Tocaro Blue’s ProteusCore™ radar processing software provides an advanced framework for extracting greater value from conventional marine radar systems.
The platform applies machine learning techniques to raw radar returns to deliver enhanced detection, classification, and situational awareness across both crewed and uncrewed maritime applications.
Since 2019, Tocaro Blue’s engineering team has built and refined one of the largest proprietary marine radar training datasets, containing more than three million radar frames. This resource underpins the machine learning models integrated within ProteusCore. By processing radar returns through its AI pipeline, the software identifies and classifies objects in real time. A dedicated configuration application enables developers to intelligently leverage data from their radar for use in autonomous control and situational awareness applications.
Automatic Gain and Range Optimization
Accurate radar interpretation relies on precisely balanced gain and range settings. ProteusCore incorporates a patent-pending AutoFocus function that automatically adapts these parameters to current conditions, including sea state, weather, and proximity to land. The result is consistent signal quality and improved object recognition without the need for manual tuning.
Contextual Object Classification
Following detection, ProteusCore applies Tocaro Blue’s proprietary algorithms to interpret radar returns and determine object type. The system distinguishes between vessels, buoys, navigation markers, shorelines, and wakes, providing operators and control systems with actionable context. This capability supports safer navigation and more informed autonomous decision-making.
Devkit Hardware from Tocaro Blue.
Multi-Target Tracking and Trajectory Prediction
ProteusCore is capable of tracking hundreds of objects, depending on configuration, automatically acquiring and releasing targets as they enter or leave range. For each tracked item, the software calculates position, velocity, closest point of approach, and predicted trajectory.
These continuous outputs can be integrated with autopilot systems or displayed within onboard navigation interfaces, ensuring real-time awareness of surrounding traffic and obstacles.
Enabling Advanced Driver Assistance
Radar’s 360-degree coverage makes it a valuable perception sensor for maritime advanced driver assistance systems (ADAS). Using data processed by ProteusCore, developers can implement collision avoidance, path planning, and automated control responses such as throttle and steering adjustments. The same functionality provides situational alerts for crewed vessels, enhancing operational safety.
Sensor Fusion for Extended Awareness
ProteusCore can be paired with onboard camera systems to deliver radar-camera fusion, improving range accuracy and object tracking stability. The radar data informs camera orientation and focus, extending visual coverage and maintaining situational awareness even in low-visibility environments.
Streamlined Deployment and Integration
Designed for straightforward implementation, ProteusCore runs entirely on CPU resources, removing the need for specialized processing hardware. The software can be installed on existing platforms via the ProteusCore Executable App, while dedicated Devkit hardware is available for sea trialing.
Broad Maritime Applications
ProteusCore technology is employed across a wide range of maritime sectors. In defense, it supports real-time situational awareness for tactical craft and uncrewed surface vessels (USVs). For OEM developers, it enables machine-learning-based radar processing for next-generation navigation and display systems. In recreational and commercial environments, the same core functionality enhances radar interpretation and improves safety for vessel operators.
