Reduce the time, number of people, & resources required for critical underwater operations.
Defence
Current practice in many MCM missions follows a two-stage process: an AUV conducts a wide-area sonar run to detect contacts, followed by separate deployments of USVs and ROVs to reacquire and visually classify them. This requires multiple deployments and sustained operator involvement.
Sonar ATR detects a contact and generates an event. Optical data is then required for visual classification, but images are typically reviewed after recovery, leaving teams to sort through large volumes of frames that are not tied to the original contact.
Cathx is working with the world’s leading vehicle manufacturers to bring detection and classification onto the AUV itself, allowing contacts to be identified and visually classified within the same workflow.
Detection & classification done during the mission. Not after.
Cathx is developing the Clarity MCM software module for AUVs supporting mine countermeasure missions and other critical underwater operations.
Clarity runs onboard the vehicle itself, processing sensor data in real time while executing machine learning workflows at the point of acquisition.
Instead of simply collecting raw data for post-mission review, Clarity extracts critical information from multiple sensor inputs and feeds it back to the vehicle during the mission, enabling it to respond dynamically to objects of interest – whether mines, infrastructure, or other seabed anomalies.
Confidence scores for object classification are generated onboard and remain tied to the original contact, rather than being separated across large volumes of imagery only to be reviewed once the vehicle has been recovered.
What this means for vehicle manufacturers
Clarity runs detection and classification models onboard the AUV during data acquisition, with outputs delivered directly into the vehicle’s autonomy stack. Classification scores are generated in-mission, with flagged events made available through the Clarity API.
For vehicle manufacturers, Clarity provides a way to advance autonomy development by rapidly integrating threat perception and classification capabilities without the need to build them from scratch. Vehicles can be delivered with these capabilities already embedded in the mission workflow, without reliance on separate reacquisition or post-mission processing.
Accelerating your delivery of these capabilities to end users enables faster time to market, at lower cost, with greater confidence & reduced risk.
What this means for end users
Executing both detection and classification within a single AUV mission represents a fundamental advance in MCM capability. Eliminating the need to recover survey data for shore-based processing, redeploy additional vehicles, and revisit targets streamlines the workflow into a single, continuous operation.
This approach reduces vessel and crew requirements, compresses mission timelines, and accelerates decision-making in contested environments.
While an operator-in-the-loop remains essential for final decision authority, single-mission execution ensures rapid delivery of high quality sonar & optical data, supporting confident threat classification and timely decision-making for navies conducting critical underwater operations.
(Above) Visual demonstration of our object detection reference ATR being executed on sonar data processed by Clarity.
Our approach is sensor-agnostic, meaning Clarity can work with several sensor inputs, including optical (images & laser) and sonar (SAS & MBES).
Solutions
Sensors
Capture ultra-clear, precision data in the harshest shallow or deep water conditions, with our subsea cameras and lasers. Provides the clarity needed for reliable processing and insight generation.
Edge Hardware
Deploy powerful data processing directly on your vehicle or sensor, enabling AI model inference without the need to wait for onshore processing.
Edge Workflow Software
Orchestrate, deploy, and manage microservices & AI models (third-party or your own) for automated event detection and actionable insights at the edge.
Edge Workflow Software
Orchestrate, deploy, and manage microservices & AI models (including third-party or your own) for automated event detection and actionable insights at the edge.
Edge Workflow Software
Orchestrate, deploy, and manage microservices & AI models (including third-party or your own) for automated event detection and actionable insights at the edge.
Interested in learning more?
Please get in touch below.