Research database

The AIMS project (Artificial Intelligence to Monitor our Seas) has the vision to develop and validate novel Artificial Intelligence (AI) algorithms to unlock the true potential of remote monitoring and enable a faster transition to a climate neutral society and economy: the AI algorithms will leverage the advantages of usual monitoring methodologies of the features of waves and offshore wind, and eventually overcome their intrinsic limitations. The value and resolution of sparse measurements of satellites and unevenly-distributed in-situ instruments will be increased, hence leading to a significant reduction of the cost and execution time of data collection, ultimately making knowledge wider and more accessible. The main field of application of AIMS is the sea, because it is a challenging environment where monitoring is difficult, long and expensive; a better knowledge of the ocean is paramount for climate science and contributes to fill knowledge gaps, as identified in the IPCC. In addition, ocean monitoring is crucial for the renewable energy sector, since offshore wind is one of the main pillars of the clean energy transition in Italy and Europe [9]. AIMS will enable much faster and cheaper surveys for site selection, permitting, design and monitoring, which are among the main obstacles that prevent the proliferation of offshore renewables. Albeit the main focus of AIMS is on the sea, the proposed AI algorithms are agnostic to the specific application, so impacts are expected across diverse fields and disciplines. AIMS considers and integrates heterogeneous data, coming from both satellites and in-situ measurements, making the developed methodologies flexible and transferable. AIMS achieves its objective via a multi-disciplinary approach, various sources of data and numerical techniques, focusing on wave height and period, and wind speed. A solid dataset will be built with satellite observations, consolidated numerical downscaling hindcasts, and in-situ measurements, including both moored wave buoys (spatial-static data in a grid layout), and gliders (spatial-varying data). Such a uniquely complete dataset, openly shared, will enable a pervasive development of a family of AI algorithms, highlighting their different fitness to various applications. AI algorithms will be developed for i) space and time gap-filling, and ii) inferring the wave period, not directly measured by satellites. A stage-gate approach will be defined for the development of AI algorithms, where quantitative performance metrics are evaluated to rank and discern among various candidates. Finally, active dissemination activities will ensure contamination in various fields of science and industry, to maximise the impact beyond the project. AIMS contributes to the PNRR in M2C2 accelerating the installation of offshore renewable energy (Investment 1.3 and Reform 1.1), and M1C2 fostering digitalization and the use of space assets and space economy (Investment 4).