/ 
Collisions between ships and large cetaceans are a known risk in certain shipping areas. For shipowners, this is both an environmental and an operational issue. A recent solution combines thermal imaging and artificial intelligence to detect whales on the surface and warn crews in good time.
Whale detection, a long-standing challenge for navigation
On shipping lanes frequented by cargo ships, ferries or industrial fishing vessels, the presence of whales represents a danger that is difficult to anticipate. Cetaceans spend most of their time underwater, appearing only briefly on the surface.
Even with a careful watch on the bridge and the use of binoculars, detection remains uncertain. The situation becomes even more complicated at night, in heavy seas or fog. These conditions largely explain the collisions observed in many parts of the world.
Some populations are particularly vulnerable. The North Atlantic right whale is a case in point. There are around 70 breeding females left. In this context, each collision can have a major impact on the future of the species.
Thermal imaging to spot marine mammals
The WhaleSpotter system is based on thermal infrared imaging. The principle is simple. A camera detects temperature differences between the water and the surface of the animal's body as it breathes or emerges. The solution uses a Boson+ thermal module. This camera captures thermal signatures even at night or in light fog. Algorithms then analyze the images to identify the characteristic shapes of a whale's breath or back.
Under the right conditions, the system signals the presence of an animal up to around 7 kilometers away. This distance corresponds to what a human observer could distinguish with binoculars in broad daylight. For the crew, a few extra miles may be enough to change course or reduce speed. On a container ship or bulk carrier, this room for maneuver becomes decisive.
Artificial intelligence to limit false alarms
Automated detection poses a classic problem. Sensors often generate numerous signals that do not correspond to cetaceans. Waves, birds or floating objects can mislead the algorithms. The WhaleSpotter system therefore relies on artificial intelligence trained to recognize the specific thermal signatures of whales. The aim is to reduce the number of false positives that tire crews.
Alerts are also verified in real time by a network of onshore marine experts. This double validation ensures an advertised efficiency rate of 99%. For a captain, the question is simple. A reliable alert enables immediate action to be taken, without the need for multiple checks.
Deployment already underway in commercial fleets
The technology is the result of over 10 years' research at the Woods Hole Oceanographic Institute. The aim was to solve the problem of "invisible whales", i.e. animals that are difficult to detect from the bridge. After several test campaigns, the solution has recorded over 250,000 confirmed detections. Some 100 systems are currently installed at over 50 sites worldwide. Some shipowners have already integrated this tool into their operations. Such is the case of Matson, a company specializing in freight transport in the Pacific.
For the time being, these technologies are intended for commercial vessels and offshore installations. But their principle could, in time, be of interest to other navigation segments. In particular, offshore yachting, where visual surveillance remains the only defence against large cetaceans.
