value of the data: “Having a system that works on such a large scale gives us a better understanding of where and when we need to take action. And, of course, it gives us more planning security, as we can estimate in advance exactly what logistics we need for our boats and the waste collected. This not only saves time, but also money. At the same time, by evaluating the material we collect, we can validate and verify to the DFKI whether the AI analysis was correct, which in turn strengthens the system’s informative value.“ Common standards for more meaningfulness Projects such as PlasticObs_plus are promising when it comes to providing local environmental projects, governments, or authorities with information on the distribution routes, sources, and accumulation areas. Such information is needed to derive and implement measures to prevent the spread or even entry of waste into the environment. “Together with weather data or information on water flow velocities, we can identify entry paths and make predictions about the spatial and temporal distribution of plastics or observe seasonal differences. This methodology would also be useful for monitoring bans, as regular monitoring would provide information on whether, for example, the actual number of bags floating in the water has declined after the introduction of a plastic bag ban,” says Floehr, describing some of the possible uses. In order to make possible applications and data generated by remote sensing for monitoring plastics in the environment as usable as possible across regions, Floehr calls for the standardization of methods, which he believes is imperative: “Experts from science, industry, civil society, government agencies, and aerospace must cooperate much more closely in order to create minimum standards for better comparability of data. Because without a basis for comparison, the data is not worth as much as it could be. If we managed to compare data from satellites with data from drones or airplanes and even with data from cellphones, we’d get much further.” The “Advances in Remote Sensing of Plastic Waste” study by the German Society for International Cooperation (Deutsche Gesellschaft für Internationale Zusammenarbeit, GIZ) recently confirmed that this is still one of the biggest hurdles. It also lamented the fact that research projects and studies often use very different metrics to identify, classify, and quantify waste.75 Source: DFKI Artificial intelligence is used in the PlasticObs_plus project to evaluate image data for plastic waste detection and classification in coastal areas. 50
RkJQdWJsaXNoZXIy ODI5MzU=