New imaging system sees through murky waters

2026-06-11 · Source: MIT News - Computer vision · Field: Technology & Digital — Robotics & Autonomous Systems, Artificial Intelligence & Machine Learning · Depth: Advanced, medium

Summary

The "Sonar-MASt3R" system, developed by engineers at MIT and the Woods Hole Oceanographic Institution (WHOI), is a new underwater mapping technique that fuses visual images from optical cameras with acoustic data from sonar sensors. Presented on June 11, 2026, at the IEEE International Conference on Robotics and Automation (ICRA), this method creates real-time 3D maps even in murky, low-visibility waters. It builds upon the MASt3R image matching algorithm, using sonar data to provide absolute scale measurements and correct the optical system's relative depth estimations. Tank experiments demonstrated Sonar-MASt3R's capability to map environments and resolve centimeter-scale details of objects in the cloudiest conditions, where cameras alone could not see. This innovation aims to guide remotely operated underwater vehicles for applications like scientific exploration, underwater construction, and deep-sea recovery, including the safe removal of unexploded underwater mines.

Key takeaway

For Robotics Engineers developing or operating remotely operated underwater vehicles (ROVs) in challenging, low-visibility environments, Sonar-MASt3R changes your operational capabilities. You can now achieve real-time, centimeter-scale 3D mapping even when sediment clouds optical cameras. This allows your ROVs to safely navigate and perform detailed inspections or recovery missions in previously untractable murky conditions, significantly expanding mission scope and safety. Consider integrating opti-acoustic fusion techniques to enhance your vehicle's perception in turbid waters.

Key insights

Opti-acoustic fusion enables real-time, high-resolution 3D underwater mapping in turbid conditions by combining sonar's scale with optical detail.

Principles

• Combine sonar for scale, optics for detail.
• Opti-acoustic fusion overcomes visibility limits.
• Real-time 3D mapping is achievable in murk.

Method

Sonar-MASt3R uses sonar data to correct the scale of MASt3R's real-time 3D maps generated from 2D optical images. A keyframe approach quickly updates the map with relevant visual detail.

In practice

• Guide ROVs in low-visibility environments.
• Map seafloor for construction/maintenance.
• Safely recover underwater ordnance.

Topics

• Underwater Robotics
• Opti-Acoustic Fusion
• Sonar Mapping
• 3D Environmental Mapping
• Remotely Operated Vehicles
• Turbid Water Navigation

Best for: Computer Vision Engineer, Robotics Engineer, Research Scientist, AI Scientist

Related on AIssential

• Human-machine teaming dives underwater — Combining human and robot strengths is crucial for effective underwater human-robot teaming in challenging environments.
• ATN3D: Density-Aware LiDAR-Radar Early 3D Object Detection Under Extreme Sparsity — ATN3D improves long-range 3D object detection in sparse LiDAR-Radar data by integrating density-aware fusion and range-specific optimization.
• UMI-3D: Extending Universal Manipulation Interface from Vision-Limited to 3D Spatial Perception — Integrating LiDAR into a wrist-mounted manipulation interface significantly enhances data quality and task reliability for embodied robot learning.
• Decart’s new world model can simulate hours of photorealistic driving — with some caveats — Decart's Oasis 3 offers real-time, photorealistic world simulation via API, targeting autonomous vehicle development despite current physics and consistency limitations.
• Bubble Robotics emerges from stealth with $5M to scale autonomous ocean robotics — Autonomous ocean robotics offer continuous, cost-effective offshore operations by eliminating vessel dependency.
• RADIO-ViPE: Online Tightly Coupled Multi-Modal Fusion for Open-Vocabulary Semantic SLAM in Dynamic Environments — RADIO-ViPE enables robust, calibration-free, open-vocabulary semantic SLAM in dynamic environments using monocular RGB video.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by MIT News - Computer vision.