not always handle well. Radar remains necessary for longer-range detection and navigation beyond the camera field of view, and AIS remains useful for cooperative vessel identification, but neither provides a complete maritime picture. Small boats, low-profile craft, buoys, floating objects, and persons in the water can be missed or misinterpreted, especially in congested waters, demanding weather, night conditions, or high sea clutter. The BlueVision uses daylight and thermal video analysis to detect these contacts visually, classify them, estimate their coordinates, and calculate whether their movement creates a collision risk.

When the risk threshold is reached, it can generate visual and audible alerts, allowing crews to focus on decisions rather than continuous screen scanning. For search and rescue, the same function is relevant because a person in the water can occupy only a small portion of a video image, while thermal detection and automated tracking can reduce the risk of losing the contact during prolonged observation. The system’s military utility includes dark-vessel detection, AIS inconsistency detection, passive close-range surveillance and support to mission-specific maritime security tasks. A vessel operating without AIS can still be detected and tracked visually, creating a warning that a real contact exists without a matching cooperative transmission.

The BlueVision can also compare what the camera sees with AIS information, helping operators identify mismatches between transmitted identity and observed characteristics, including cases where a vessel’s declared type, position, or behavior does not fit the visual picture. This is relevant for illegal fishing, smuggling, sanctions evasion, infiltration, gray-zone maritime activity, and coastal-security missions. Because daylight and thermal cameras are passive sensors, the BlueVision can also support navigation and surveillance when radar transmissions are restricted under emission-control conditions. The system does not replace radar for long-range surveillance, but it gives vessels an additional passive perception layer for close-range contacts, hazards, and suspicious maritime activity. 

The Sancar is the clearest operational case because unmanned surface vessels need perception data that software can use directly. Integrated with the Advent aboard the Sancar Armed USV, the BlueVision supplies classified object tracks, estimated positions, movement data, and risk assessments to support autonomous navigation and mission execution. Under the Turkish Presidency of Defence Industries (SSB) program, the system completed collision-avoidance and search-and-rescue test activities linked to the Sancar. For an armed USV, that capability is operationally important because the vessel may patrol near naval bases, ports, coastal waters, or dense maritime traffic while encountering merchant vessels, patrol craft, RHIBs, buoys, floating debris, and persons in the water.

A remote operator cannot be expected to interpret every video frame during long-duration missions, and autonomous navigation software cannot safely maneuver on raw imagery alone. The BlueVision addresses that gap by converting camera imagery into structured, updated tracks that can be used for route adjustment, obstacle avoidance, and mission awareness. The BlueVision is also designed for broader integration beyond one naval unmanned vessel. It has already been integrated into Rafnar boats through a cooperation with VN Maritime and Piloda Shipyard, and it is also used in commercial maritime applications, including container vessels operating in difficult navigation environments.

The product family includes dedicated versions for commercial vessels, yachts, autonomous surface vessels and defense vessels. For cargo, tanker and container ships, the emphasis is advanced navigation safety during port approaches, straits, anchorages and high-traffic sea lanes. For yachts, the focus is compact situational awareness. For autonomous surface vessels, the priority is perception for unmanned navigation. For defense vessels, the system adds ISR, coastal security, search and rescue, asymmetric-threat awareness, dark-vessel detection and spoofing detection when combined with multi-sensor inputs. This segmentation matters because the same core architecture can be adapted to civilian safety, naval surveillance and autonomy without changing the basic optical suite. 

The software-defined architecture is central to the system’s growth path. The BlueVision can add radar integration, AIS integration, route planning and monitoring, autonomous navigation and remote-control support, external camera integration, ISR functions, search-and-rescue performance improvements for night and harsh weather, and mission-specific detection models. Its adaptable object-detection model means the system can be updated as new operational needs emerge, such as detecting different small-craft profiles, floating hazards, coastal objects, or maritime threat patterns.

The result is a sensor-processing layer that sits between cameras and command systems, turning visual information into tracks that can be displayed to operators, passed to navigation systems, or used by combat management and autonomy functions. Operationally, the main significance is not that the BlueVision adds another camera to a ship, but that it shifts electro-optical surveillance from manual observation to continuous AI-assisted perception, with every detectable object in view treated as data for navigation safety, mission management, and maritime security.

Written by Jérôme Brahy

Jérôme Brahy is a defense analyst and documentalist at Army Recognition. He specializes in naval modernization, aviation, drones, armored vehicles, and artillery, with a focus on strategic developments in the United States, China, Ukraine, Russia, Türkiye, and Belgium. His analyses go beyond the facts, providing context, identifying key actors, and explaining why defense news matters on a global scale.

Explore More Defense News

• Land Defense News
• Naval Defense News
• Defense Aerospace News