Tocaro Blue offers ProteusCore™, a radar processing software platform that applies artificial intelligence and machine learning to enhance the capability of conventional marine radar systems.
Since 2019, Tocaro Blue has been assembling one of the largest proprietary maritime radar datasets. With more than three million frames, the software improves detection, classification, and tracking performance across varied marine environments.
AI Radar Processing, Configured for the Vessel
Marine radar is fundamental to navigation, yet interpretation often depends on operator expertise. ProteusCore embeds trained machine learning models into off-the-shelf radar hardware to deliver automated tuning, adaptive processing, and contextual understanding of radar returns. A configuration application allows users to visualize the machine learning pipeline and adjust radar and NMEA 2000 parameters.
A patent-pending AutoFocus capability automatically optimizes radar gain and range with respect to sea state and proximity to land. By fine-tuning these parameters, ProteusCore supports accurate object detection and classification across the radar field of view and provides real-time perception data for systems integrators and autonomous vessel control systems.
Object Understanding and Multi-Target Tracking
After initial detection, ProteusCore applies Tocaro Blue’s proprietary machine learning models to interpret radar imagery and determine object classes. The software distinguishes among vessels, navigation markers, buoys, land, and wake patterns, supplying navigation and autopilot platforms with context that conventional radar processing does not provide.
The tracking engine automatically acquires, maintains, and removes hundreds of tracks without manual input. For each tracked object, ProteusCore outputs position, velocity, predicted trajectory, and closest point of approach, maintaining a real-time list of nearby objects for both manned and uncrewed vessels.
Enabling Advanced Marine Systems
With a 360-degree field of view, radar is a useful perception sensor for marine advanced driver assistance systems. Developers can use ProteusCore’s classified tracks for collision detection and avoidance, enabling autopilot systems to make steering and throttle adjustments in response to detected targets. Output data can also drive alarms and alerts to support safe operation.
ProteusCore supports camera and radar fusion. Radar-based tracking informs camera pointing and provides distance measurements, improving target tracking and 360-degree awareness when paired with machine vision systems. This approach underpins hybrid perception solutions such as Sea.AI Sentry.
Deployment and Use Cases
ProteusCore runs on standard CPU resources, which can reduce cost relative to solutions that require specialized EO/IR or LiDAR hardware. In many installations, the software operates on existing onboard computers via the ProteusCore App or SDK. Tocaro Blue also supplies dedicated hardware options through the Proteus DevKit for integration and testing.
The technology is used across the maritime sector. In defense, it provides real-time situational awareness tools for tactical craft and uncrewed surface vessels. For OEMs, it supports next-generation user interface design and advanced driver assistance capabilities. For recreational boaters, it enhances radar-based navigation and situational awareness.
By applying machine learning to radar data interpretation, ProteusCore™ provides a mature, software-based path to higher precision and usability in maritime radar perception.




