Marine Data Observability using KPIS: An MDSE Approach

The 2023 climate change report states that the current temperature rise has led to recurring and hazardous weather events, devastating communities and the planet. Ocean observation systems and marine data generated by them are crucial for predicting these extreme events, understanding the ecosystem states, and regulating marine industries. Many regional and global initiatives have been supporting the collection and sharing of more data, filling gaps in ocean observation. However, some challenges can impact the quality of marine data at different points of data delivery pipelines: from acquisition and transmission at the Internet-of-Underwater-Things (IoUT) level up to storage and sharing. IoUT devices can have challenges due to limited battery, rough underwater terrain, error-prone wireless underwater communication, or low communication bandwidth to the cloud. Thus, mechanisms must be put in place to allow monitoring of data quality throughout the delivery pipeline, to optimize the usage of data and improve decision-making based on the data. This study explores observation of marine data quality on a data platform using Key Performance Indicators (KPIs). We have created a model of the platform and specified KPIs. Both are fulfilled by platform-collected data quality metrics, with the purpose to infer the state of the data in the platform over different periods. Our results show that the model-based implementation is able to function as a semantic translator between a metric monitoring toolkit and the platform objectives, integrating it into an observable subsystem for the overall middleware data platform.