Plastic pollution in the seas and oceans is a well-known phenomenon. The cause behind this global environmental problem is also very well known: our excessive consumption of plastic, which is processed into almost everything. However, much less is known about the exact distribution, varieties, quantities and specific sources of all that 'marine' plastic waste. VITO is helping to close this information gap, for example by monitoring waste from the air and from space, but also from bridges and river banks.
Today, marine plastic waste is still largely monitored by local sampling, for example by casting nets from ships and seeing what comes to the surface. Or by studying plants and animals in the sea (or sea birds) to see how many and which plastics have entered the food chain.
Remote sensing, with images from cameras fitted in satellites, planes or drones but also in vessels or buoys or in devices on land, can take that monitoring to a higher level. Not only to estimate the bulk quantities of marine plastic waste, but also, for example, to recognise the types of plastics. "This is possible thanks to spectral analysis, whereby we examine plastics through different parts of the sunlight spectrum," says Els Knaeps of VITO. "With infrared, for example, different types of plastic show a unique fingerprint that we can detect with our spectrometers." This technology is already used in plastic recycling (on land), where cameras and sensors help sort plastic waste. However, the distance to the plastics is usually small and there isn't usually (sea) water in between or on top of them. “With plastics either floating on the water's surface or just below, it is more difficult to recognise the different types by spectral analysis."
Chosen techniques, with AI
Exactly how difficult this is, and what can be done about it, is what VITO has been researched in the HYPER project. Experiments were carried out to collect spectral data from a whole range of plastics, both new ('virgin') and weathered plastics, in various conditions: dry, wet, immersed, covered with algae, etc. All these data were gathered in a unique database that VITO made publicly available.
The experiments took place in a large water tank at the Hydraulics Research Laboratory in Antwerp. Cameras and sensors were installed to examine the influence of various parameters (the turbidity of the water, the intensity of light in the tank, the types of plastic waste and the depth at which it floats in the water). The research is part of the European Horizon 2020 programme ATTRACT. Knaeps: "With the results of these experiments in a controlled environment, we aim to select the best camera and sensor technologies for the detection of marine (macro)plastic debris in the near future." The chosen technologies may also contain some artificial intelligence (AI), so that the plastic waste detection and recognition can be done independently and possibly also enable the monitoring of large areas of water in the future.
This monitoring will not only take place on the high seas. "Most plastic waste ends up in the sea via rivers," says Knaeps. "By placing cameras on bridges, for example, we can spot the waste in time to remove it. That is much more difficult once it reaches the sea." For example, the plastic waste can be intercepted by an unmanned river drone that locates and fishes it out based on the data analysis from the camera images. Other remote-sensing technologies such as drones could be used to monitor wide stretches of river, for example in Southeast Asia where a lot of plastic waste flows into the ocean on major rivers.
In Vietnam, VITO has been participating in the AIDMAP project since September 2020. It experiments with drones to detect, recognise and quantify marine plastic waste - using AI technologies such as deep learning. The project is supported by the European Space Agency (ESA), as the drone images can also serve as a monitoring tool for high-resolution images taken by satellites.
Democases in the Scheldt
Smart processing of remote-sensing data is still in its infancy. Many steps still need to be taken before it can be rolled out in concrete applications. These applications will be developed according to the specific circumstances and problems. In the Scheldt at Antwerp, for example, it may be useful to automatically detect waste larger than, say, 20 centimetres so that it does not obstruct navigation. In addition, more and more companies that are at the source of the plastic waste stream (for example, because they produce packaging materials) want to know whether and to what extent their plastics end up in rivers and the sea.
But companies are also joining forces to tackle plastic pollution, although much more needs to be known about the flow of plastic waste from rivers into the sea. After all, it is extremely complex: plastic moves differently according to size, shape, composition, state of weathering, whether algae grow on it, water temperature, salinity and depth. Anyone who wants to tackle the problem must therefore take all these factors into account. That is precisely the aim of the PLUXIN project, a collaboration between knowledge institutions such as the Flemish Marine Institute (Vlaams Instituut Voor de Zee - VLIZ), which is leading the project, and 13 companies from the so-called Blue Cluster. The companies include large players such as Colruyt Group, which conducts research into aquaculture in the North Sea, but also small innovative companies such as Xenics from Leuven, which develops and improves infrared cameras to detect plastic waste.
Within PLUXIN, VITO is jointly responsible for the project part on remote sensing, in which technology is tested in several demo cases, including in the Scheldt basin. Cooperation with companies, often from very different sectors, is important. "We learn from each other what the technology's possibilities are, and how we can support each other in choosing the right techniques and applying them," says Knaeps.
The fact that the VITO Remote Sensing unit has become so involved in recent years in projects to help combat plastic pollution is a result of its rich experience and broad expertise in monitoring water quality, typically combining data from different sources. The data are also processed in various ways, such as in the MAPEO Water platform, which is currently also being used (with fixed cameras and drones) to map plastic pollution in rivers and coastal areas.
Other VITO units are also closely involved in the plastics issue. The Sustainable Materials unit is looking at how more plastics can be better recycled or reused. And the Sustainable Chemistry unit is investigating how microscopic plastic particles (microplastics) in water can be detected more easily. And then, of course, there are the possible harmful effects of macro and (especially) microplastics on human health and the environment. This is what the researchers of the VITO Health unit are looking into.
More info
els.knaeps@vito.be
Image: VIETNAM VIGAC
