University of Plymouth and MSEIS LIMITED KTP 24_25 R5

A new camera system will use artificial intelligence to spot whales, seals, and floating debris from ships in real time, fusing thermal and optical video feeds to detect hazards that human lookouts often miss. Current offshore monitoring relies on crewed watches or basic radar, which struggle to identify marine mammals in rough seas, fog, or at night. This blind spot forces vessels to slow down or alter course unnecessarily, wasting fuel and time, while also risking collisions that can injure animals or damage propellers. The project combines a dual-sensor camera—one thermal, one visible-light—with machine learning software that processes both streams simultaneously, distinguishing a porpoise from a wave or a shipping container from a log. If the system works, it could be deployed on survey vessels, offshore wind farm support boats, and cargo ships. The immediate impact would be safer navigation and fewer costly delays. For marine industries, it means automated compliance with environmental regulations without relying on spotters. For conservation, it offers a continuous, data-rich record of animal presence in busy waters—information that currently exists only in patchy sighting logs. The technology is applied engineering, not fundamental science, but its success depends on solving a hard real-world problem: teaching a computer to recognise a living creature in a chaotic, moving seascape.