The oceans hold a vast, ongoing challenge: estimates range from 26 to 66 million tons of waste, with about 94% settled on the seabed. Until recently, cleanup focused mainly on surface litter, while underwater debris demanded divers working in dangerous conditions. Now a European Union funded initiative aims to deploy a fleet of intelligent autonomous robots to tackle debris both on the surface and on the ocean floor.
The SeaClear2.0 project, backed by the EU, introduces a first autonomous robotic and artificial intelligence system capable of locating and removing debris from the seabed. The concept hinges on a coordinated fleet of smart machines that first detect waste and then retrieve it, combining aerial drones, underwater rovers, autonomous surface vessels, and purpose-built robotic grippers.
Robots guided by artificial intelligence can identify, locate, and collect refuse automatically. The team hopes to expand on the earlier SeaClear1.0 system, which successfully passed its tests. SeaClear2.0 is designed to lift heavier items and address debris on both the sea floor and the surface, pushing the technology to deeper waters and greater efficiency.
The core of SeaClear2.0 is a sophisticated robotic setup for automated waste collection that can operate deeper than its predecessor, handle heavier loads, and manage surface litter as well.
The operation is named SeaCat, a seaworthy mothership that coordinates the fleet. SeaCleantech allies refer to it as the central hub for data, power, and control, enabling a seamless flow of tasks to robots deployed across the water column.
SeaClear2.0 will use high-resolution sonar and diverse sensors, along with an intelligent, maneuverable claw capable of picking up large debris such as tires, bicycles, or pipes.
First tests in the Mediterranean
Scientists will design a modular compartment with adjustable chambers to separate different trash fractions, much like drawer dividers, to better handle bulky or heavy waste.
Surface debris will be collected by autonomous mobile equipment capable of working in narrow or shallow areas. With roughly 500 million tons of plastic entering European seas each year, supporters insist that technology must pair with social interventions to reduce both production and release of garbage by citizens.
SeaClear2.0 aims to empower communities through outreach events, geographic storytelling, a gamified garbage reporting app, cleanup activities, exhibitions, competitions, and art installations. In addition, the project will propose improved sorting and recycling approaches and collaborate with local communities to refine policies to close existing gaps.
The new approach will be demonstrated worldwide with three full-scale demonstrations and three pilot tests in the Mediterranean. Five subprojects will explore validating different parts of the system in various regions.
The SeaClear2.0 effort began on January 1 and is scheduled to run for four years. A kickoff meeting took place in Delft, Netherlands, on February 14.
A ‘mothership’, a drone and two ‘turtles’
The process starts with SeaCat, an unmanned surface vehicle that scans the seafloor, creates a 3D bathymetric map, and marks any large debris it detects.
SeaCat acts as the central hub for the fleet: all other robots stage, communicate, and return to it. The mothership also provides power and hosts the computing resources required for detection, control, and AI components.
When conditions allow, an unmanned aerial vehicle, functioning as a flying sensor, surveys for litter in the air and helps identify larger bags that will guide the underwater robot in the next stage.
A compact underwater vehicle, nicknamed miniTurtle, is deployed from SeaCat to investigate the seafloor up close. It uses a forward-looking camera, sonar, and metal detectors to locate smaller pieces of trash and adds these coordinates to the map.
Garbage is identified with artificial intelligence and deep learning object recognition, teaching the system to distinguish litter from marine life so that only relevant debris is collected.
The next steps involve a second marine robot named Turtle, equipped with a clamp and suction device to reach debris in challenging conditions and remove each item with precision.
Budget and collaboration
A budget approach guides route planning and the movement of observation and collection robots, leveraging reinforcement learning and data-driven control methods to optimize performance.
After collection, SeaCat and Turtle deposit debris into a specialized basket designed to interface smoothly with the gripper, preventing released material from drifting back into the water. The basket also communicates to guide the robot to optimal positions for unloading.
The SeaClear team is supported by the Horizon Europe program with a total budget near nine million euros, most of which comes from community funds. A consortium of partners across multiple countries brings expertise in public engagement, policy, sensing and control, AI, marine technology, and recycling processes. SeaClear2.0, which stands for Scalable Full-Cycle Marine Litter Recovery in the Mediterranean and beyond, is planned to conclude in December 2026.
SeaClear2.0 activity will be complemented by the broader SeaClear project portfolio, which shares goals around cleaner seas and smarter waste management. A dedicated project website provides ongoing updates and resources for researchers and communities alike.
The SeaClear initiative envisions a future where advanced robotics and citizen participation converge to reduce ocean litter, protect marine ecosystems, and inform smarter environmental policies. The project continues to explore scalable, bottom-up solutions that can be adapted to other sea basins as part of a global effort to restore cleaner oceans.